Skip to main content

Full text of "DTIC ADA054705: Reliability and Maintainability Improvement Program for the F-106 Avionics."

See other formats


AD-A054 70b ARINC RESEARCH CORP ANNAPOLIS HD F/6 9/3 

RELIABILITY AND MAINTAINABILITY IMPROVEMENT PROGRAM FOR THE F-l— ETC(U) 
NOV 67 F09603-67-A-0003 

UNCLASSIFIED 711-01-1-850 NL 





AD A 05 47 0 5 



Prepared for 

Warner Robins Air Materiel Area 
Air Force Lofistics Command 
under Contract 
F09(603)-67-A- 0003-0001 




RESEARCH CORPORATION 


j document has bc-en approvtxj 
r public relcc-'t ar.d solo; itB 
.."ribution is unlimited. 




SECURITY CLASSIFICATION OF THIS PAGF. (When >>'„« Inter ed) 


RKAD INSTRUCTIONS 
Hr. FORE COMPLETING FORM 


REPORT DOCUMENTATION PAGE 


1. Kl PORT NUMUI H 


3. RECIPIENT'S CATALOG NUMBER 


2. GOVT ACCESSION NO, 


b. TYPE OF REPORT d PERIOD COVERED 


4. TITLE (and Subtitle) 


RELIABILITY AND MAINTAINABILITY IMPROVEMENT PRO- 
GRAM FOR THE F-106 AVIONICS 


PERFORMING ORG, 

711 - 01 - 1-850 


RCPOR r NUMBER 


8. CONTRACT OR GRANT NUMBER^?.) 


7. AUTHORfs; 


F09 ( 603 ) -67-A-0003-0001 


10. PROGRAM ELEMENT, PROJECT, TASK 


PERFORMING ORGANIZATION NAME ANU ADDRESS 

ARINC Research Corporation »/ 

2551 Riva Road 

Annapolis, Maryland 21^01 


AREA ft WORK UNIT NUMBERS 


12. REPORT DATE 

November 3, 1967 


CONTROLLING OFFICE NAME ANO ADDRESS 

Warner Robins Air Materiel Area 
Air Force Logistics Command 


15. SECURITY CLASS, (of this report) 


14. MONITORING AGENCY NAME & ADDRESSf// dllterent from Controlling Olltcc) 


Warner Robins Air Materiel Area 
Air Force Logistics Command 


UNCLASSIFIED 


DECL ASSI FI CATION ' DOWNGRADING 
SCHEDULE 


16 DISTRIBUTION STATEMENT (ol this Report) 


UNCLASSIFIED /UNLIMITED 


17. DISTRIBUTION STATEMENT (of the abstract entered In Block 20, If different from Report) 


18. SUPPLEMENTARY NOTES 


19. KEY WORDS (Continue on revorso side if necessary end identify by block number) 


20. ABSTRACT (Continue on reverse side II necessary and Identify by block number) 

ARINC Research conducted a l!*-month program to improve the reliability 
and maintainability of the F-106 avionics. Corrective actions were developed 
for human-factors and equipment problems Identified under previous contracts. 
The F-106 model analysis was updated to conform to the current equipment con- 
figuration. Analysis of the F-106 power subsystem resulted in recommendations 
for expanding test procedures and making minor modifications to improve unit 
reliability and performance. 


EDITION OF 1 MOV RMS OBSOLLTl 





Eoi 


LIABILITY AND 


IMPROVEMENT 
^ THE F-10 


i 


' INTAINABI LIT Y 
GRAM FOR 
VIONICS* 


IUX5* f' 


Prepared for 

Warner Robins Air Materiel Area 
Air Force Logistics Command 
Under Contract 
F09( 6 03 )-67-A-0003-0001 


ARINC RESEARCH CORPORATION 
a subsidiary of Aeronautical Radio, Inc. 
2551 Riva Road 
Annapolis—Jiaryland 21401 


Publics/ 


f711-il-l-85d 


I Jbia doi~i (|| | 

/ °r public relent# J??* 


too -W 





ARINC Research conducted a 14-month program to improve the reliability and 
maintainability of the F-106 avionics. Corrective actions were developed for 
human-factors and equipment problems identified under previous contracts. The 
P-106 model analysis was updated to conform to the current equipment configuration. 
Analysis of the F-106 power subsystem resulted in recommendations for expanding 
test procedures and making minor modifications to improve unit reliability and 
performance . 




r 





SUMMARY 

This report presents the results of a 14-month program conducted by ARINC 
Research to Improve the reliability and maintainability of the F-106 avionics. 

The work was performed for Warner Robins Air Materiel Area under Contract 
F09603-67 -A -0003-0001 . 

The principal activities of this program were as follows: 

• Development of recommendations for correcting equipment and human-factors 
problems defined during earlier contract activities 

• Updating the F-106 model analysis and the report on the quantification 
of F-106 avionics reliability and maintainability to conform to the 
current equipment configuration 

• Analysis of the power subsystem of the F-106 to define deficiencies and 
provide corrective recommendations where feasible 

Updating of the F-106 quantification report and the F-106 model analysis 
demonstrated that the change in equipment configuration as a result of the Group 31 
Interceptor Improvement Program did not degrade total system reliability. It also 
Identified specific units, added during the modification program, that could be 
improved to increase the rate of mission success. 

The findings of the power-subsystem evaluation include the following: 

• Correction and expansion of test procedures can reduce the number of 
serviceable units rejected during bench check and the number of un- 
serviceable units being bench-checked as serviceable. 

• Certain minor modifications can improve unit reliability, performance, 
and maintainability. 

• The addition of certain units and certain portions of the power sub- 
system to the periodic Inspection requirement can reduce total system 
failures . 


v 



CONTENTS 

Page 


ABSTRACT lii 

SUMMARY v 

1. INTRODUCTION 1 

1.1 Background 1 

1.1.1 Requirements of the Previous Contracts 1 

1.1.2 Requirements of Current Contract 2 

1.2 Scope of Report 2 

2. APPROACH 3 

2.1 ARINC Research Methods 3 

2.1.1 Basic Approach 3 

2.1.2 Reliability Prediction 3 

2.1.3 Computerized System Analysis 5 

2.1.4 Reliability and Maintainability Measurements 5 

2.2 Implementation 5 

2.2.1 Program Organization 6 

2.2.2 Data System 6 

2.2.3 Reports 6 

2.2.4 Coordination Activities 7 

3. REVISION OF MATHEMATICAL MODEL 9 

3.1 Introduction 9 

3.1.1 Goals 9 

3.1.2 Recapitulation of Earlier Work 9 

3.2 Selecting Standard Missions 10 

3.3 Determining the Requirements for Successful Phase and Function 11 

3.3.1 Functions Required for Phase Success 11 

3.3.2 LRUs Required for Function Success 13 

3.3.3 Failure Rates for Functions and LRUs 14 


vii 


l 


i 

I 












CONTENTS (continued) 

3.4 Determination of Reliability 

3.4.1 Reliability Profile for Mission A 

3.4.2 Reliability Profile for Mission B 

3.4.3 Reliability Profile for Mission C 

3.4.4 Reliability Impact of the Function 

3.4.5 A Theoretical Mission Reliability 

3.5 Conclusions 

3.6 Recommendations 


Page 

30 

30 

31 
31 
31 
37 

37 

38 


4. REVISION OF SYSTEM QUANTIFICATION 39 

4 . 1 Int roduc t ion 39 

4.2 Theoretical Reliability Characteristics 39 

4.3 Observed Reliability Characteristics 40 

4.3.1 Data Collection 40 

4.3.2 Data Organization 40 

4.3.3 Observed Reliability Data 40 

4.3.4 Theoretical and Measured MTBF 4l 

4.4 Observed Maintainability Data 4l 

4.4.1 MA-1 System Total Downtime 4l 

4.4.2 MA-1 System Repair Time 42 

4.5 Shop Repair Time for Line Replaceable Units 42 

5. INVESTIGATION OF POWER SUBSYSTEM 49 

5.1 Background 49 

5.2 General Description of the MA-1 Power Subsystem 50 

5.3 Investigation Methods 56 

5.3.1 Approach 56 

5.3.2 Equipment Used 59 

5.4 Findings Related to D-C Power 62 

5.4.1 Voltage Problems 63 

5.4.2 Unit Problems 84 





vlli 


CONTENTS (continued) 


Page 


5.5 

Findings Related to A-C Voltages 

89 


5.5.1 

Voltage Problems 

91 


5.5.2 

Unit Problems 

97 

5-6 

Recommendations 

114 

HUMAN FACTORS 

119 

6.1 

Introduction 

119 

6.2 

Problem Definition and Approach 

121 


6.2.1 

Problems Defined in Terms of Objectives 

121 


6.2.2 

Problems Defined in Terms of Elements Studied 

122 

6.3 

General Assumptions Limiting Objectives 

123 


5.3.1 

Previously Acquired Data 

123 


5.3.2 

New-Data Acquisition 

125 

6.4 

Analysis of Data 

125 


6.4.1 

Correlation of Factors Influencing System Effectiveness 

125 


6.4.2 

Analysis of Maintenance Actions 

127 


6.4.3 

Analysis of Pilot's Symptom Reporting 

127 

6.5 

Peripheral Influences 

127 


6.5.1 

Unofficial Ratings 

127 


6.5.2 

Outside Demands 

128 


6.5.3 

Spare -Parts Allocation 

128 


6.5.4 

New Test Equipment 

128 

6.6 

Conclusions and Recommendations 

129 


6.6.1 

Reduction of Number of Removals 

129 


6.6.2 

Replacement of Good Units 

129 


6.6.3 

Reduction of Repair Actions 

130 


6.6.4 

Reduction of Maintenance Events 

130 


6.6.5 

Delineation of Performance Goals 

130 


6.6.6 

Use of Test Equipment 

131 


6.6.7 

Validation of Recommendations 

131 


ix 




C ONTENTS ( continued ) 


7. SPECIAL TASKS 135 

7.1 Evaluation of the General Electric Rapid Tune Test Set 135 

7.1.1 Task Definition 135 

7.1.2 Task Summary 135 

7.1.3 Conclusions 135 

7.1.4 Recommendations 136 

7.2 Reliability Investigation of IFF Control Switch 136 

7.2.1 Task Definition 136 

7.2.2 Task Summary 136 

7.2.3 Recommendations 137 

7.3 Evaluation of the Fault Detection Tester (FDT) and IRAM 

Computer 138 

7.3.1 Task Definition 138 

7-3.2 Task Summary 138 

7.3.3 Task Findings 138 

7.4 Investigation of Ground Loops 140 

7.4.1 Task Definition 140 

7.4.2 Task Summary ....... . 2.40. 

7.4.3 Conclusions 140 

7.4.4 Recommendations 140 

7.5 Investigation of Rotary-Joint Failures 141 

7.5.1 Task Definition 141 

7.5.2 Task Summary l4l 

APPENDIX A - PART FAILURE RATES USED IN RELIABILITY PREDICTIONS FOR F-106 

ELECTRONIC SYSTEMS A -3 

APPENDIX B - AN OPTIMUM STRATEGY FOR SINGLE-UNIT SEQUENTIAL TESTS B-l 


1. INTRODUCTION 


This report presents the results of a fifteen-month program, performed 
under Contract FO 96 O 3 - 67 -A-OOO 3 -OOOI, to improve the reliability and maintain- 
ability of the F-106 avionics. The objective of these improvements was to 
increase the effectiveness of the F-106 Fighter Interceptor. 

1.1 Background 

The work completed in this program was largely a follow-on to work performed 
by ARINC Research under the direction of Warner Robins Air Materiel Area in two 
earlier contracts, AF09(603) -48024 and AF09(603)-60655*- These earlier contract 
activities are described briefly in this report to provide a better understanding 
of the more recent activities. 

1.1.1 Requirements of the Previous Contracts 

The general objectives of the earlier programs were to maximize F-106 
mission success, while minimizing repair time, at the lowest possible cost. 

These earlier programs required the following actions: 

• Investigate: 

(1) Reliability on the basis of frequency of component failures 
and unsatisfactory performance 

(2) Maintainability of components, subsystems, and overall 
avionics 

• Perform : 

(1) Deficiency area determinations, evaluations, and corrections 

through: (a) analysis of supplemental data and design, 

(b) redesign of circuits, and (c) improvement of maintenance 
techniques 

(2) Reliability and maintainability analyses of ECPs submitted 
by other contractors 

(3) Prototype and flight-test studies 


♦Final Engineering Report, Reliability and Maintainability Improvement Program 
for the F-106 Avionics, ARINC Research Publication 51B-01-2-o39, 1$ July 19bb . 


Evaluate : 


(1) Modification tests 

(2) Requirements for more effective support 

(3) Minimum MTEF requirements for proposed equipment 

(4) Impact of measured and predicted reliability values on 
mission success 

1.1.2 Requirements of Current Contract 

Under the current contract WRAMA, WRNE directed ARINC Research to 
"continue efforts initiated under Contract AF09(603) -48024 and AF09( 603) - 60655 . 
Equipments previously analyzed did not contain Group II IIP (Rapid Tune and 
Paramp). In view of location of contractor personnel at organizations new oper- 
ating with these modifications, the following services shall be accomplished 
against equipment containing Rapid Tune and Faramp: 

(a) Updating of Measured Reliability Data 

(b) Updating of Predicted Reliability Data 

(c) Updating of Reliability Model 

(d) Updating of Measured Maintainability Data 

(e) Provide corrective actions to improve reliability where appropriate 

(f) Review of maintenance procedures, analysis of adequacy of AGE and 
providing recommendations for improvement where appropriate . " 

1.2 Scope of Report 

This final report discusses the tasks performed by ARINC Research during 
this contract period . 

Section 2 outlines ARINC Research's approach, method, and reporting procedures. 
Supporting information and reliability standards are included. Problems associated 
with the implementation and quantification of the latter are given in Appendix A. 
Sections 3 and 4 present the complete results of updating tasks; Section 5> the 
power subsystem investigation; and Section 6 , the human- factors study. A method 
to determine optimum troubleshooting strategy for a particular symptom is presented 
in Appendix B. Seven special tasks were completed during the course of this pro- 
gram. Complete results of the efforts have been previously submitted to the con- 
tracting agency. These tasks are presented in Section 7. 


[ ‘ 1 

L . d 


} 

I 


H 


2. APPROACH 

2.1 ARINC Research Methods 

The scope of the P-106 contract allowed ARINC Research to use Its 
experience in reliability research. Some of the basic concepts Involved are 
discussed here. 

2.1.1 Basic Approach 

The five elements of the scientific method, ARINC Research's basic approach, 
are identified in Figure 2-1 and discussed below. 

Element 1, Problem Definition: Pertinent data are collected and analyzed 
to define specific problems. 

Element 2, Evaluation and Analysis: Once a problem is defined, more detailed 
analyses of the problem and its related equipment are performed to determine if 
a solution is possible. 

Element 3, Solution Derivation: Alternate solutions are evaluated and 
quantified to select the optimum corrective action. 

Element 4, Verification: The selected solution is first verified in the 
laboratory and then installed for operational testing. 

Element 5 , Refinement: During the course of the verification tests, the 
results are analyzed to determine whether the solution can be refined to achieve 
better results. 

2.1.2 Reliability Prediction 

Many reliability prediction techniques are currently in use by ARINC Research 
and others in the field. The technique used by ARINC Research for this program is 
consistent with the constraints and requirements of the program. Details of the 
prediction technique have been presented in special reports submitted by ARINC 
Research*, and additional information pertinent to this program is presented 
throughout this report. 

The principal purpose of predictions in the earlier program was to indicate 
problem areas rather than to quantify equipment reliability exactly. The purpose 
of prediction in this program was to provide a baseline for comparing the pre- 
Group II and the post-Group II equipments. 


*ARINC Research Publication 329-01-1-492, Quantified Reliability and Maintainability 
Characteristics of the F-106 AWCIS, 1 March 19&5 . 


3 


r 


Define Identify Problem Investigate Problem 

Hardware Areas Areas 

for Each k , 1 



Achievement 




























2.1.3 Computerized System Analysis 

ARINC Research has developed a Computerized Reliability Analysis Method (CRAM) 
for machine analysis of the reliability Indices of systems or subsystems. The 
CRAM program has the advantage of rapid execution once It Is set up, particularly 
for Iterative processes. By Its nature, the CRAM program Is less susceptible to 
the computational and translational errors usually associated with human processing 
of large quantities of data. 

In the earlier contracts, the CRAM program was used In the P-106 avionics 
system to determine the relative Influence of specific LRUs and their associated 
subfunctions on mission success. This approach was modified In the present appli- 
cation to provide the maximum Information while still taking Into consideration 
future changes In the P-106. 

2.1.4 Reliability and Maintainability Measurements 

The many problems associated with the collection and analysis of data 
to quantify reliability and maintainability Indices of military systems are well 
known to ARINC Research. The principal problem is lack of detailed information 
and inadvertant errors in the AFM 66-1 data records. With this in mind, ARINC 
Research field engineers are required to cross-check data forms and correct them 
where necessary before transmitting the data. This approach, with frequent moni- 
toring of the maintenance procedures, ensures a high-quality Input for data 
analyses . 

The quantification and Interpretation of reliability Indices are often 
misunderstood, usually because of the definitions of terms such as "time" and 
"failure" in a mean-time-between-f allure (MTBF) Index. This problem Is discussed 
in some detail jn ARINC Research Publication 518-01-2-639. 

The measured MTBF values were contrasted with predicted values for LRUs of 
Interest In the F-106 avionics. ARINC Research assumed that the predicted values 
represented an average, or state-of-the-art, value and developed a priority list 
for corrective actions. The desired result was thus a maximum increase In LRU 
(and system) MTBF for a given expenditure of resources. 

2 .2 Implementation 

The basic methods discussed above were implemented to the extent possible 
within the typical constraints, time and money, of the contract. In addition, 

ARINC Research operated on a noninterference and a nonduplicatlon-of-effort 
basis -- that Is, the activities of ARINC Research conformed to and in no way 


*ARINC Research Publication 

329-01-1-492, q 

.uantifled Reliability and Maintain- 

ability Characteristics of 

the F-106 AWCIS 

, 1 March 1985 . 


5 


Interfered with the operation of the military, and ARINC Research coordinated 
all actions with the contract administrator to ensure no Inadvertent duplication 
of efforts involving other agencies. 




2.2.1 Program Organization 

The basic program organization was substantially the same throughout the 
earlier two-year program and the present program. The program administration and 
primary engineering responsibility were at ARINC Research headquarters in Annapolis 
during both programs. The field-engineering and data-collection efforts were 
performed at Dover, Selfridge, and Langley Air Force Bases during the two previous 
contracts. The availability of modified aircraft dictated that Dover Air Force 
Base and Tyndall Air Force Base be used in the current program. 

2. 2. 1.1 Field-Base Operations 

ARINC Research personnel at the field bases operated under the technical and 
administrative control of the home office. At the F-106 operational bases these 
field personnel collected maintenance and operational data, verified them, and 
transmitted them to headquarters. In addition, they performed special tasks such 
as bench testing, developing proposed modifications, measuring temperature or 
performance, monitoring flight-tests, and interviewing operators and maintenance 
personnel. 


2. 2. 1.2 Laboratory Analysis 

Laboratory analysis of some of the equipment of the F-106 MA-1 AWCIS was 
necessary as part of the improvement program. This was performed in the ARINC 
Research laboratory at Annapolis, Maryland. The rest of the work performed at 
the ARINC Research laboratory included fabrication and checkout of the in-flight 
tape recorder used in the power- sub system study. 

2.2.2 Data System 

The data systems employed throughout the F-106 programs were based primarily 
on available data sources. For example, the reliability and maintainability data 
system was based on AFM 66-1 data. ARINC Research engineers checked the data for 
validity and completeness and, where required, supplemented the data with addi- 
tional information such as equlpment-operate time and aircraft flight time. 

2.2.3 Reports 

The progress of the program was reported by means of monthly and special 
reports. The frequency and format of these reports conformed with the preferences 
of the contract administrator. 



6 


2.2.4 Coordination Activities 


Administrative and technical liaison requirements were fulfilled through 
close contact of ARINC Research headquarters and resident personnel with the 
contracting agency and with other agencies participating in the overall F-106 
MA-1 AWCIS Improvement program. 

Administrative liaison was maintained with Warner Robins. Coordination with 
other contractors was maintained through the Warner Robins Service Engineering 
group. In addition, ARINC Research attended and participated in all Technical 
Advisory Group (TAG) meetings held since September 1964. 


7 


M* 




. ■■■ 



3. REVISION OF MATHEMATICAL MODEL 

3.1 Introduction 

Since July 1964 ARINC Research Corporation has been engaged In a program to 
Improve the reliability and maintainability of the MA-1 Aircraft Weapons Control 
and Intercept System (AWCIS) as used In P-106 aircraft. As an aid to this 
study, ARINC Research personnel developed a mathematical model* of MA-1 system 
reliability. Since then, however, the model has become obsolete as a reference 
standard because of extensive modifications to the avionics of the P-106. The 
current MA-1 system Is no longer the system portrayed In earlier reports. As a 
result. In August 1966 , Warner Robbins Air Materiel Area (WRAMA) directed ARINC 
Research to re-assess the MA-1. 

3.1.1 Goals 

This new task has three goals: (l) a comparative measure of system, sub- 
system, and unit reliability of the new (Group II) configuration; (2) establish- 
ment of a new reliability- index base line (based on the current MA-1 system); 
and (3) identification of those Line Replaceable Units (LRUs) within the various 
subsystems that with the least modification will provide the greatest potential 
gain In system reliability. 

3.1.2 Recapitulation of Earlier Work 

The following paragraphs summarize the results of previous work In which 
the original mathematical model was used. This work was assigned to ARINC 
Research in October 1964 by the P-106 Technical Advisory Group (TAG). Com- 
putations were based on ln-house data and data collected especially for the 
program. ARINC Research's Computerized Reliability Analysis Method (CRAM) was 
employed : 

(1) Functional reliability profiles for the F-106A weapons system were 
developed. This required selecting "standard" missions on the basis 
of their importance and being typical of their group. The times 
selected for the various phases were average rather than maximum or 
minimum figures; they were not related to the capability limits of 
the aircraft. 

(2) Functions (or elements) necessary for successful completion of each 
phase of each mission were defined. Here again "standard" requirements 
had to be selected from a great number of possibilities. If the 
requirement occurred rarely, the function was not included. 

* Analysis of the Reliability Model for the MA-1 AWCIS of the F-106A Aircraft . 

ARINC Research Publication 329-01-1-491, 1 March 1965 . 


f * 1 

f * i 


i , 


i/ 


% 


t 


(3) The LRUs Involved in each function were defined, and the LRUs required 
for mission success were identified. Only LRUs in which a substantial 
portion of the circuitry was related to the function were Included. 

Units contributing only small items (e.g., one relay contact) were 
not included in that function. , , 

During the unit -identification effort, it was found that approximately 
13 LRUs were common to more than one function. These LRUs were 
separated from their functions during the computer program and treated 
as a function. For calculations in which relative answers were suf- 
ficient, these LRUs were not separated, thus increasing the speed of 
the calculations. 

(4) Operation and failure characteristics were established for each LRU. 

The only functions with redundant IHUs were function U (command UHF) 
and function ADF (Automatic Direction Finder). For all other functions, 
the probability of success was defined as the product of the prob- 
abilities of success for the individual LRUs that comprise the function. 

(5) Boolean expressions for mission success were devised. Symbols were 
assigned to each function, and subscripts were added to the symbols 
to denote the phase. These expressions were solved to obtain the 
probability of mission success. 

(6) The results were analyzed. 

3.2 Selecting Standard Missions 

The following were selected (with the approval of the Air Defense Command) 
as standard missions: 

Mission A - a high-altitude intercept mission (also called a standard 
recommit mission). Two different targets are intercepted. 

Mission B - a low-altitude intercept mission. Two different targets 
are intercepted. 

Mission C - a flush-intercept mission, combined with two standard recommit 
missions (similar to mission A). There are three stages in 
this mission: 

(1) An aircraft takes off, hovers for a certain time, and lands. 

(2) It takes off again for a high-altitude intercept mission 
and lands. 

(3) It takes off again for another high-altitude Intercept 
mission and lands. During the landings, fuel and armament 
may be replaced; but it is assumed that no malntsnance 

is performed and no failures occur. Ground time during 
Intermediate landing is not counted in the cumulative 
phase time. 


i 

I 


.1 

I 

( 

! 


! 


i 

I 

i 

l 



10 




Each mission is divided into several phases; and each phase requires 
specific functions to be successful (see Section 3.3). The time period for each 
phase was selected from numerous possibilities. Six experienced pilots at Dover 
Air Force Base assisted by completing questionnaires indicating the maximum time, 
minimum time, and average time for each phase in each mission. The time for each 
phase used in the calculations was the mean value of the typical time on the 
questionnaires . 

Tables 3-1, 3-2, and 3-3 list the mission phases, the time for each phase, 
and the cumulative phase time for each of the three missions. 


3.3 Determining the Requirements for Successful Phase and Function 

3.3.1 Functions Required for Phase Success 

It is difficult to determine which functions are essential to the success of 
each phase. Under some circumstances, every function is necessary for phase 
success; under other circumstances, none of the listed functions influence phase 
success, although function A (armament) appears absolutely necessary to the 
launch phase. As a compromise measure, only those functions necessary for 
launching all weapons in the majority of possible circumstances (especially 
those most useful in practice flights for which substantial data are available) 
are included. 


It would not have been surprising if different missions required different 
functions for corresponding phases. For example, a low-level mission might 
require different equipment from that required by a high-level mission. The 
calculations were simplified, however, when it was found that corresponding phases 
required identical functions regardless of the mission. 


3.3.1.1 Function Symbols 

Symbols were assigned to each of the functions, to permit a concise pre- 
sentation of functions in the tables, and to facilitate working with the Boolean 
expressions. Table 3-4 lists the symbols for the functions. 


3. 3. 1.2 Boolean Expressions 


As an example of a Boolean expression, T^ represents the premise that the 
function T (TACAN) operated successfully through phase 15 . This type of notation 
(a capital letter to indicate the function and a numerical subscript to represent 
the phase) is used in a Boolean expression to describe a mission in mathematical 
terms. Table 3-5 shows each phase of Missions A and B and the corresponding 
Boolean expression. By arranging the 15 expressions according to the rules of 
elementary mathematical logic, it is also possible to show in Table 3-5 the con- 
cluding Boolean expression that can be equated to mission success. 


11 


TABLE 3-1 

MISSION A PHASE TIME 


TABLE 3-2 

MISSION B PHASE TIME 





Phase 

Phase 

Time 

(Minutes) 

Cumulative 
Phase Time 
(Minutes) 

Climb 

11 

11 

Vector 

15 

26 

Offset 

0 

26 

Acquire 

3 

29 

Track 

2 

31 

Launch 

0 

31 

Pullout 

1 

32 

Vector 

14 

46 

Offset 

0 

46 

Acquire 

3 

49 

Track 

2 

51 

Launch 

0 

51 

Pullout 

1 

52 

Return 

17 

69 

Land 

6 

75 





Phase 

Phase 

Time 

(Minutes) 

Climb 

4 

Vector 

15 

Offset 

0 

Acquire 

3 

Track 

1 

Launch 

0 

Pullout 

1 

Vector 

13 

Offset 

0 

Acquire 

3 

Track 

1 

Launch 

0 

Pullout 

1 

Return 

17 

Land 

6 


Cumulative 
Phase Time 
(Minutes) 



TABLE 3-3 


MISSION C PHASE TIME 



Phase 

Cumulative 

Number 

Phase Time 

Phase Time 


(Minutes) 

(Minutes) 


Climb 

Loiter 

Return 

Land 

Climb 

Vector 

Acquire 

Track 

Launch 

Pullout 

Vector 

Acquire 

Track 

Launch 

Pullout 

Return 

Land 

Climb 

Vector 

Acquire 

Track 

Launch 

Pullout 

Vector 

Acquire 

Track 

Launch 

Pullout 

Return 

Land 


TABLE 3-4 


FUNCTION SYMBOLS 


Function 

Symbol 

TACAN 

T 

Command UHF 

U 

Data Link 

D 

Computer 

C 

MA-1 Power 

P 

Identification Friend or Foe (IFF) 

I 

Search Radar 

SR 

Track Radar 

TR 


Infrared 

Armament 

Automatic Direction Finder 
Instrument Landing System 
Flight Control and Measurement 


\U AVAILABLE COPT 





























BEStlAVAlLABLE .COPY 


TABU 3 5 

BOOLEAN EXPRESSIONS FOR MISSIONS A AND B 

Phase 

Number 

Phaae 

Boolean Expression 

1 

Climb 

Pi 4 Fi 4 Ci t Ii 4 (U or D) x 

2 

Vector 

P 2 Pj 4 C 2 Sc 1 2 4 U 2 4 Dj 2 4 SR 2 ^ IR 2 Sc T 2 

3 

Offset 

P 3 4 F a 4 C 3 4 I 3 4 U 3 4 D a 4 SR 3 4 IR 3 Sc T 3 

4 

Acquire 

P« 4 P 4 4 C« 4 I 4 4 U« 4 D« 4 TO* 4 IR« 4 T 4 

5 

Track 

Ps Sc P 5 4 C 5 4 TR 5 4 IR 5 4 A 5 

6 

Launch 

P 0 4 F a 4 C a 4 TR« 4 XR« 4 A e 

7 

Pullout 

— 

8 

Vector 

Pa 4 Pa 4 Ca Sc la 4 Ua 4 Da 4 SRa 4 XRa 4 Ta 

9 

Offset 

Pa 4 P 9 4 Ca 4 la 4 Ua 4 Da 4 SRa 4 IRa 4 Ta 

10 

Acquire 

Pi o 4 Pio 4 Ci o 4 I 10 4 Uio 4 Dio 4 TRio 4 IRio 4 Tio 

11 

Track 

Pii 4 Pn 4 Cn 4 TRu 4 IRu 4 An 

12 

Launch 

Pi 2 4 Pi 2 4 Ci 2 4 TRia 4 IR 12 4 A 12 

13 

Pullout 

— 

14 

Return 

Pi 4 * Fi 4 * Ii 4 * [(ADF & U) or (D Sc C) or (SR & C) or (T & C)]i 4 

15 

Land 

Pis & Pis Sc iis & Uis Sc [ (SR & C) or (T & C) or L]i s 

Success 

= Pi 5 * 

Fis Sc I 15 & U 15 Sc [(SR & C) or (T & C) or L]i S & 


[ADF or [C & (D or T or SR)]}i 4 Sc Ciz & TRiz Sc IRi 2 Sc 


A 12 4 Dio 4 Tio 


1 


II 


3 . 3 . 1 . 3 Tabu la t Io n of Functions and Phases 

Table 3-6 lists the combination of functions required for successful com- 
pletion of each phase of the missions. The ampersand (&) Indicates that the 
requirement applies to both of the functions It connects, while the "or" indi- 
cates that the requirement applies to either one of the functions. The entire 
expression within the parentheses is affected by the symbols (&, or) preceding 
the parentheses . 

3.3.2 LRUs Required for Function Success 

Each of the functions depends on the successful operation of its LRUs. Some 
LRUs are necessary for more than one function; others are not necessary to any 
function (for the completion of the missions). Many of these unnecessary LRUs 


13 








% 


; 


t 





are used only for testing during maintenance and are not shown In the reliability 
model for an operating MA-1 system. Other LRUs normally listed In F-106 docu- 
ments but omitted from this model are those used only on aircraft with the cockpit 
configuration known as "round." Other aircraft have a cockpit configuration known 
as "vertical." This study used the vertical configuration, which Is 'common to 
80 percent of the F-106 aircraft. Table 3-7 lists the LRUs Intentionally omitted 
from all functions. 


TABLE 3-6 

RELATIONSHIP OF FUNCTION TO MISSION PHASES 

Phase 

Required Functions 

Climb 

P&F&C&I& (U or D) 

Loiter 

P&F&C&I& (U or D) & T 

Vector 

P&F&C&I&U&D&T&SR&IR 

Offset 

P&F&C&I&U&D&T&SR&IR 

Acquire 

P&F&C&I&U&D&T& TR & IR 

Track 

P&F&C&TR&IR&A 

Launch 

P&F&C&TR&IR&A 

Pullout 


Return 

P & F & I & (U& ADF) or (D & C) or (SR & C) 


or (T & C) 

Land 

P & F & I & U & L or (SR & C) or (T & C) 


3.3.3 Failure Rates for Functions and LRUs 


Tables 3-8 through 3-20 list the LRUs for the 13 functions. Tables for 
those functions not affected by the Group II program* are shown as they appeared 
In the original report. They list the observed failure rate for each of the 
LRUs, the percentage of the function failure rates, and the possible Improvement 
ratio for each LRU. The tables for functions affected by the Group II program** 
present data on both the observed and predicted failure rates for comparisons of 
the Pre- and Post-Group II equipment configurations. Each function and the related 
data (Tables 3-8 through 3-20) are discussed In detail In Section 3.4.4 of this 
report . 


* Tables 3-8, 3-9, 3-10, 3-13, 3-l8, and 3-19- 


** Tables 3-11, 3-12, 3-14, 3-15, 


3-16, 3-17, and 3-20. 



j 

1 

] 



14 






TABLE 3-7 

LRU'S NOT INCLUDED IN THE FUNCTIONS 


LRU 

Part 

Number 

Remarks 

Digital Computer Test Set 

464296 

Used for self-test 

Radar Test Set No. 1 

464096 

Used for self-test 

Photographic Recorder 

464149 

Used for self-test 

Radar Test Set No. 2 

464196 

Used for self-test 

Electrical Equipment Rack, Photo 
Recorder 

464702 

Used for self-test 

Dehydrator Rack 

464674 

Similar to P/N 464774 

Rate Gryoscope Transmitter 

TRU/2/A 

Replaced by P/N 

131311-01 

Tactical Display Horizontal 
Situation Indicator (HSI) 

464180 

Replaced by P/N's 

464181 and 464920 

Vertical Situation Amplifier 

464306 

Round configuration only 

Air Data Signal Data Converter 
(SDC) 

464823 

Round configuration only 

Command and Target Altitude 
Indicator Assembly 

464980 

Round configuration only 

Mach Number Indicator Assembly 

464880 

Round configuration only 

Bar Setting Analog Signal Control 

464259 

Round configuration only 

Static Pressure and Angle- 
of-Attack Air Data Compensator 

464521 

Round configuration only 

SDC - Flight Director 

464720 

Round configuration only 

Electrical Equipment Rack - 
Flight Director 

464511 

Round configuration only 

Horizon Indicator 

329B3 

Round configuration only 

Course Indicator 

331A3 

Round configuration only 

Steering Computer 

562A3B 

Round configuration only 

Radio Frequency (RF) Transmission 
Line Switch 

464263 

Disconnected 

Liquid Level Indicator 

464281 

Used for self-test 

Airborne Moving Target Indica- 
tion ( AMTI ) Video Amplifier 

464495 

Disconnected 

AMT I Signal Comparator 

464150 

Disconnected 

Pressure Meter 

464024 

Used for self-test 

AC-DC Generator 

464089 

Replaced by P/N 464056 

Voltage Regulator Assembly 

464992 

Similar to P/N 464892 

Infrared (IR) Range Unit 

464746 

Not necessary In this 
model 


Non-Computing Fixed Sight 


464169 


Not used In these 
missions 







TABLE 3-8 

FAILURE RATES FOR TACAN LRU'S (FUNCTION T) 


<D 

-p 

c 

rH 

o 

X) 

e= o 

•H 

<V -H 

co 

> P 

co 

o cd 

o 

P PC 

CL, 

P. 


5 



I 4 CO 4 OrHOI I I ^3- t — -=J- I 


H h 'O CM CVJ CVJ 
OJ 


O 




73 -P 

o 

C\J CO 

lPi 

d) cd 

o 

00 rH 

Cvj 

■p pc 
o 

o 

rH i — 1 

f- 

•H 0) 

1 — 1 

O '-O 

-=4 

'O 

73 P 


00 rH 

OJ 3 




P rH 


CM 


(X «H 


rH 


Cti 









^ 4 co w s o\ ia 






-=t 

rH 

O 

-=t 

'O 

on 

-=* 

•o 

on 

CO 

on 

rH 

on a 

1 

o 

m 

-3* 

ON 

O 

00 

on 

CM 

m 

00 

o 

o 

o o 


OJ 




rH 

rH 



rH 
















4- 




LT\ 

-=t 

on 

in 

ON 

ON 

ON 

r- 

rH 

-=T 

o 

CM 

oj m 


O') 

CO 

CM 

-4* 

t— 


00 

vo 

o 

on 

■3- 

-=t 

CM rH 



ON 

in 

o 

CM 

vd 

CM 


s 

VO 

r- 

ON 

oo o 

* 

in 

OU 

rH 

on 

m 

CO 

ON 


in 

CM 

rH 

-=4 LfN 

* 

c^ 


rH 

Cvj 

on 

o 

On 

CM 

■=t 

co 

CM 

rH 

h- CM 


*\ 

•» 


•v 

•» 


•» 

•» 

•* 

•* 





c- 

-=f 


CO 

ON 

VO 

CM 

CM 

on 






rH 





rH 



rH 


















rH -=t 
rH C\ 


on 

■=t 

in 

ON 

ON 

ON 

o 

o 

rH 

o 

CM 

o 

on 


OJ 

f- 

o 

CM 

CM 

CM 

CVJ 

CM 

co 

CM 

o 

m 

CM rH 

CO 

rH 

o 

-=t 

on 


CM 

in 

VO 

rH 

ON 

VI) 

CM 

1 1 

-=t 

-=t 



-=t 

•=* 

■=* 

■=? 

-=* 


-=t 

-4- 

i 

CM CM 

0J 

VO 

VO 

VO 

VO 

VO 

VO 

vo 

MJ 

VO 

VO 

0 

CM CM 

CM 

-=t 

•=t 



-3- 

-=t 



■3* 


-=t 

in in 


P 














o 









P 





-p 

d 




U 

a> 


8 8 


d) 

P 





to 



P 

p 


co 

co 


p 





o 


p 

0) 

p 




p 





p 


o 

p 


p 

rH 

CM 


> 





p 


p 

p 

£ 

<D 



H 

c 





<d 


O 

•H 

CO 

Ti 

• 

• 

a 

o 





p 


<D 

E 

§ 

o 

o 

o 

o 





c 


rH 

to 

o 

s 

s 







H 


<D 

§ 

p 




>> 






1 


CO 

Eh 

p 

>> 

>> 

cd 

<d 


p 



p 



p 


o 

aJ 

cri 

rH 

rH 


O 



a) 


rH 

Eh 

to 

p 

rH 

rH 

a 

a 


P 

p p 


> 


(1) 


5 

aJ 

a 

a 

CO 

CO 


cti 

O 0) 


•H 


c 

a; 

rH 

CO 

co 

•H 

•H 


o 

P «H 


0) 


§ 

hO 

p 

d 

•H 

•H 

Q 

Q 


•H 

cd <p 


O 


§ 

cd 

■g 

Q 

M 




Ti 

O «H 


a) 


£ 

& 



rH 

rH 


c 

•H rH 

cd 

cn 


O rH 

PC 

s 

P 

P 

cd 

ed 


FH 

T3 Ph 

G 



O 




•H 

•H 

o 

o 



0 1 

a> 

§ 

a 


§ 

§ 

§ 

8 8 

*H 

P 

•H 

P 


§ 

p 

o 

o 

o c 

o 

Q 

o 

o 

O 

O 

O 

o 

O 

M M 

c 

< 

cd 

< O 


< 

< 

o 

o 

cd 

cri 

£d 

< 

s a 

5 

EH 

PS 

EH O 

Eh 

EH 

EH 

o 

o 

EH 

Eh 

PC 

Eh 



C'-LrvONonoooNvooocorn ro 

fflCO'OHOWHHin'OtO VO 

,-h in o .=4 on cnj t— on^a- in 





P *H 

<U <T5 

* § 


t Failure rates were taken from AFLC D056B-2 of 18 November 1964 (6 months) and modified 
to 80 # because of use in only 80 $ of aircraft that have a vertical configuration. 















TABLE 3-9 

FAILURE RATES FOR COMMAND UHF LRU'S (FUNCTION U) 



Part Number 


UHF Receiver, or 

Time Division Data Link 
(TDDL) Receiver 

Frequency Selector Control 
Power Supply 
UHF Transmitter 
Rack 

Audio and ADF Electronic 
Control Amplifier (ECA) 

UHF Antenna and Lead 

Headset, Microphone, 
Personal Leads 



46*1059 

464074 

464606 

74524 

96100 


Observed 
Failure Rate 


50,505.05 

8,230.45 

13,468.01 

11,597.46 

29,554.81 

4,489.34 

3,367.00 
9,523.81 
285.71** 


74,000.00 




Percent of 
Total 

Failure Rate 

Predicted 
Failure Rate 


Possible 

Improvement 

Ratio 



11,036.96 

8,676.04 

1,460.76 
3,652.44 
8,750.06 
8 


4,946.78 



* These represent the calculated contribution to apparent overall failure rate considering 
the redundant configuration. 

** Failure rates were taken from AFLC DO56B-2 of 18 November 1964 (6 months). 



TABLE 3-10 

FAILURE RATES FOR DATA LINK LRU’S [FUNCTION D| 


Part Number 


TDDL Antenna 
TDDL Receiver 

TDDL Converter Receiver 
Control 

TDDL Digital -Digital 
Converter 

Rack 

Rack 

Computer Mode Annunciator 

Auto Navigation Selector 
Control 


464019 

464220 

464811 

464074 

464034 

464955 


Observed 
Failure Rate 

Percent of 
Total 

Failure Rate 

Predicted 
Failure Rate 

337.84* 

1.8 

10.00 

8,230.45 

42.8 

8,676.04 

2,618.78 

13.7 

1,127.89 

2,618.78 

13.7 

21,509.59 

77.96* 

0.4 

6.00 

4,489-34 

23.4 

161.18 

467.73* 

2.4 

27.55 

374.11 

1.8 

17.63 

19,214.99 


31,535.88 


Possible 

Tiprovement 

Ratio 



* Failure rates were taken from AFLC D056B-2 of 18 November 1964 (6 months). 



























TAKE 3-11 

FAILURE RATES FOR COMPUTER LRU'S [FUNCTION C) 




I 


I 




Pre-Group II 

Post-Group II 

LRU 

Part 

Number 

Observed 
Failure Rate 

Predicted 
Failure Rate 

Observed 
Failure Rate 

Predicted 
Failure Rate 

AC Inputs SDC 

464023 

3,367.00 

1,314.96 



Analog-Digital Signal 

Comparator 

464050 

3,741.11 

1,626.98 



Digital Output Phase Change 

Relay Assembly 

464064 

5,237.56 

7,143.42 



Digital Computer Inter- 
connection Box 

464318 

1,117.32* 

715.^7 



Analog Signal Sampling 

Electronic Switch 

464051 

6,734.01 

4,994.07 



Analog Outputs SDC 

464123 

16,460.91 

9,972.57 



Input-Output SDC 

464255 

18,705.57 

18,184.11 



Digital SDC 

464323 

4,863.45 

8,412.51 



Control Electronics 

Digitizer Computer 

464446 

13,468.01 

16,584.21 



Arithmetic Electronic 

Digitizer Computer 

464146 

20,950.24 

22.8l4.48 



Shift Register Digitizer 

Memo ry 

464157 

2,992.89 

5,027.19 



Clock Pulse Generator 

464489 

1,122.33 

723.76 



Read-Write Memory Amplifier 

Assembly 

464457 

10,475*12 

8,717.26 



Data Storage Magnetic Drum 

464057 

5,237.56 

891.13 



Read Memory Diode Cate 

Assembly 

464657 

2,618.78 

7,365.56 



Electronic Equipment Rack, 

Right Hand (RH) Forward 

464273 

3,367.00 

248.50 

N V 

N/C 

Total 


120,458.86 

114,736.18 

120,458. 8b 

114, 73b. 18 


Failure rates were taken from AFLC DO56B-2 of 18 November 1964 (6 months). 


J 



TABLE 3-12 

FAILURE RATES FOR MA-1 POWER LRU S (FUNCTION P) 



Part 

Number 

Pre -Group II 

Post-Group II 

mo 

Observed 
Failure Rate* 

Predicted 
Failure Rate 

Observed 
Failure Rate* 

Predicted 
Failure Rate 

System Power Control 

464905 

3,741.11 

440.99 

13,503.20 

738.59 

Undervoltage-Overvoltage Relay- 
Assembly 

464062 

1,496.45 

2,545.41 



+28V and -140V DC Voltage 

Regulator Assembly 

464692 

2,992.89 

1,052.08 



Interconnection Box No. 1 

464018 

4,115.23 

1,499.04 



Interconnection Box No. 2 

464n8 

2,688.17* 

1,331.55 



AC-DC Generator 

31056 

4,306.63** 

145.08 



Power Transfer Relay Assembly 

464162 

2,992.89 

3,298.12 



-250V DC Power Supply 

464192 

2,618.78 

2,078.04 



400 cps and 1600 cps Voltage 
Regulator Assembly 

464892 

6,734.01 

2,291.10 



100 Millihenry Reactor 

464135 

298.60* 

2.97 



40 Millihenry Reactor 

464035 

l8l. 92* 

10.00 



DC Slip Ring Generator 

464689 

2,618.78 

100.00 



+300V and -150V DC Voltage 
Regulator Assembly 

464792 

2,244.67 

1,200.22 



+300V DC ?ow=r Filter 

464092 

1,496.45 

521.28 



+150V DC Power Filter 

464991 

748.22 

821.13 



+300V DC Power Filter 

464591 

5,237.56 

997.34 



+150V DC Power Filter 

464891 

1,122.33 

551.21 



-140V DC Power Filter 

464791 

2 ,092 .39 

395.77 



fioov and -140V DC Voltage 
Regulator 

464292 

4,863.45 

764.39 



-155/55, loOO cps, +300 

Reference Voltage Regulator 

464491 

9,352.79 

447.23 



+50V, -50V and -15V Transistor 
Pow n r Supply 

464326 

2,992.89 

945.35 

1,761.29 

N/C 

Comput-r Zl, +50V DC R"fer“nce 

464489 

1,122.33 

723.76 



Total 


66,959.04 

22,662.11 

75.489.53 

22,959.71 


Failure rates were taken from AFLC DO56B-2 of 18 November 1^64 (6 months;. 


Failure rate from minutes of the 11 September 1964 TAG meeting. 



































TABLE 3-14 

FAILURE RATES FOR SEARCH RADAR: LRU’S (FUNCTION SR) 




Pre-Group II 

Post-Group II 

LRU 

Part 

Number 

Observed 
Failure Rate 

Predicted 
Failure Rate 

Observed 
Failure Rate 

Predicted 
Failure Rate 

Master Synchronizer 

46*1003 

2,992.89 

2,684.59 



Radar Transmitter-Receiver 

464o65 

19,453.30 

18,281.89 

16,438.68 

12,345.76 

Directional Coupler 

464oS4 

1,496.45 

133.65 

1,761.29 

N/C 

Waveguide 

464oi6 

374.11 

505 . 00 

** 

254.52 

Waveguide 

464216 

374.11 

504.22 

587.10 

254 . 52 

Antenna 

464017 

11,971.57 

7,276.95 



Dehydrator 

464097 

7,518.80* 

- 



Rack 

464774 

207.90* 

- 



Compressor 

464o45 

4,363.45 

380.50 



Valve 

464107 

4,115.23 

104.30 



Mode Selection Radar Set Control 

464305 

374.11 

326.44 

318.02 

922.72 

Radar Set Control 

464855 

3,367.00 

2,091.29 

2,544.12 

1,743.00 

Amplifier Computer 

464241 

10,101.01 

5,885.11 



Scan Generator 

464663 

6,734.01 

7,078.24 



Antenna Servo Amplifier 

464206 

5,237.56 

2,802.76 



Elevation Drive Amplifier 

464506 

7,482.23 

2,209.44 



Azimuth Drive Amplifier 

464lo6 

4,489.34 

1,931.33 



1600 cps Filter 

464425 

374.11 

12.31 



Electronic Equipment Rack, 

Left Hand (LH) Forward 

464073 

5,985.78 

427.51 

** 

N/C 

Rack 

464173 

441.70* 

- 

** 

N/C 

Cooling Hat 

464190 

374.11 

- 



Intermediate Frequency (IF) 
Amplifier 

464295 

7,108.12 

1,542.37 

26,419.30 

1,546.55 

Video Amplifier 

464095 

9,352.79 

3,173.42 



Clutt :r Gates 

464082 

2,244.67 

3,058.87 

2,348.38 

3,051.60 

Sweep G'-n"rator-Ampllf ler 

464195 

8,230.45 

3,674.90 



Search and Attack Flight 

Indicator 

464080 

25,813.69 

3,943.67 



Rack 

464002 

374.11 

- 



Cathole-Ray Tub” (CRT) Light 

Filter 

464025 

1,013.17* 

- 



CRT Visor 

464125 

748.22 

- 



Indicator Sweep Generator 

464389 

3,367.00 

3,436.08 



TOTAL 


156,581.49^ 

71,464.84 w 


H 


Failure rates were taken from AFLC D056B-2 of l8 November 196^ (6 months). 

♦No failures repelled, therefore the Pre-Group II failure rate was used again. 






















TABLE 3-14 (Cutilii*) 




Pre-Group 

II 

Post-Group II 

LRU 

Part 

Number 

Observed 
Failure Rate 

Predicted 
Failure Rate 

Observed 
Failure Rate 

Predicted 
Failure Rate 


New Units Added During Group II Modification 


PA/RT Power Supply 

464026 

N/A 

N/A 

587.10 

1,842.35 

4 Port Circulator 

464432 

N/A 

N/A 

Est. 406.86 

220.00 

Waveguide Directional Coupler 

464484 

N/A 

N/A 

Est. 406.86 

110.00 

Waveguide Assembly Radar 

464516 

N/A 

N/A 

Est. 406.86 

394.00 

Computer Programmer 

464541 

N/A 

N/A 

2,935.48 

3,755.04 

AFC High Voltage Power Supply 

464641 

N/A 

N/A 

29,354.78 

6 , 016.17 

Low Voltage Power Supply 

464741 

N/A 

N/A 

1,761.29 

1,447.25 

Hydraulic Drive Unit 

464841 

N/A 

N/A 

4,109.67 

3,708.94 

Total 


156,581.49 

71,464.84 

212,549.02 

82 , 767.18 




















TABLE 3-15 

FAILURE FOR TRACK RADAR* LRU'S [FUNCTION TR) 


Total from Table 3-H 
Hand Control 

Antenna Tracking Amplifier 

Automatic Gain Control 
(AGC) and Angle Track 
Converter 

Torque Generator-Amplifier 

Radar Relay Switch Assembly 

Attack Display Amplifier 

Attack Display SDC 

Time -Sharing Electronic 
SDC 

HIG-4 Rate Gyro 

HI 3-4 Rate Gyro 

Range Track Synchronizer 

Accelerometer 

Steering Signal Amplifier 

Steering Signal Computer 


Part Number 


464083 

464141 

464020 

464041 

464063 

464395 

464223 

464523 

74148 

7414a 

464103 

464o6l 

464341 

464346 


Pre-Grouc II 

Post-G 

roup II 

Observed 
Failure Rate 

Predicted 
Failure Rate 

Observed 
Failure Rate 

Predicted 
Failure Rate 

156,581.49 

71,464.84 

212,549.02 

B2,7b7.l8 

8,978.68 

1,818.13 

6,042.30 

N/C 

6,734.01 

3,160.31 



22,446.69 

4,463.09 



2,244.67 

1,955.29 



1,496.45 

3,602.12 

1,174.19 

3,952-57 

4,489.34 

5,236.41 



2,992.89 

3,067.66 



10,475.12 

6,388.78 



753.58** 

490.00 



259.80** 

490.00 



19,827.91 

3,674.32 



1,122.33 

637.72 



4,115.23 

3,144.92 



8,230.45 

4,303.04 

5,283.86 

4,321.39 

250,748.64 

113,896.63 

300,510.94 

125,607 . Tt 


* In addition to the LRU's listed above. Track Radar also includes those LRU's listed 
for Search Radar (Table 3-U) • 

** Failure rates were taken from AFLC D056B-2 of 18 November 1964 (6 months). 










TABLE 3-16 

FAILURE RATES FOR INFRARED LRU'S (FUNCTION IR) 




Pre-Group II 

Post-Group II 

LRU 

Part Number 

Observed 
Failure Rate 

Predicted 
Failure Rate 

Observed 
Failure Rate 

Predicted 
Failure Rate 

Nitrogen Tank 

464099 

9,726.90 

159.11 



Rack 

464902 

207-90* 

10.00 



Infrared Receiver 

464767 

4,115.23 

2,096.97 



Rack 

464802 

77.96* 

0.83 



Hose 

464399 

- 

- 



Special Cable 

464499 

2,992.89 

- 



IR SDC 

464040 

6,359.90 

5,668.67 



Passive Detector Synchro 

Signal Amplifier 

464441 

8,604.56 

3,918.92 



IR Relay Switch Assembly 

464663 

6,734.01 

7,078.24 



Indicator Video 

464195 

8,230.45 

3,674.90 



Indicator 

464080 

25,813.69 

3,943.67 



Audio and ADF ECA 

464606 

3,367.00 

4,946.78 



Headset and Personal Leads 

96100 

285.71* 

- 



Hand Control 

464083 

8,978.68 

1,818.13 

6,042.30 

N/C 

Mode Operations Set Control 

464855 

3,367.00 

2,091.29 

2,544.12 

1,743.00 

Antenna Control Amplifier 

464241 

10,101.01 

5,885.11 



Attack Display Amplifier 

Assembly 

464395 

4,489.34 

5,236.41 



Attack Display SDC 

464223 

2,992.89 

3,067.66 



Steering Signal ECA 

464341 

4,115.23 

3,144.92 



Time Sharing SDC 

464523 

10,475.12 

6,388.78 



Steering Signal Computer 

464346 

8,230.45 

4,303.04 

5,283.86 

4,321.39 

Total 



63,433.43 

122,560.07 

63,103.49 


Failure rates were taken from AFLC D056B-2 of 18 November 1964 (6 months). 











TABLE 317 

FAILURE RATES FOR ARMAMENT LRU'S (FUNCTION A) 




Pre-Group II 

PosWJroup II 

LRU 

Part Number 





Hand Control Flight Stick 

464083 

8,978.68 

1,818.13 

6,042.30 

N/C 

Armament Self-Test Panel 

464596 

. 748 . 22 

274.75 



Armament Test Relay 

Assembly 

464446 

13,468.01 

16,584.21 



Parameter-Setting Relay 

Assembly 

464364 

1,122.33 

5,346.17 



Launcher 

464054 

748.22 

1,000.40 



Inte rvalomete r 

299 

779.42* 

- 



Mode-Select Relay Assembly 

464464 

1,870.56 

4,961.93 



Armament Control Relay 

Assembly 

464264 

3,367.00 

6,844.84 



Transmitter-Tuning ECA 

464866 

5,611.67 

2,448.61 

7,045.15 

3,505.01 

Missile Automatic Flight 

Control ^AFC) Channel 

Selector 

464043 

2,992.89 

3,427.59 



Missile Antenna Test ECA 

464366 

748.22 

3,089.09 



Missile Antenna ECA 

464266 

748.22 

2,409.97 



Gyro Power Control Panel 

464008 

748.22 

712.33 



Armament Control Powei 

Supply 

464087 

2,618.78 

1,793.84 



Armament Control Relay Box (ACRB) 

RY437A 

545.85* 

- 



Air Control Timer (ACT) 

T228a 

25.99* 

- 



Armament Control Panel (ACP) 

8-62221 

1,428.57* 

- 



Special Weapons Rack (75311) 

8-57200 

6,410.26* 

- 



Total 


52,961.11 

50,711.86 

51,458.21 

51,768.26 


* Failure rates were taken from AFLC D056B-2 of 18 November 1964 (6 months). 





















p 

a) c 

P 05 
43 E O 
P 05 P 
m > p 
05 o as 
o DC 
cl, a 




VO-afOOOOOOVOCMO 

OVOC^-t'-PVOOO 


ro o 
co ro in 


VO VO O VO rl (7l ID O I CO CO CO 

(nt-N40WHH LOCO VO 
o 10 to m h w ^ m 4 t- 



riWlftritO^COWCO 
• •••••••• 

P O CO O CO O £ — CM O 
H H CO 










* 











* 




in 

in 

rH 

0 


rH 

- 3 - 

0 

CO 

CM 

co 

0 


1 — 1 

0 

CO 

0 

co 

- 3 - 

c- 

CM 

rH 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

in 

0 

- 3 - 


< 7 \ 

CO 

vo 

r— 

CO 

CO 

0 

0 

no 

c- 

vo 

CO 

VO 

in 

CM 

rH 

■=fr 

in 

in 

cm 

CO 

ro 

• 3 * 

-=t 

co 

C\J 

VO 


CM 

•\ 



•* 

•* 

•s 

•v 


•v 



0 

CO 


CO 


CO 

c— 


OJ 



in 





rH 















rH 

-=t 










rH 

(TV 

c- 

f- 

c- 

vo 


rH 

0 

CM 

OJ 


CO 

00 

VO 

rH 

0 

r- 

CO 

OJ 

O 

0 

C\J 

1 

rH 

1 — 1 

VO 

rH 

vo 

0 

rH 

ov 

VO 

rH 

1 

-=t 

•=t 


-=t 


-=t 

-=t 

-=t 

-=t 

CM 

CM 

VO 

vo 

vo 

vo 

VO 

vo 

vo 

VO 

VO 

CM 

CM 


-=t 

-=t 



- 4 - 


■=t 

-=t 

in 

in 







u 











05 











P 

































> 










H 

c 










3 

0 

0 








< 







u 



0 


>> 

>> 





0 



w 


aj 

aj 










rH 

rH 





•0 



Dh 


a 

a 




u 

Ch 

05 


S 


ra 




0 

<u 

<D 

> 

al 


«H 

•H 



■p 

•H 

> 

•H 

§ 


« 

0 



c 5 

p 

•H 

05 

T) 






O 

•H 

05 

0 

0 ) 



rH 

rH 



p 

rH 

0 

05 

-p 


ct 5 

CtJ 




p. 

<D 

DC 

£ 



O 

0 



c 

g 

DC 


O 


•H 

•H 



H 



>0 


•H 


p 

p 

X 

a 



Ph 

Q 


T 3 

O 

O 

O 

0 

O 

M 

P 

B 

O 

tr> 

§ 


a 

cc 

ctf 

EH 

<ti 

EH 

a 

PC 

aj 

PC 

a a 




jC 


0 


a] 


05 


c 


P 


0) 


05 


a) 


4C 

• 

a 

■■ — v 


05 

to 

43 

c 

P 

p 

C 

s 

O 

0 

6 

p 


p 

VO 

0 

' — *■ 

p 

-3- 

05 

VO 

43 

CTV 

p 

P 

c 


p 

0) 


43 

T3 

i 

05 

0) 

u 

> 

p 

O 

3 

z 

O' 


05 

CO 


1 — 1 

05 

p 

u 

0 

a5 

CM 

>» 

0) 

(h 

43 

vo 

P 

in 

05 

8 

05 


3 

8 

0 

O 

§ 

O 


*3 

§ 

CD 

fi 

TJ 

p 

3 


rH 

c 

O 

0) 

G 

.* 

•H 

a) 


P 

-P 


O 

G 

g 


0) 


s 

G 


cti 

05 


05 

CO 

P 

— 

aJ 

£> • 

h 

3S 

05 

0) § 

3 

05 P 

P 

0) 05 

P 

c. c 

aj 

E-t P 

a 

* 

* 


* 













-p 
0 ) C 

rH CJ 

f E O 

•rH d *H 
W > -P 
W O id 
O Pi K 
Pi & 


rH 

iH 

rH 

rH 

rH 

H 

rH 

• • 

• • 

• • 

• • 

• • 

• • 

• • 

00 

in 

CO 

VO 

m 

co 

rH 

CO 

00 

t- | 

OJ 

CO 

rH 

t- 

• 

• 

• 

• 

• 

• 

• 

CVI 

m 

CO 

rH 

c- 

o\ 

in 

VO 




CV) 

CVJ 




o 

•d -p 

o 

-=t 

in 

co 

CO 

00 

CVJ 


CVJ 

<D (0 

o 

CO 

•=t 

C'j 

CVJ 

rH 

ON 

o\ 

o\ 

■P K 

• 

• 

• 


• 

• 

• 

• 

• 

o 

o 

rH 

VO 

vo' 

o 

rH 

1 00 

CO 

CO 

•H <D 

rH 

OA 

o 


t- 1 

i vo i 


o 

CTv 

•d Pi 

£3 

P-< *H 


VO 

•» 

CO 

VO 

* 

co 

o\ 

•v 

■ 3 - 

VO 

CVJ 

rH 


CO 

•t 

rH 

CO 

•s 

VO 

rH 


, 41 

<P +5 

O cd 

K 

■P rH 

CO 

CO 

in 


•=* 

in 

in 

0^ 

H 

OJ 

C cd oj 

CVJ 

o 

in 

VO 

vo 

o 

co 

rH 

in 

CO 

d 43 fn 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

o o d 

rH 

CO 

VO 

vo 

vo 

o 

CO 

rH 

o 

o 

U fcl rH 
d -H 
Ph d 


CVJ 

cvj 







CVJ 


d 

43 

n cd 

CO 

•st 

rH 

o 

o 

* 

G\ 


* 

CO 

* 

o 

* 

* 

CO 

CVJ 

> 

o 

a 

> d 

o > 
S5 cd 

.c 

4) K 

CO 

CVJ 

o 

o 

o 

CTv 

CO 

c- 

00 

cvj 

cvj 

C0 

CO 


• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 


rH -P 

M 0) 

CO 

VO 

00 

C- 


in 

CTv 

c- 

CTv 

CTv 

-=t 


td 

a> u 

CJ 

rH 

VO 

vo 

VO 

CVJ 

CO 

CTv 

in 

O 

cvj 

<p 

<p x: 

n d 

VO 

CJ 

-=t 

CO 

CO 


-=t 

in 

CVJ 

CO 

C~- 

o 

O -P 

rH 

O tH 
cd 

Ph 


* 

•=t 

rH 

* 

CO 

rH 

* 

CO 

•» 

00 


•» 

•=* 



•V 

o 

rH 

* 

o 

in 

CVJ 

m 

vo 

CVJ -P 

1 Ch 

cq td 

VO Pi 


d 

X> 

in 



VO 

o 

H 

•=* 


t- 

rH 

rH 

O 

-=t 

25 

in 

vo 

VO 

o 

CVJ 

rH 

r- 


Cl 

rH 


c- 

cvj 

CO 

VO 

t- 

m 

o 

•=T 

vo 

CO 

-P 

Pi 



•=t 

vo 

•=* 

vo 

-=t 

VO 

3 

-=* 

vo 

h- 

cn 

i 

rH 

CO 

cd 

0* 


«=*- 


*=t 

•=f 

■=t 


CO 

CO 

rH 


Pi P< C O 
a) ai 3 co 
^ > 

Pi -H O Pi 
cd 0) tH O 

a o •g -p 

0) 3 o 

0) K < £ 

& Pi Pi <H 
Hf( O « Cl 
to 0) N tH 
> H <P -P 

llrl ri rl £ 

•d 0) cd H bo 
rl O O Q. -H 
rl 0) O E rH 
OK Hi < K 


•d o 

cd 

C O 

6 

3 co 

S 

as 


M M i! 
■H O O 
h (d bJ 
K K K 


0 ) <! -H 

& h Q 
rH 0) 0) 

co n -d 

rl 3 
d rH -P 

•d 3 tH 

•H O -P 

3 ^ 





r 







! 



L 


i 

TABLE 3-20 

1 

FAILURE RATES FOR FLIGHT CONTROL AND MEASUREMENT LRU'S (FUNCTION F] 



Pre- Group II 

Post-Group II 

LRU 

Part Number 

Observed 
Failure Rate 

Predicted 
Failure Rate 

Observed 
Failure Rate 

Predicted 
Failure Rate 

Stable Element 

464289 

4,489.34 

4,464.28 



Rack. 

464474 

129. 9 1 ** 

- 



Roll and Pitch ECA 

464109 

2,992.89 

661 . 17 



Azimuth ECA 

464309 

4,489.34 

2,336.50 



Junction Box, Latitude 

Control 

464409 

2 , 618.78 

3,956.20 



System Power Control 

464905 

3,741.11 

440.99 

13,503.20 

738.59 

Demodulator Channel 

464209 

2,244.67 

979.98 



Integrator ECA 

464009 

3 , 367.00 

6,481.60 



Rack 

464374 

1 , 870.56 

7.28 



Static Pressure and Angle- 
of-Attack Data Compensator 

464721 

2,111.49*- 

6 , 666 . 00 



Rack 

464573 

441.70* 

- 



Air Data Computer 

464646 

4,863.45 

18 , 602.63 



Rack 

464773 

25.99* 

- 



Air Data Signal Data 

Converter 

464420 

2,244.67 

19 , 177.83 



Rack 

464673 

181 . 92 * 

- 



Altitude Rate SDC 

464320 

1,493.43*' 

4 , 617.70 



Rack 

464611 

64.97* 

- 



Bearing Select Converter 

Control 

464463 

1,039.07 *' 1 

560.01 



Roll and Pitch Rate Gyro 

464127 

3,741.11 

3 , 542 . 84 



Control Surface Command 

Amplifier Computer 

464121 

7,482.23 

2,375.30 



Rack 

464873 

1,496.45 

8.31 



Automatic Flight Control 

System (AFCS) Flight Mode 

Control 

464163 

1,122.33 

1,035.77 



Tactical Display HSI 

464181 

13,468.01 

19,229.66 



Rack 

464102 

2 , 618.78 

- 



Tactical Display Converter 

464920 

7,856.34 

5,716.02 



Rack 

464602 

227.40* 

10.00 



Cockpit Display No. 1 SDC 

464520 

2,992.89 

12 , 413.26 



Cockpit Display No. 2 SDC 

464620 

2,244.67 

7 , 903.22 



* Failure rates were taken from 

AFLC D056B-2 

of 18 November 1964 (6 months). 

** These failure rates were taken from AFLC D056B-2 of 18 
modified to 8o?S because of use in only So% of aircraft 

November 1964 (6 months) and 
that have a vertical configuration. 



.1 

l 

J 

J 

l| 

1 

) 

.1 

J 

1 

\ 

1 

J 

J 

J 

J 


u 


r 



TABLE 3-20 (Cutinuid) 

! Pre-Group II 


Post -Group II 


LRU 

Part 

Number 

Observed 
Failure Rate 

Predicted 
Failure Ra 

Aerodynamic Amplifier 

Computer 

464021 

1,122.33 

1,726.20 

AFCS Normal Aircraft 

Accelerometer 

464161 

748.22 

281.92 

Navigation and Landing 

Approach Amplifier Computer 

464221 

4,115.23 

4,373.85 

Communication and Naviga- 
tional Subsystem Test Set 

464396 

4,863.45 

4,144.55 

Steering Signal Converter 

Amplifier Computer 

464421 

3,367.00 

6,701.79 

Automatic Attack Amplifier 
Converter 

464621 

7,856.34 

7,438.21 

Angle-of-Attack Transmitter 

2562A-2 

1,122.33** 

508.96 

Temperature Probe 

3225-1A 

339.90* 

- 

Damper Amplifier 

1131-39101 

29,411.76* 

28,833.31 

Turn-Rate Transmitter 

11671G1 

4,255.32* 

24,778.96 

Pitch G Limiter 

8-61100-003 

374.11* 

9,018.17 

Linear Accelerometer 

24522K 

1,377.41* 

280.00 

Magnetic Azimuth Detector 

DT173/AJN 

363.77** 

- 

Compass Adaptor 

131316-01 

3,240.44** 

2,383.17 

Switching Rate Gyro 

MC-1 

255.89** 

490.00 

Power Supply Amplifier 

131313-01 

4,098.36** 

1,553.15 

Compass Controller 

131312-01 

1,432.66** 

907.49 

Displacement Gyro 

129370-01 

13,698.63** 

4,425.24 

Linear Accelerometer 

Transmitter 

TRU-3A 

767.46** 

153.38 

Vertical Speed and Altitude 
Amplifier 

15461-1-A1 

3,533.57** 

1,470.35 

Vertical Speed and Altitude 
Indicator 

18001-2A-5 

16,393.44** 

29,669.87 

Rate Gyro Transmitter 

131311-01 

2,915.45** 

1,153.73 

Altitude Director 

Indicator 

13131^-01 

10,309.28** 

1,803.97 

Mach Safe Speed Airspeed 

Amplifier 

154621 

2,506.27** 

1,261.46 

Mach Safe Speed Airspeed 

Indicator 

18000-1A-1 

8,333-33** 

23,308.23 

Aircraft Flight Director 

Computer 

CPU-4/A 

124.69** 

2,181.37 

Horizontal Situation 

Indicator 

522-2411 

748.22 

5,363.63 

HSI Amplifier 

522-1394 

250.13 

433.34 

Attitude Memory 

Amplifier Computer 

464321 

4,340.00 

3,920.00 



213,925.52 289,751.05 


BtSlIMMUBLLO 


290,048.65 



3.4 Determination of Reliability 

In chapter 4 of the original P-106 model report,* the following topics were 


discussed : 


• The reliability of each phase of each of the three standard missions, 
i.e., the probability that each phase can be successfully completed, 
given that the equipment operated correctly at the beginning of the 
mission. 

• The impact of each of the 13 functions on the reliability of Mission A, 
i.e., the percentage of mission reliability improvement that can be 
achieved by halving the failure rate of each function, in turn, while 
keeping the failure rates of the other functions constant. 

• The improvement in Mission A reliability that could be expected if the 
most unreliable LRUs in each function were improved by the ratio indi- 
cated for the predicted failure rates. 

These same general areas were considered during the preparation of this 
report. The comparison of the mission reliability profiles (Tables 3-21, 3-22, 
3-23) show very little change in reliability due to the group II modification. 
Therefore, very little could be gained by the reassessment of items two and three. 
For this reason, this discussion Is limited to the assessment of the differences 
in the Pre- and Post-Group II reliability profiles. 

The original model report pointed out that the probabilities for mission 
and phase success presented were apparent probabilities. The actual probabilities 
of mission success should be greater than the apparent probability because of the 
following factors : 

• The clustering effect at the function and system levels 

• The possibility of operating with less than full capability 

• The possibility that an LRU part might fail, but the function could 
be successful because that part was not critical 

3.4.1 Reliability Profile for Mission A 

Table 3-21 compares the Pre-Group II probabilities taken from the original 
model report and the Post-Group II probabilities. The differences shown here 
are not considered significant, since a number of variables other than the 
equipment modification could account for them. The differences result partly 
from differences in the operational requirements of the bases used as data 
sources for the two programs.** 


* Analysis of the Reliability Model for the MA-1 AWCIS on the F-106A Aircraft, 
Contract AF09(603)-4d024, ARINC Research Publication 329-01-1-491, 1 March 
1965. 

♦♦Dover and Selfridge AFBs are operational Air Defense bases, while Tyndall AFB 
is primarily committed to transition training. 





3 • 4 . 2 Reliability Profile for Mission B 

The probabilities of successful phase completion are compared in Table 3-22. 
Here, the model differs from Mission A In that the phase times are shorter and 
the total mission time is shorter. 

The differences shown here can be attributed to the differences In the data 
sources; they do not indicate a significant difference In the reliability of the 
two equipment configurations. 

3.4.3 Reliability Profile for Mission C 

The probabilities of successfully completing each phase and all the phases 
in Mission C are listed in Table 3-23. The low probability of mission success 
for both configurations can be attributed to the comparatively long mission time 
involved, and to the constraint that no maintenance can be accomplished during 
intermediate landings. It can also be attributed to the fact that no degree of 
reduced capability was considered in these calculations; i.e., either the system 
operated at full capability at each point in time, or else the system failed. 

Here, again, the small difference in the reliability of the Pre- and Post- 
Group II configurations does not indicate a significant difference in equipments. 

3.4.4 Reliability Impact of the Function 

Section 4.4 of the original model report presented discussions of potential 
function improvement and the impact of such Improvements on the mission. 

The material presented here is concerned largely with the contribution of 
unit unreliability to a function. The reasons for using this approach (rather 
than that of the original plan) are the following: 

• Although the general approach used in the original analysis could be 
used "as is", if it should be desirable to update Tables 4-4 and 4-5 
of the original model report, it appears that little could be gained 
by performing this exercise because of the forthcoming modification 
program . * 

• The Group II modification program caused no change at all in many 
functions (U, T, L, ADF, I, and D) and created only small changes 
in others (F, C, and A). Functions whose failure rates have 
undergone the greatest change because of the modification program 
have little effect on the mission-reliability values. 

3 . 4 . 4 . 1 Function F (Flight Control and Measurement) 

The only unit within function F to be modified during the Group II program 
is the System Power Control Unit (P/N 464905). Both the pre- and post-modifi- 
cation predicted failure rates (Table 3-20) indicate a low-complexity unit. 


* 


Changes to the TACAN, Computer, and UHF equipment. 




TABLE 3-21 




MISSION A RELIABILITY PROFILE 


Phase 

Number 

Phase 

Cumulative 

Time 

(Minutes) 

Pre-Group II 
Probability of 
Success 

Post-Group II 
Probability of 
Success 

1 

Climb 

11 

0.92581233 


2 

Vector 

26 

0.71110353 


3 

Offset 

26 

0.71110353 


4 

Acquire 

29 

0.66597171 


5 

Track 

31 

0.64117352 

0.62197155 

6 

Launch 

31 

0.64117352 

0.62197155 

7 

Pullout 

32 

0.64117352 

0 . 62197155 

8 

Vector 

46 

0 . 52302640 


9 

Offset 

46 

0.52302640 


10 

Acquire 

49 

0.49492620 


11 

Track 

51 

0.47917021 

0.45584903 

12 

Launch 

51 

0.47917021 

0.45584903 

13 

Pullout 

52 

0.47917021 

0.45584903 

14 

Return 

69 

0.43738146 


15 

Land 

75 

0.39139574 

0.37140336 


Phase 

Number 






Climb 

Vector 

Offset 

Acquire 

Track 

Launch 

Pullout 

Vector 

Offset 

Acquire 

Track 

Launch 

Pullout 

Return 

Land 


TABLE 3-22 

MISSION B RELIABILITY PROFILE 


Cumulative 

Time 

(Minutes ) 



Pre -Group II 
Probability of 
Success 

Post-Group II 
Probability of 
Success 

0.97236838 

0.77946580 

0.77946580 

0.734817&1 

0.71826223 

0.70185316 

0.71826223 

0.70185316 

0.71826223 

0.70185316 

0.59540550 

0.59540550 

0.56394685 

0.55264452 

0.53091975 

0.55264452 

0.53091975 

0.55264452 

0.53091975 

0.50464328 

0.45532100 

0.43631359 
























TABLE 3-23 


TABLE 3-23 

MISSION C RELIABILITY PROFILE 

Phase 

Number 

Phase 

Cumulative 

Time 

(Minutes ) 

Pre- Group II 
Probability of 
Success 

Post-Group II 
Probability of 
Success 

1 

Climb 

10 

0.93232622 


2 

Loiter 

54 

0.65084237 


3 

Return 

72 

0.59396107 


4 

Land 

79 

0.50081302 

0.4948175 

5 

Climb 

90 

0.46643225 


6 

Vector 

104 

0.25570040 

' 

7 

Ac quire 

107 

0.22229724 


8 

Track 

108 

0.20831394 

0.18785840 

9 

Launch 

108 

0.20831394 

0.18785840 

10 

Pullout 

109 

0.20831394 

0.18785840 

11 

Vector 

122 

0.17268244 


12 

Acquire 

125 

0.16338739 


13 

Track 

126 

0.16028069 

0.14210626 

14 

Launch 

126 

0.16028069 

0.14210626 

15 

Pullout 

127 

0.16028069 

0.14210 626 

16 

Return 

143 

0.14703964 


17 

Land 

150 

0.12437449 

O.IO987635 

18 

Climb 

161 

0.08670410 


19 

Vector 

175 

0.08396335 


20 

Ac quire 

178 

0.081957998 


21 

Track 

179 

0.074078776 

0.062473412 

22 

Launch 

179 

0.074078776 

0.062473412 

23 

Pullout 

180 

0.074078776 

0.062473412 

24 

Vector 

193 

0.061407804 


25 

Acquire 

196 

0.058102390 


26 

Track 

197 

0.056997614 

0.047258280 

27 

Launch 

197 

0.056997614 

0.047258280 

28 

Pullout 

198 

0.056997614 

0.047258280 

29 

Return 

214 

0.052287890 


30 

Land 

221 

0.042033101 

0.034726108 


33 















r 


However, a comparison of these two values shows an Increase in unit complexity 
of approximately 67 percent. When this 67 -percent increase in the predicted 
failure rate is compared with a 360 -percent increase in the observed failure 
rate, it is seen that this unit is a candidate for a detailed engineering 
investigation toward correction of this situation. 

It is necessary to consider the impact that Improvement of this unit might 
have on function F and function P (MA-1 Power) before the total gain can be 
estimated . 


3. 4. 4. 2 Function TR (Track Radar) 

The TR function is made up of the SR (Search Radar) units plus the addi- 
tional units needed to perform the track operation. For the purpose of this 
discussion, only the "additional" units will be included. Units common to both 
functions are discussed below. 

The units within function TR affected by the Group II Program are the Hand 
Control (P/N 464083), the Radar Relay Switch Assembly (P/N 464063), and the 
Steering Signal Computer (P/N 464346). 

The predicted failure rates (Table 3-15) for two of the three units 
(P/N's 464063 and 464346) show an increase following modification, and the third 
unit (P/N 464083) shows no change at all. The observed failure rates (Table 3-12) 
have decreased for all three units. Table 3-15 provides the comparison of the 
two equipment configurations and the base line for evaluation of future changes . 
The third area of interest, units showing high improvement potential, has been 
eliminated since the units are performing as well as, or better than, expected 
on the basis of the predictions. 

3. 4. 4. 3 Function SR (Search Radar) 

The changes to the equipment during the Group II program were largely con- 
centrated within the SR function units.* Ten of the 30 SR function units were 
modified to some extent. Three of the units (P/N's 464084, 464073, and 464173) 
showed no change in predicted values due to modifications. Of these units, only 
the P/N 464084 unit failed during the Tyndall AFB data -acquisition program. For 
the two units exhibiting no failures during this program, the Pre-Group II failure 
rates were used** in the current model exercise. The slight increase in the 
failure rate of the P/N 464084 unit does not represent a change that is considered 
significant to the function or to the system. However, it should be noted that 
the observed failure rate for this unit is still higher than should be expected 
on the basis of the predicted values. 

* These units are also used in the TR function. 

**The "old" values were used in the revised model for these units to prevent 
introducing any unwanted bias, as would be possible if "new" values were 
calculated on the basis of a new sample size and test time. 


1 

1 

:j 


i 


} 

1 

] 





1 



34 


One additional modified unit (P/N 464305) showed a small increase in failure 
rate following the modification. This was expected since the predicted failure 
rate Indicated an increase in unit complexity and failure rate. Here, again, 
the changes shown for this unit will have no measurable impact on either the 
function of the system. 

The P/N 464082 unit exhibits a very small (less than 1 percent) reduction 
in predicted failure rate end an insignificant (2i percent) increase in the 
observed value. These changes are well within the errors inherent in the method 
of quantification. 

P/N's 464065, 464855, and 464295 of this function (and the TR function) 
exhibit higher failure rates following the Group II modification program. Of 
these three units, P/N 464065 and P/N 464855 showed a reduction in predicted 
failure rates and demonstrated a corresponding improvement in the observed rate 
changes. The third unit of this group (P/N 464295) has demonstrated the poorest 
reliability of the modified units. Although there is no significant change in 
the predicted values, there is a striking difference in the two measured values.* 
Because of this difference and the fact that this unit contributes a large per- 
centage of the total unit failures of both the SR and TR functions, its main- 
tenance history should be investigated to determine the reason for its poor 
performance. Then action should be taken to correct the deficiencies that con- 
tribute to the reduction in reliability. 

The final group of SR-function units consists of the eight units added to 
the system during the Group II program. For the purpose of this discussion, 
these new units can be divided into three categories: 

(1) Those with very low predicted and observed rates of failure 

(2) Those with demonstrated observed failure rates as good as, or 
better than, expected on the basis of the predicted rates 

(3) Those which do not perform as Indicated by the predicted failure rates 

The first category consists of three units (P/N's 464432, 464484, 464516) 
with very low predicted failure rates and no observed failures during the data- 
collection portion of this program. One unit (P/N 464026) had an observed 
failure rate of approximately 31 percent of the predicted value. The combined 
failure rates of these four units contribute less than 1.5 percent to the total 
unit failures within the function and, therefore, present little or no oppor- 
tunity for improvement. The second category consists of one unit (P/N 464541) 
that has demonstrated an observed failure rate better than anticipated on the 
basis of the predicted failure rate, and two units (P/N's 464741 and 464841) for 
which the observed values closely approximate the predicted values. Thus these 
three units offer very limited improvement potential. 

*Pre -Group II Failure Rate: 7,108.12 failures per million hours. 

Post-Group II Failure Rate: 26,419.30 failures per million hours. 


33 



* 




t 








The remaining category consists of only the P/N 464641 unit. This unit 
alone accounts for more than 13. 5 percent of all SR-functlon unit failures and 
more than 9-5 percent of the total TR-functlon unit failures. While ranking 
highest in observed failure rate for both functions on the basis of the predicted 
values, it ranks fourth in the SR function and fifth in the TR function. Thus 
this unit becomes a prime candidate for engineering investigation directed at 
improving reliability. 

3. 4. 4. 4 Function IR (Infrared) 

The three units related to the TR function are also applicable to functions 
discussed earlier in this report. The Hand Control, P/N 464083, and the Steering 
Signal Computer, P/N 464364, have already been discussed, as has the Mode Oper- 
ations Set Control, P/N 464855* No further discussion of the units is necessary. 

3.4.4. 5 Function C (Computer) 

The one unit associated with the "C" function that was part of the Group II 
modification program was the equipment rack, P/N 464273. The modification to 
this unit produced no change to the predicted failure value. Since no electrical 
changes were made during the Group II modification program, the earlier observed 
failures were re-used in the revised model. 

3. 4. 4. 6 Function U (Command UHF) 

The Group II program did not include modification to the units of function 
U. Therefore, these units appear in this report exactly as in the original 
document . 

Function U units are to be replaced in the near future, thus precluding 
any requirement to analyze the Improvement potential of this equipment. 

3. 4. 4. 7 Function P (MA-1 Power) 

Two units within function P were modified during the Group II program: the 
Transistor Power Supply, P/N 464326, and the System Power Control, P/to 464905. 

The first of these (P/N 464326) showed no change in the predicted value as a 
result of the modification and showed an improvement in reliability as reflected 
in a reduction in the Post-Group II observed failure rate. The second unit, 

P/N 464905, was included in the discussion of the F function since it is part 
of that function. The only additional comment is that the total impact of any 
improvement to this unit must be based on its being used in more than one 
application. 

3. 4. 4. 8 Function T (TACAN) 

The equipment associated with function T was not modified during the Group 
II program. In addition, this subsystem is to be replaced in the near future. 

For the purpose of this model and assessment, the original failure-rate values 
were re-used in the current effort. 


3 

J 

i 


) 

i 

) 

] 


] 


36 



k 



% 





3-4. 4. 9 Function L (Instrument Landing System) 

The original failure-rate values for function L were re-used In the current 
effort since the units concerned were unaffected by the modification program. 

3. 4.4. 10 Function ADF (Automatic Direction Finder) 

The original failure-rate values for function ADF were re-used In the 
current effort since the units concerned were unaffected by the modification 
program. 

3.4.4.11 Function A (Armament) 

The original failure-rate values for function A were re-used In the current 
effort since the units concerned were unaffected by the modification program. 

3.4.4.12 Function I (Identification Friend or Foe) 

The original failure-rate values for function I were re-used In the current 
effort since the units concerned were unaffected by the modification program. 

3.4.4.13 Function D (Data Link) 

The original failure-rate values for function were re-used In the current 
effort since the units concerned were unaffected by the modification program. 

3.4.5 A Theoretical Mission Reliability 

No attempt Is made here to calculate a theoretical percentage of improvement 
for mission A as was presented in the earlier report. Although Important to the 
initial model exercise, a similar assessment of the total system based In part 
on potential Improvement to the TACAN, communications, and computer subsystems 
now in use would be of little value because of the present plan to replace these 
systems entirely. 

3.5 Conclusions 

In general, comparison of the Pre- and Post-Group II mission profiles 
shows a small decrease in the probability of mission success following Group II 
modifications. This decrease could be expected as a result of the increase in 
system complexity caused by the modifications. 

A closer look at the functions and units revealed that most of the observed 
unit failure rates changed in accordance with the changes in complexity and 
predicted failure rates. Some LRUs did not react as expected. These show the 
greatest potential for reliability Improvement. They are: 

(1) System Power Control, P/N 464905 

(2) IF Amplifier, P/N 464295 

(3) AFC H.V. Power Supply, P/N 464641 


37 


3.6 Recommendations 


The maintenance history of P/N 464905 and P/H 464295 should be Investigated 
In detail to determine the cause of the Increase In observed failure rate that 
followed this modification program and to determine the required corrective 
actions. 

The history of P/N 464641 should be Investigated to determine the causes 
of the substantial difference between the measured and predicted failure rates 
and to provide the necessary corrective actions. 


38 







4 . 


REVISION OF SYSTEM QUANTIFICATION 


4.1 Introduction 

The data presented In the original report were for all MA-1 units In the 
F-106A aircraft. Changes in configuration of the units that were made too late 
to be accounted for in Technical Order 1F-106-701, dated 3 February 1964, were 
not included in the program. 

For rapid generation of usable data on which initial planning and direction 
could be based, the MA-1 system was first defined in terms of the " — 06" code 
book, with only 7400 series codes being used. 

The system configuration was reviewed in greater detail upon receipt of 
basic system Technical Orders (particularly T.O. 1F-106A-2-27-1) . This review 
disclosed that a number of units supported by the instrument maintenance facili- 
ties were, in fact, part of the MA-1 system. The system redefinition resulted 
in the addition of several units to the data-collection program. Rather than 
delay publication, the report was submitted without the measured reliability 
values for these units. These values were provided when sufficient maintenance 
information had been accumulated.* 

4.2 Theoretical Reliability Characteristics 

Full advantage can be taken of theoretical MTEF figures only when the method 
of derivation is completely defined and understood. The values presented here, 
as were those in the original document, are based on a simple part count, assuming 
no redundant or parallel circuit design. The part failure rates used in both this 
and the original report represent field and operational failures and thus have a 
built-in average with respect to stress and operating conditions. These rates 
provide a good estimate of the performance that can reasonably be expected of 
any particular unit, assuming that typical engineering practices have been applied 
at the time of design and construction. The failure rates have been determined 
largely from studies on airborne weapon systems similar to the MA-1, with opera- 
tional environment and equipment age taken into account, and are Judged to be the 
most pertinent of available data. 

A detailed discussion of part failure rates was given in the original report. 
Since that discussion applies equally to both the original and current programs, 
it is included here as Appendix A. 


*See Table 1-1, ARINC Research Publication 329-01-1-492, 1 March 1965. 


39 



t 

: 5 . 




4.3 Observed Reliability Characteristics 

Here, again, the original program and the updating effort are necessarily 
based on similar "ground rules." This section is a discussion of these rules, 
their similarities, their differences, and the reason for their differences. 

4.3.1 Data Collection 

The measured reliability values presented in the original report were 
derived from failure and time data (supplemented AFM 66-1 data) collected by 
ARINC Research engineers stationed at Self ridge and Dover Air Force Bases. 

These measured values were based on two months of Selfrldge flight operations 
and four months of Dover flight operations, for a combined total of 2,673 F-106 
flying hours. 

The measured values for the Post-Group II equipment configuration presented 
in this updating document are also derived from failure and time data. However, 
in this case the measured values are based on four months of Tyndall Air Force 
Base flight operation, for a total of 1,703 F-106 flying hours. 

4.3.2 Data Organization 

The information presented in the tables of the original report were shown 
in descending order of complexity, i.e., MA-1 System, Subsystem, Equipment Groups, 
and Line Replaceable Units. In addition, the tables were organized to present 
measured reliability values for Dover and Selfridge Air Force Bases separately 
and combined . 


The tabular presentation in this report compares the measured reliability 
values for the equipment configurations for Pre- and Post-Group II. For simplic- 
ity, the Dover-Selfridge combined values are used as the Pre-Group II entries. 

4.3.3 Observed Reliability Data 

The original report presented the mean time between maintenance actions 
(MTBMA) and mean time between failures (MTBF) for the MA-1 AWCIS. 

The MTBMA values were derived by dividing the number of flight hours by 
the total maintenance actions initiated against the system, the subsystem, and 
the LRUs, respectively. 


The MTHF values were derived by dividing the number of flight hours by the 
number of maintenance actions necessary to satisfy complaints against the system, 
the subsystem, and the LRUs, respectively. MA-1 elements in which no malfunction 
was found were not included in the MTBF calculations. 


In all cases, the system or subsystem was considered to be maintained (MTBMA) 
or failed (MTBF) only one time during any one maintenance action even though more 
than one LRU of the system or subsystem was Involved. 





) 

J 



1 

) 




I 




40 






1 


I* 





r • 


These same rules were used In the collection and preparation of the 
revision data. However, the effort was limited to Include only those units 
(and associated subsystems) which were affected by the Group II modification 
program. Table 4-1 of this report presents a comparison' of Pre- and Post-Group II 
MTBMA's and Pre- and Post-Group II WTEF's for all units affected by the modifica- 
tion program. 

4.3.4 Theoretical and Measured MTEF 

Theoretical and measured MTBFs were compared In the original report. The 
original data were tabulated to show values for round- and vertical-instrumented 
aircraft and combined values. 

Similar material Is given in this report; however, the tabular presentation 
compares only those units which were affected by the Group II modification program. 

Table 4-2 presents a comparison of the following: 

(1) Pre-Group II predicted (or theoretical) MTEF values 

(2) Post-Group II predicted (or theoretical) MTEF values 

(3) Pre-Group II measured (or observed) MTEF values 

(4) Post-Group II measured (or observed) MTEF values 

The conditions and assumptions that applied to the acquisition and presenta- 
tion of the MTEF values were discussed In earlier sections of this report and in 
the original report and will not be repeated here. The same constraints were 
imposed in both programs. 

4.4 Observed Maintainability Data 

In the original report, maintainability data were presented in terms of 
total downtime and total repair time at the system and subsystem levels, with 
separate presentations for each category of data from each base. The method 
of data presentation has been modified for this report in that it is limited 
to the total MA-1 system downtime and repair time. 

4.4.1 MA-1 System Total Downtime 

System-total-downtime data (see Table 4-3) indicate the probability of 
completing maintenance within time t for the Post-Group II equipment configura- 
tion. These data encompass elapsed time from landing to correction of all sys- 
tem deficiencies. The number of actions completed at time t and the cumulative 
total of actions completed through time t are shown. Figure 4-1 is a graphic 
presentation of these data and the Pre-Group II data from Dover and Selfridge 
Air Force Bases. 


41 


J 


4.4.2 MA-1 System Repair Time 

System-repair time data (see Table 4-3) indicate the probability of 
completing maintenance within time t for the Post-Group II equipment configu- 
ration. These data encompass all time consumed in direct maintenance to correct 
all system discrepancies, the number of actions completed at time t c , and the 
cumulative total of actions completed through time t c . Figure 4-2 is a graphic 
presentation of these data and the Pre-Group II data from Dover and Selfridge 
Air Force Bases. 

4.5 Shop Repair Time for Line Replaceable Units 

The original report reflected average repair time in man-hours for MA-1 
system LRUs for which measured data were available. This same tabular presenta- 
tion has been modified for this report to show measured data from Dover and 
Selfridge Air Force Bases as the Pre-Group II values and data from Tyndall as 
the Post-Group II values. 

The data entries are presented here in Table 4-4 in the same way that 
Pre-Group II data were presented in the original report; i.e., LRUs with less 
than four maintenance actions were not considered representative. However, the 
LRUs from the Tyndall data are included in this report to show the observed LRU 
repair times. For those units with less than four failures, the number of 
repair actions is shown. 


42 



TABLE 4-1 

DATA COMPARISON FOR THE RADAR SUBSYSTEM. PRE AND POST-GROUP H IIP 
MODIFICATION: AT 95 PERCENT CONFIDENCE LEVEL 



Measured MTBMA 


Measured MTBF 


Pre- Group II 
UCL* 


Post-Group II 


Pre -Group II 


Post-Group II 


Fire Control 

Radar Subsystem Group I 


Rack, Indicator Pressure 

11045 

2673 

482 




11045 

2673 

482 




Synchro, Master Timer 

270 

167 

103 




650 

334 

170 




Waveguide Assembly 

11045 

2673 

482 




11045 

2673 

482 




Antenna, Radar 

99 

72 

52 




118 

83 

59 




Converter, Waveform 

39 

32 

23 




58 

45 

35 




Amplifier, Torque Generator 

251 

157 

98 




950 

445 

205 




Compressor, Air 

292 

178 

108 




349 

206 

120 




Relay, Switch Assembly 

431 

243 

136 

1048 

341 

146 

1645 

668 

261 

7038 

852 

237 

Receiver, Transmitter Radar 

68 

51 

39 

85 

56 

40 

68 

51 

39 

93 

61 

42 

Rack, LH Forward Compartment 

219 

141 

90 

414 

189 

100 

270 

167 

103 

774 

284 

131 

Indicator, Flight Command 

41 

33 

27 




49 

39 

31 




Gate Clutter 

387 

223 

128 

605 

242 

118 

950 

445 

205 

1563 

426 

166 

Control Manual 

118 

83 

59 

97 

63 

44 

165 

111 

75 

£80 

166 

107 

Coupler, Directional 

1645 

668 

260 

2747 

568 

195 

1645 

668 

260 

£747 

568 

195 

Amplifier, Video 

91 

67 

49 




157 

107 

72 




Meter, Radar Self Test 

11045 

2673 

482 




--- 

— 





Synchro, Range Track 

42 

34 

27 




66 

50 

39 




Amplifier, Azimuth Drive 

183 

121 

80 










adar Subsystem Group II 













Valve, Auto, Regulator 

349 

206 

120 




431 

243 

136 




Visor, Cathode Ray Tube 

2452 

891 

305 




4311 

1336 

371 




Amplifier, Antenna Track 

55 

96 

70 




233 

148 

94 




Recorder, Photographic 

118 

83 

59 




138 

95 

66 




Comparator, Signal 

2452 

891 

305 




11045 

2673 

482 




Sight, Fixed 

349 

206 

120 




349 

206 

120 




Rack, Atenna, Transmitter Group 

11045 

2673 

482 




--- 

... 

— 




IXict, Assembly, R/M Cool 

2452 

891 

305 




11045 

2673 

482 




Amplifier, "weep Generator 

132 

92 

64 




183 

121 

80 




Relay, Switch Assembly 

557 

297 

156 




1215 

535 

229 




Amplifier, Antenna Servo 

138 

95 

66 




318 

1191 

113 




Waveguide Assembly 

11045 

2673 

482 

67324 

1703 

307 

11045 

2673 

482 

67324 

1703 

307 

Converter, Signal Data 

270 

161 

103 




650 

334 

170 




Amplifier, Computer 

66 

50 

39 




144 

99 

68 




Amplifier, IF 

113 

81 

58 

49 

36 

27 

219 

141 

90 

50 

38 

28 

Control, Radar Set 

1215 

535 

229 

124288 

31^5 

567 

11045 

2673 

482 

124289 

3145 

567 

Amplifier, Filter Assembly 

165 

ill 

75 




431 

243 

136 




Computer, Steering Signal 

106 

76 

55 

152 

90 

58 

183 

121 

80 

414 

189 

100 

Generator, Sweep 

207 

167 

103 




557 

297 

156 




Filter, Band Pass 

2452 

897 

305 




11045 

2673 

482 




Amplifier, Video 

473 

267 

145 




1644 

668 

261 




Amplifier, Elevation Drive 

127 

89 

62 




206 

134 

86 





Radar Subsystem, Group III 
464026 Power Supply 

464432 4-Port Circulator 

464484 Directional Coupler 

464516 Waveguide Assembly 

464541 Computer Programmer 

464641 AFC Power Supply 

464741 LV Power Supply 

464841 Hydraulic Drive 

464855 Control, Radar Set 292 178 1 

464866 Amplifier, Transmitter Tuning 206 143 

464523 Converter, Signal Data 79 59 

Non-Radar System Units 

464273 Rack, RH Forward Compartment 431 243 1 

464326 Power Supply 3**y 206 1 

464905 Control, System Power 206 134 


* UCL = Upper confidence limit. LCL = Lower confidence limit. 
** Calculated LCLs. 




2747 

568 

195 




67324 

'1703 

307 



2747 

568 

307** 






407** 



2747 

568 

307** 






407** 



2747 

568 

307** 






407** 



605 

243 

118 




1048 

3^1 

146 



31 

24 

19 




46 

34 

26 



1562 

426 

166 




2747 

568 

195 



310 

155 

86 




605 

243 

118 

£92 178 

108 

456 

242 

141 

557 

297 

156 

910 

393 

200 

£06 143 

86 

223 

122 

73 

292 

178 

108 

275 

142 

81 

79 59 

44 




138 

95 

66 




431 £43 

136 

2747 

568 

195 

557 

297 

156 

7038 

852 

237 

349 206 

120 

1562 

42o 

166 

650 

334 

170 

119 

74 

50 

206 134 I 

86 

93 I 

61 

42 

473 I 

267 

145 

119 

74 

50 


BESl AVAILABLE COfl, 











TABLE 4-2 

DATA COMPARISON, RADAR SUBSYSTEM PRE-GROUP S 
PART-GROUP D DP MODIFICATION 



Unit 

Theoretical MTBF 

Part 

Number 


Pre-Group 
II IIP 

Post- Group 
II IIP 

464002 

PI re Control 

Radar Subsystem Group 1 

Rack, Indicator Pressure 

NE 


464003 

Synchro, Master Timer 

373.5 


464016 

Waveguide Assembly 

1980.0 


464017 

Antenna, Radar 

137.4 


464020 

Converter, Waveform 

242.1 


464041 

Amplifier, Torque Generator 

511.4 


464045 

Compressor, Air 

2628.1 


464063 

Relay, Switch Assembly 

277.6 

253.0 

464065 

Receiver, Transmitter Radar 

54.7 

81.0 

464073 

Rack, LH Forward, Compartment 

2339.1 

2339.1 

464080 

Indicator, Flight Command 

253.8 


464082 

Gate Clutter 

327.8 

327.8 

464083 

Control Manual 

531.0 

550.0 

464084 

Coupler, Directional 

7482.2 

7482.2 

464095 

Amplifier, Video 

315.1 


464096 

Meter, Radar Self Test 

10134.8 


464103 

Synchro, Range Track 

272.2 


464106 

Amplifier, Azimuth Drive 

517.8 


464107 

Valve, Automatic, Regulator 

9587.7 


464125 

Visor, Cathode Ray Tube 

NE 


464141 

Amplifier, Antenna Track 

316.4 


464149 

Recorder, Photographic 

141.9 


464150 

.Comparator, Signal 

707.4 


464169 

Sight, Fixed 

1293.4 


464173 

Rack, Atenna Transmitter Group 

NE 


464190 

IXict, Assembly R/M Cool 

NE 


464195 

Amplifier, Sweep Generator 

272.1 


464196 

Relay, Switch Assembly 

243.9 


464206 

Amplifier, Antenna Servo 

356.8 


464216 

Waveguide Assembly 

1983.3 

3929.0 

464223 

Converter, Signal Data 

326.0 


464241 

Amplifier, Computer 

170.0 


464295 

Amplifier, IF 

648.4 

646.6 

464305 

Control, Radar Set 

3062.4 

1083.8 

464341 

Amplifier, Filter Assembly 

318.0 


464346 

Computer, Steering Signal 

232.4 

231.4 

464389 

Generator, Sweep 

291.0 


464425 

Filter, Band Pass 

81234.8 


464495 

Amplifier, Video 

545.7 


464506 

Amplifier, Elevation Drive 

452.6 



BEsavaucit copy 



Measured MTBF at 95- Percent 
Confidence Limit 


2673 **82 
334 170 

2673 482 
83 59 

45 35 

445 205 
206 120 
668 261 
51 39 

167 103 
39 31 

445 205 
ill 75 
668 260 
107 72 

66 50 3 

431 243 13 

4311 1336 
233 148 

138 95 

11045 2673 
349 206 

1 

2673 482 

121 80 
535 229 
1191 113 
2673 482 

334 
99 
141 
2673 
243 





















Total System Downtime 


Total System Repair Time 


Number of 


Probability of 


Actions at 
End of 
Time t 


Completing 
Actions Within 
Time t 


Cumulative 

Actions 


Calendar Time, t 
(Hours) ' 


Cumulative 

Actions 


0.004862 

0.047002 

0.047002 

0 . 178282 
0 . 178282 


0 . 000000 
0.024311 
0.024311 
0.113452 

0 . 113452 


0.247974 

0.285251 

0.356564 

0.356564 


0.325770 

0.325770 

0.470016 

0.502431 


0.487844 


0.589951 
0 . 661264 
0.661264 
0.717990 
0.755267 


0.763371 

0.773096 

0.787682 

0.794165 

0.813614 


0.871961 
0 . 894652 
0.901135 
0.915721 


0 . 880065 


0.883306 

0.894652 

0.897893 

0.904376 


0.9335^9 

0.935170 

0.936791 

0 . 944895 


0.904376 

0.907618 

0.917342 

0.925446 

0.927066 


0.951379 
0.957861 
0.962723 
0 . 964344 
0.969206 


0.969206 


0.931929 

0.933549 

0.941653 

0.948136 

0.951378 


0.972447 

0.975689 

0.975689 

0.975689 


0.990276 


0.977310 


0.982172 

0.987034 

0.988655 

0.988655 


0.995138 

0.998379 

0.998379 

0.998379 


0.990276 


0.998379 

1.000000 


0.487844 

6 

0.567261 

44 

0.586710 

0 

0.643436 

35 

0.653160 

23 

0.695300 

5 

0.695300 

6 

0.737439 

9 

0.737439 

4 

0.758509 

12 


10.00 

-- 11.00 

10 

11.00 

— 12.00 

14 

12.00 

— 13.00 

20 

13.00 

— 14.00 

13 

14.00 

— 15.00 

9 




























































TABLE 4-4 


ACTIVE REPAIR TIME FOR 


LINE REPLACEABLE UNITS OF THE MA-1 SYSTEM 


Part 

Unit 

Average Man-Hours Per 

Repair Action Per LRU 

Number 


Pre- 

Group II 

Post -Group II 



Dover 

Selfrldge 

Tyndall 

332D6 

Fire Control 

Radar Subsystem Group I 

Gyro, Vertical 




463097 

Dehydrator 

— 

— 

— 

464002 

Rack, Indicator, Pressure 

— 


— 

464003 

Synchronizer, Master Timer 

0.6 

2.1 

— 

464016 

Waveguide Assembly 

— 

— 

— 

464017 

Antenna, Radar 

3.2 

4.2 

— 

464020 

Converter, Waveform 

0.8 

1.5 

— 

464024 

Indicator, Pressure 

— 

— 

— 

464025 

Filter, Light CRT 

— 

— 

— 

464041 

Amplifier, Torque Generator 

0/6 

1.8 

— 

464045 

Compressor, Air 

2.8 

1.2 

— 

464063 

Relay, Switch Assembly 

1.6 

0.8 

2.1 

464065 

Receiver, Transmitter Radar 

6.8 

2.7 

3-4(3) 

464073 

Rack, LH Forward Compartment 

7.0 

3-3 

4.0(2) 

464080 1 

Indicator, Flight Command 

1.9 

3-3 

— 

464082 

Gate, Clutter 

0.5 

2.4 

2.1 

464083 

Control, Manual 

1.1 

2.4 

2.8 

464084 

Coupler, Directional 

— 


3.0(1) 

464095 

Amplifier, Video 

0.6 

■ I 

— 

464096 

Meter, Radar Self Test 

wBM 


— 

464103 

Synchronizer, Range Track 



— 

464106 

Amplifier, Azimuth Drive 

ESI 


— 

464107 

464125 

Radar Subsystem Group II 

Valve, Automatic Regulator, 
Pressure 

Visor, Cathode Ray Tube 

2.6 

1.4 

— 

464141 

Amplifier, Antenna Tracking 

0.8 

1.3 

— 

464149 

Recorder, Photographic 

■ 

2.3 

— 

464150 

Comparator, Signal 


— 

— 

464169 

Sight, Fixed 


— 

— 

464173 

Rack, Antenna Transmitter Group 


— 

— 

464190 

464195 

Duct Assembly, Receiver Modu- 
lator Cooling 

Amplifier, Sweep Generator 

2.1 

1.3 



— 


47 






















TABLE 4 4 (CHtiuii) 


Part 

Number 

Unit 

Average Man-Hours Per 

Repair Action Per LRU 

Pre- 

Group II 

Post Group II 

Dover 

Selfrldge 

Tyndall 

464196 

Relay, Switch Assembly 

0.8 

1.2 

— 

464206 

Amplifier, Antenna Servo 

1.5 

1.1 

— 

464216 

Waveguide Assembly, Radar 

— 

— 

— 

464223 

Converter, Signal Data 

0.6 

1.0 

— 

464241 

Amplifier, Computer 

1.4 

1.5 

— 

464295 

Amplifier, IF 

1.3 

1.0 

2.2 

464305 

Control, Radar Set 

— 

1.8 

3-3(1) 

464341 

Amplifier, Filter Assembly 

1.0 

1.6 

— 

464346 

Computer, Steering Signal 

0.9 

2.1 

2.5 

464389 

Generator, Sweep 

1.1 

1.3 

— 

464395 

Amplifier, Attack Display 

0.5 

2.0 

— 

464425 

Filter, Bandpass 

— 

— 

— 

464495 

Amplifier, Video 

0.4 

— 

— 

464506 

Amplifier, Elevation Drive 

1.7 

1.3 

— 


Radar Subsystem Group III 




464674 

Rack, Dehydrator 

— 

— 

— 

464702 

Rack, Remote Scope Record 

— 

— 

— 

464774 

Rack, Dehydrator and Filter 

— 

— 

— 

464796 

Panel, Self-Test IAWCS 

— 

— 

— 

464855 

Control, Radar Set 

— 

1.2 

2.5 

464866 

Amplifier, Transmitter Tuning 

0.9 

1.2 

2.1 

464523 

Converter, Signal Data 

1.0 

1.2 

— 

464026 

Power Supply 

— 

— 

1-3(3) 

464432 

4-Port Circulator 

— 

— 

1.0(1) 

464484 

Directional Coupler 

— 

— 

— 

464516 

Waveguide Assembly 

— 

— 

— 

464541 

Computer Programmer 

— 

— 

5.0 

464641 

AFC Power Supply 

— 

— 

5.0 

464741 

LV Power Supply 

— 

— 

1.5^3) 

464841 

Hydraulic Drive 

— 

— 

2.2 


Non-Radar System Units 




464273 

Rack, RH Forward Compartment 

— 

— 

2.1(1) 

464326 

Power Supply 

— 

— 

3.0 

464905 

Control, System Power 

— 

— 

2.1 


Note: Number In parentheses indicates number of observed repair actions for 
units having less than 4 failures. 



5. INVESTIGATION OF POWER SUBSYSTEM 


5.1 Background 

During the earlier F-106 contract activity, ARINC Research published, 
measured, and predicted NTTEF values for units of the F-106 MA-1 system*. Analysis 
of these values showed that the system could be improved significantly by Improving 
the reliability of certain units in the power subsystem. Several recommendations 
for unit Improvement were made during the continuing program. However, because 
of the priorities given to the UHF and TACAN systems, the power subsystem received 
limited attention. More recently, greater attention has been given to the power 
subsystem since it now appears to be a major contributor to MA-1 unreliability. 
Modification programs to other portions of the system incorporating units or 
subsystems with more critical power requirements have contributed to this 
problem. 

Additional factors that affect system performance but are a part of the 
power-subsystem are changes in power-system loads resulting from modification 
programs and changes to wiring, shielding, and ground returns. The more recent 
investigation disclosed situations in which modifications caused the generation 
of transients, which were being introduced into the power busses. 

Analysis of flight -symptom data showed that most malfunctions reported by 
the operator directly against the power subsystem (ADCR 66-28 Codes: PP-) are 
normally verified by the maintenance technician and readily corrected. The more 
common reliability and maintainability problems associated with the power sub- 
system are well known, and a qualified flight -line technician has no trouble 
isolating actual failures in this subsystem. Hov;?ver, marginal performance, 
power-line noise, and other abnormal conditions occurring in the power-distribution 
busses in the different operational models of the F-106 weapons system cannot be 
readily Identified by the operator as power subsystem problems. This results in 
an operator complaint that cannot be verified by the maintenance technician. 

The operator usually identifies the problem as being related to the subsystem 
that displayed degraded performance. The radar, computer, and automatic -flight - 
control subsystems are the most critical to power requirements and are frequently 
reported as degraded when the problem is actually in the power being supplied to 
the subsystems. Problems of this type have been found to contribute to the high 
rate of unit adjustments (e.g., the steering and tracking functions), which in 
turn can upset the balance of the functional subsystems. They also reduce the 

* ARINC Research Publication 329-01-1-492, Quantified Reliability and Maintainability 
Characteristics of the F-106 AWCIS, 1 March l$b5. 


49 


percentage of next -flight success. With the incorporation of the more recent 
modifications -”ch as multi-mode storage display unit, P/N 464080-181, new 
adjustment proteins are being encountered. The new Multi-Mode Storage Tube 
(MMST) 464080 unit has been found to be extremely critical to the presence of 
noise in the ±50 Vdc supplied by the P/N 464326 unit. 

Earlier investigations were concerned with units 892/992, 491, and 326. 
However, the overall system effects of marginal unit performance were Identified. 
Design deficiencies, technical -order and procedural inadequacies, and test- 
equipment and AGE (Aerospace ground equipment) deficiencies were defined. These 
findings and those of earlier efforts indicated that a more detailed study of 
the power subsystem was necessary to define its characteristics and deficiencies. 

Since most components in the power subsystems were designed approximately 
ten years ago, the state of the art and the standards for aiivraft electrical 
power have changed substantially. Deficiencies that appear, by present standards, 
to be design inadequacies, such as deficient components (transistors, diodes, 
and resistors), are a problem common to most equipment designed in that time 
period. An additional problem is the location of heat -sensitive parts next to 
heat -producing comonents. 

The MA-1 power subsystem is representative of the state of the art for the 
time of its design. Thus many problems associated with this system are common 
to most systems of similar age. This study was directed toward better definition 
of the exact conditions present in the MA-1 power subsystem and the Impact of 
any anomalous conditions on F-106 reliability. 

5.2 General Description of the MA-1 Power Subsystem 

The F-106 MA-1 power subsystem develops the voltages necessary to operate 
the MA-1 AWCIS. In addition, the power subsystem performs switching, transfer, 
time-delay, and over/under -voltage -protection functions, as diagramed in Figure 
5-1. The basic power subsystem consists of 21 individual units. Table 5-1 lists 
the 21 units of power subsystem and the function of each. A more detailed 
distribution diagram, not including the timing, switching, or over/under -voltage 
circuitry, is presented in Figure 5-2 to show the high degree of Interdependency 
of the power-subsystem functions. 

The aircraft installation of the power-subsystem units is shown in Figure 5-3> 
with the location of the ARINC Research flight -recorder package. The recorder 
was installed in the space normally occupied by the P/N 464296 unit for the in- 
flight recording of the power subsystem's parameters. 

In addition to the 21 basic power -subsystem components, there are a number 
of specialized power supplies in the MA-1 system. Including units 464229, 464392, 
464489, 464741, and 464746. Although these units are not considered a part of 


50 




3AJJQ psads ^UB^SUOO 


c 



rH 

g o 



0 

0 P 



c 

Qh P 



c 

to o 



a 

c c 


rH 

x: 

Cd 3 


0 

o 

G <P 

r-H 

c 


Eh 

0 ' ' 

c 

c 

>> 

G rH 

cd^ 

o 

0 cd 

C ON 

JZ OJ 

p 

bDr-t 

cd in 

OV£> 

P 

Cd 0 

x: * 

w o 

ocx) 

P 'O 

o oj 

>> - 

0 rH 

rH i 

VO 

cd 00 

C rH 

O 0 

G rH 


G^- 

> 6 

0 ^ 

0 O 

O 

P 

<PCO 


O 

O P 

CO rH 


G 

P 

c o 

0 

0 

CO T3 

cd-^ 

6 

P 

cd C 

g 

P 

G 

m cd 

Eh 

£h 

H 



rH 

rH 

p 

T3 

0 

0 

0 

G 

G 

G 

G 

cd G 

C 

C 

G 

o 

cd 

cd 

cd 

G P 

.c 

x: 

x: 

O P 

P o 

o 

o 

o 

p G 

G^" 

G^ 

G 

cd 3 

O rH 

O rH 

O 

rH |x. 

p On 

P ON 

P r ' 

3 

P Lf\ 

P ON 

P P 

bO bO 



Cd ON 

0 G 

P OJ 

rH rH 

p t>- 

PC p 

3 On 

3 ON 

3 — ' 

G 

bOO 

bOCO 

bO 

0 0 

0'—' 



0 

bOP 

cc 

PC 

K 

Cd rH 

P P 

> 

> 

> 

O 

o 

o 

O 

> 

o 

LTN 

•=t 


CO 

p 

P 


+ 

+ 

l 



1 



G 

P 


0 

o 

G 

0 

Gr> 

O 

0 

m vo 

0 OJ 


G 

P OJ 

<P ON 

co 

C 

i m 

0 OJ 

G 

cd 



Gv> 

0 G 

x: 

V 


> O 

Or-* 

G P 

> P 

C P 

OJ 

cd 0 

0 

O P 

>> On 

G 

O C 

o o 

p P 

t~ c 

XT G 

G 

P. — ' 

cd 

p cd 

0 3 

(X 

o x: 

i x: 


3 

Ln o 

o 


CO 

i 

T3 

P 


>> 

G >> 


> 

•v P 

cd p 

O 


> PX 

PX 

> 

O 

PX 

o px 


in 

O 3 

O 3 


OJ 

in co 

P CO 


i 

+ 

+ 


C 


| 

P 




0 

P r\ 

P ° 

G 

G 


G 

G 

P 4-> 



cd 

S.S 

x: vu 

O ON 


€ 

0 P 

ON 


C ' 

COW 

G \ 


O OJ 

G G 

O OJ 


P ON 

T) D. O 

p On 


P ON 

P 1 p 

pcx) 


o\ 

0 0 P 

cd •> 


0 OJ 

p bfl o P OJ 


p On 

<d c 

3 ON 


OCX) 

P 3 

bOt-~ 


G * 

G P fc 

0 * 


(X OJ 

O O 

PC OJ 


1 VO 

P > 

ON 


0 O 

cd 

T) VO 



G TJ 

P ' 


cd 

0 G 

0 


p 

G cd 

P 


P 

0 



o 

C5 



> 


c 


G 

o 


O 

p 


P 

p 

G 


o 

O 

G C. 

G 

P 

O 0 

3 

cd 

P G 

fc 

G 

cd 0 


0 

G C3 

G 

G 

0 

O 

0 

G O 

P 

C5 

0 1 

cd 

G 

O 


0 

< 

o o 

G 

> 

1 1 

0 

O 

o 

p < 


•Power is supplied from either the ground power unit or the aircraft generators. 









> 

ft 


" — — 


bbi available COPY • 1 

— .1 




TABLE 5 1 



. a 



COMPONENTS OF POWEO SUOSTSTEM 



1 

ITEM 

NOMENCLATURE 

FUNCTION 

PART NUMBER 

LOCATION 

J 

1 

Interconnecting 

Box No. 1 

Delays application of 
certain voltages; dis- 
tributes voltages to 

MA-1 system buses. 

464018-153 

Armament 

Rack 

J 

2 

40 -Millihenry 

Reactor 

Filters the -150 v and 
-140 v generated 
voltages . 

464035-152 
(2 Req.) 

CSD-Gen. 

Compartment 


3 

Undervoltage, 
Overvoltage Relay 
Assembly 

Senses d-c voltages and 
disconnects a-c gener- 
ator fields in case of 
over- or under-voltage ; 
also disconnects a-c 
generator field in case 
of an over or under value 
in a-c voltage on receipt 
of a signal from 892 or 
992. 

465062-151 

Armament 

Rack 


4 

Alternating-Current/ 
Direct -Current 
Generator 

Generated +28 V, 115 V, 
400-Hz, 3-phase; and 

115 V, 

l600-Hz, 1-phase. 

Jack Sc Helntz 
31056-002 

(HugheB 464089-150) 

CSD-Gen. 

Compartment 


5 

+300 V Direct- 
Current Power 

Filter 

Filters high frequency 
ripple from +300 v 
supply. 

464092-150 

Armament 

Rack 


6 

Interconnecting 

Box No. 2 

Distributes 400-Hz and 
generated d-c voltages. 

464118-153 

Armament 

Rack 


7 

100-Millihenry 

Reactor 

Filters +300 V supply. 

464135-152 

CSD-Gen. 

Compartment 


8 

Power -Transfer 

Relay Assembly 

Distributes a-c voltage 
and +28 V supply. 

464162-152 

AFT-Fuselage 
(Sta 145) 


9 

-250 Vdc Power 

Supply 

Develops -250 from 

1600 Hz supply. 

464192-175 

Radar Rack 


10 

+100 V, -140 V 

Voltage Regulator, 

Ref. to +300 V 

Develops accurate 
+100 v and -140 v 
referenced to +300 v 
supply. 

464292-150 

Armame.it 

Rack 


11 

±50 V 8c - 15 V 
Transistor Power 

Supply 

Develops ±50 v and 
-15 v from l600-Hz 
supply. 

464326-150 

Radar Rack 









BtSI I AV AUABLt _COPX 

TABLE 5 1 leiltimi) 

COMPONENTS OF POWEI SUISYSTEM 


ITEM 

NOMENCLATURE 

FUNCTION 

PART NUMBER 

LOCATION 

12 

115/55 V, 1600 Hz 
+300 Ref. Voltage 
Regulator 

Develops accurately 
regulated 115 v and 55 v 
from 1600-Hz supply. 

464491-150 or 
464491-175 

C.N. & L. 
Rack 

13 

+300 Volts Direct 
Current Power 

Filter 

Filters the high fre- 
quency ripple from 
+300 v supply. 

464591-151 

Radar Rack 

14 

Direct Current 

Slip Ring 

Oenerator 

Generates +300 v, 

+150 v, and -140 v. 

Jack & Helntz 
31055-006 

(Hughes 464689-151) 

CSD-Qen. 

Compartment 

15 

+28 V & -140 V 

D. C. Field Volt- 
age Regulator 

Assembly 

Filters low frequency 
ripple from +28 v and 
-140 v supplies. 

464692-150 

Armament 

Rack 

16 

-140 Volts Direct 
Current Power 

Filter 

Filters high-frequency 
ripple from -140 v 

Bupply . 

464791-151 

Radar Rack 

17 

+300 V 8c +150 V 

D. C. Field Volt- 
age Regulator 

Assembly 

Filters low-frequency 
ripple from +300 v and 
+150 v supplies. 

464792-150-MD 1 

Armament 

Rack 

18 

+150 Volts Direct 
Current Power 

Filter 

Filters high-frequency 
ripple from +150 v 
supply. 

464891-150 

Armament 

Rack 

19 

400-Hz 

1600-Hz Field 

Voltage Regulator 
Assembly 

Regulates a-c voltages 
and provides over-volt- 
age/under-voltage pro- 
tection. 

Jack & Helntz 
51117-004 

(Hughes 464892-154) 

C.N. & L. 
Rack 

20 

+150 Volts 

Direct Current 

Power Filter 

Filters ripple frequency 
from +150 v supply. 

464991-150 

Radar Rack 

21 

400 -Hz & 

1600-Hz Field 

Voltage Regulator 
Assembly 

Regulates a-c voltages 
and provides over- 
voltage/under- voltage 
protection. 

464992-150 

C.N. & L. 
Rack 















FIGURE 5-2 

BASIC POWER SUBSYSTEM DISTRIBUTION -ROT .DING 
OV/UV FUNCTIONS AND TIMING/SWITCHING Clt> ulTRY 





















ARINC TAPE RECORDER 
INSTALLED HERE 


FIGURE 5-3 

AIRCRAFT LOCATION OF POWER-SUBSYSTEM UNITS 



the basic power subsystem, the characteristics of the power they supply were 
observed, and in some cases recordings were made. 

The six basic generated voltages originating from the Hughes Aircraft 
Company generators are listed in Table 5-2. The frequency control of MA-1 a-c 
voltages is dependent on the proper operation of the Constant Speed Drive unit 
and the Frequency Controller (CVAC P/N 689305). 

The Frequency Controller samples the output frequency of the 115-Vac, 400-Hz 
generator (Convair) to provide fine -frequency speed correction to the Constant 
Speed Drive unit. 

Failure of the Frequency Controller or the 115-Vac, 400-Hz generator will 
result in failure of MA-1 system power. Observed failures include frequency 
excursions as high as 423 Hz for the MA-1 400-Hz system (1692 Hz for the l600-Hz 
system) . 

When the aircraft is operated on ground power (AF/ECU-10M) , the six basic 
voltages are generated and regulated within the ground -power unit. When ground 
power is used, the normal functions of P/N 464892/464992, 464692, and 464792 are 
not utilized ; however, the over/under -voltage sensing and control circuits of 
these units are utilized. The ground -measurement tasks include the six basic 
voltages, but most of the effort was devoted to the secondary, or derived, 
voltages. The secondary voltages and parameters are listed in Table 5-3- 

Detailed operational theory was not included as part of the power subsystem 
investigation. This area is adequately covered in the MA-1 Field Maintenance 
Technical Manual for the power subsystem, T.O. 11F1 -MAI -12-1, 1 November 1961, 
Revised 16 November 1964. However, as shown in Figure 5-2, the system is unusually 
complex and highly interrelated; therefore, circuit descriptions are presented 
wherever appropriate in this discussion of the power subsystem. 

5.3 Investigation Methods 
5.3.1 Approach 

Units of the MA-1 power subsystem are operated in three different environments 
in the aircraft on aircraft power, in the aircraft on ground power, and in the 
maintenance shops. One objective of this investigation was to compare equipment 
operation and determine whether there was a correlation between events (such as 
voltage transients) that occur in the three environments. If correlation could be 
established, it would simplify identifying and isolating any problems that might 
be observed. For this task it was necessary to acquire a recording device capable 
of operating in all three environments, especially in the aircraft during flight. 
ARINC Research purchased a magnetic-tape recorder and associated electronic 
components that were packaged and used in several aircraft during flight lists. 




Voltage (Levels Within Tolerance) At Model Lf-440 Load Bank 
.n an Operational System (T-1C1 Digital Voltmeter). 
























TABLE 5-3 

GENERAL PARAMETERS OF SECONDARY VOLTAGES 


Secondary Voltage 

Unit Number 
(464 ) 

Derived From 

Referenced to 

Maximum 

A-C 

Component 

-250 Vdc 

+100 Vdc (Reference) 
-140 Vdc (Reference) 
+50 Vdc (Transistor) 
-50 Vdc (Transistor) 
-15 Vdc 

55/115 V, 1600 Cycle 
(Reference ) 

192 

292 

292 

326 

326 

326 

491 

115V, 1600 Hz 
+300 Vdc 

-250 Vdc 

115V, 1600 Hz 
115V, 1600 Hz 
-50 Vdc 

115V, 1600 Hz 

+300 Vdc 

+300 Vdc 

+100 Vdc (Ref) 

Internal Reference 

Internal Reference 

No Reference 
(Zener Diodes) 

+300 Vdc 

200 Mv P-P 

70 Mv P-P 

100 Mv P-P 

200 Mv P-P 

200 Mv P-P 

100 Mv P-P 

The voltages below are generated in units which are not 
part of the power subsystem. 

+50 Vdc (IR) 

746 

115V, 400 Hz, 
0C 

Internal Reference 

200 Mv P-P 

-50 Vdc (IR) 

746 

115V, 400 Hz, 
0C 

Internal Reference 

200 Mv P-P 

+50 Vdc (Computer) 

489 

+100 Vdc 
Reference 

+100 Vdc Reference 

Unknown 


A study of the power subsystem and its operation on the aircraft* shows that 
with the exception of the source of basic generated voltages, there is little 
difference between operation on aircraft power and operation on ground power. 

In both cases the loads, regulation, etc., are the same. This lack of difference 
made it possible to obtain much of the needed data from an aircraft operating 
on ground power. One aircraft was made available to be used with the Fault 
Detection Tester (FDT), which was being tested at Dover AFB. Aircraft serial 
number 500 was designated to be used for the FDT tests. Extensive work had been 
performed to optimize the power subsystem of this aircraft. During the FDT test, 
ARINC Research observed and recorded voltage waveforms for this optimized system. 

The investigation of the power subsystem consisted of measuring and recording 
voltages in the aircraft on both ground and aircraft power, making in-flight 
recordings as aircraft availability would permit, and recording specific voltages 
of the subsystem in the shop mock-ups. Direct monitoring was accomplished by 
connecting a Tektronix storage oscilloscope, type 564, to units of the power 
subsystem in the maintenance shops or in aircraft on the ground. Other observa- 
tions were made by recording voltage waveforms on magnetic tape and then playing 
back the tapes into the storage oscilloscope. This method was used primarily for 
in-flight recordings. However, the ability to store and play back the tape on 
the recorder was also used to a limited degree in the aircraft during ground tests 

♦In-flight recordings were made using several operational F-106 aircraft. 



|: 

li 

I 

3 

) 

i! 
















and during tests In the shop. It was thus possible to make permanent photographic 
records of some transitory waveforms that otherwise could not be photographed. 

The flight -test phase of the measurement program also included recording data 
on MA-1 performance and follow-up maintenance actions. This information was 
required for analysis of F-106/AWCIS power-subsystem characteristics in the 
operational environment. The format for collection of flight -recorder data is 
shown in Figure 5-4. 

The data collected and tabulated throughout the flight test phase of this 
program included the full range of system parameters and reflect the real-time 
relationships of events. 

5.3.2 Equipment Used 

5 . 3 .2.1 Magnetic Tape Recorder - A magnetic tape recorder was purchased 
from Kinelogic Corporation to be used for in-flight recordings of voltages in 
the power subsystem. The recorder included the following components: 

• Tape Transport 

• Record and Playback Head 

• Recorder Operation Logic 

• Seven Direct Record Amplifiers 

. Two Direct Playback Amplifiers 

• Bias Oscillator 

• Drive Motor Inverter 

• Inverter Voltage Regulator 

A special chassis was constructed at the ARINC Research laboratory in 
Annapolis, Maryland, to house the recorder components. All interconnecting 
wiring for the above components, additional control circuitry, and external 
wiring were installed in the ARINC Research laboratory. 

The recorder package was designed so that it could be installed in an F-106, 
for in-flight recordings, in the space normally occupied by the 464296 unit. 

This provided a convenient location with easy access to the power-subsystem test 
point located in the 464489 unit. No aircraft -wiring changes were required. 

The wiring required for normal operation of the MA-1 computer subsystem was 
Included in the recorder package. 

The recorder package was designed for both automatic and manual operation. 
Automatic operation was necessary for in-flight recordings, and manual operation 
was used for playback and during the recording of other than in-flight operation. 


59 




















Automatic operation was initiated when the nose -wheel -well door closed and 
continued until the end-of-tape signal was sensed by the recorder control 
circuitry. 

The electrical and mechanical specifications of the tape recording equipment 
used in this program are as follows: 

Mechanical 

Tape Speed: 7*5 ips primary 

15 ips via inverter frequency change (switched electronically) 

Number of Channels: Seven Record 

Seven Playback 

Tape Length: 600 feet 

Tape Type: 0.5-inch wide 

1 . 1 -mil thick 0 . 92 -mll mylar base, 0 .l 8 -mll oxide coating 
Record Time: 16 minutes at 7.5 ips 

Flutter: 2.0 percent maximum peak to peak; Pass band - 0.2 Hz to 

2.5 Hz, 7.5 ips speed 

Size: 7 inches by 6 inches by 3 7/16 inches (exclusive of ARINC 

Research modifications) 

Weight: 5.5 pounds (exclusive of ARINC Research modifications) 

Electrical 

Power: +27 Vdc +3 volts, approximately 0.7 amp 

Frequency Response: 100 Hz to 32 KHz +3 dB at 7*5 ips* 

Input Sensitivity: 0.1 to 1.5 volts rms 

Input Impedance 100K ohm minimum (unbalanced) 

Bias Oscillator Frequency: 750 KHz 
Output Impedance: 100 ohms 

Output Level: 1 volt rms, nominal, into 1 kilohm minimum 

5. 3 .2. 2 Additional Equipment - In addition to the flight -recorder 
package, the following items were required for ground measurements and recorder 
playback: 

• Tektronix Model 56 A, Storage Oscilloscope 

• Tektronix Model C-19, Oscilloscope Camera 

*This was the frequency response used by ARINC Research during this study, 
although the unit is capable of operation to 300 KHz depending on tape speed. 



61 


r 


% 


i 




• Hewlett-Packard Model 428B, Clamp-on D-c Ammeter 

• Tektronix Type P6016 A-c current probe 

• Voltmeters, Harmonic Analyzer, Frequency and Modulation meters available 
in the Air Force Inventory at Dover 

• Miscellaneous, patch cables, equipment carts 

5. 4 Findings Related to D-C Power 

Some of the problems In the power subsystem are associated with the secondary 
regulator-filter units and circuitry (unit P/Ns 464092, 464192, 464292, 464591, 
464791, 464891, and 464991), which are highly Interrelated and dependent on 
each other for proper performance. In Figure 5-2, it can be seen that these 
units provide reference voltages and operating bias voltages (which are derived 
from sources external to the units). These conditions create feedback loops, 
making the identification of specific problems, such as transients, extremely 
difficult. 

The 115-volt, l600-Hz source also provides inputs to this group of units. 

These include tube heater voltage and power to compensate for undervoltage 
excursions in the related output voltage through the series clampax circuits 
in the 464092, 464591, and 464791 units. The series clampax circuits are designed 
to compensate for transient low-voltage conditions of the -140, +150 and +300 Vdc. 
These d-c voltages originate in the field regulators. The purpose of the parallel 
or shunt clampax circuits is to maintain the incoming voltage at a specified 
maximum level. 

The clampax configuration gives rise to a problem whenever there is a 
momentary loss (less than 50 seconds) of the d-c source voltages from the field 
regulator units. The load imposed oh the series clampax and transformer- 
rectifier supply (fed by the 115-volt, l600-Hz source) as a result of this momentary 
loss causes an immediate catastrophic failure of components in the 115-volt, 
l600-Hz transformer and rectifier-filter circuits. A secondary result of this 
momentary failure is a short in the power transformer, causing a power dump of 
the 115-volt, l600-Hz source and consequently loss of the entire power subsystem. 

The loss of transformer Tl, known to be a high failure item, is typical of this 
type of failure. 

Several failed transformers were recovered as failed parts during earlier 
ARINC Research reliability studies. Examination showed that the transformer 
case had exploded, with a resultant loss of potting tar. 

Overload protection is not provided to the transformer and rectifier-filter 
circuits. If protection were provided for these circuits, the MA-1 system could 
in many cases complete an operational mission with some possible degradation of 
performance. 





! 



I 

1 





I 



62 


An additional problem associated with the clampax units is the abnormally 
high replacement rate of the type-609A regulator tube. The failure mode of most 
of these tubes was complete loss of emission. The failure of one or two of these 
tubes does not result in complete failure of the unit functions, because of the 
parallel configuration; however, ARINC Research has found that the noise levels 
recorded on units with failed tubes were much higher than normal. 

In one case during the ARINC Research measurement program, a 591 unit was 
discovered in an operational aircraft in which three of the four series clampax 
regulator tubes had failed. Tube-tester checks showed no emission at all for any 
of the three tubes. The only Indication of the failure of this 591 unit was the 
high level of noise and transients recorded on the +300 Vdc power buss in the air- 
craft. Separate clampax circuits, two for each voltage, are used to filter the 
+300-Vdc and +150-Vdc power in the aircraft. One clampax circuit is located for- 
ward and one aft in the aircraft. When voltage noise is checked in the aircraft 
at the front of one of these clampax units, the noise may be within tolerance. 
However, because of long lead lengths between units, the noise level at another 
MA-1 system unit may exceed the specified tolerance. To alleviate this problem, 
a more complete noise check at several units may be required. 

Another problem related to the clampax circuits is illustrated in Figure 5-5. 
During certain loading variations caused by system-mode switching, high-level 
oscillations occur on the +150 Vdc line. 

A cursory investigation of the effects of adding additional filter capacitance 
across the outputs of units 326, 192, 791, 991, 891, 092, and 591 was not fruitful. 
The additional capacitance in most cases did reduce the noise level; however, the 
added capacitance created phase shift in the feedback loops, which made the regu- 
lator-clampax circuitry even more susceptible to oscillation. Further investigation 
into this oscillation problem is required. 

5.4.1 Voltage Problems 

The power-subsystem voltages, both generated and conditioned, were observed 
and photographed in both aircraft and shop environments. Tables 5-4 and 5-5 list 
the voltages recorded and the location in which these photographic recordings were 
made. Where available, the specification for maximum allowable ripple is also 
listed for the applicable voltages. Out-of -tolerance conditions shown in tables 
5-4 and 5-5 are discussed later in this section of the report. 

All measurements listed in Table 5-4 were made in aircraft serial number 500, 
which had been specially prepared for tests of the Fault Detection Tester (FDT) 
at Dover AFB. The voltages observed in this aircraft represent the optimum for 
an F-106 aircraft because extensive work was performed to prepare the power sub- 
system of this aircraft for the FDT tests. 

Measurements shown in Table 5-5 were made in the various maintenance shops 
(radar, computer, electric, etc.) as identified in the table. Specifications for 
maximum permissible ripple on each voltage are provided for comparison. 

63 






















TABLE 5-5 


I* 

■ 

p. 


¥ . 



RIPPLE VOLTAGE MEASUREMENTS MADE IN SHOP MOCK-UP 
(O-C VOLTAGES) 


Voltages 

Maximum 

Allowed 

Ripple 

Radar Shop 
Mock-Up 

Electric Shop 
Test Stand 

±50V Power Supply 
464326 

No Load 

100$ Load 

Good Unit 

Defective 

Unit 

+28 Vdc 

4 volts 

4 volts* 





-140 Vdc 

200 

800 





+150 Vdc 

200 

240 

260 

500 



+150V Field 



380 

1.3 volts 



+300 Vdc 

300 


900 

1.3 volts 



+300 Field 



360 

1.4 volts 



-250 Vdc 

200 

460 





+100V (Ref) 

7u 

280 





-140V (Ref) 

100 

560 





+50V (Transistor) 

200 

560 



20 

150 

. -50V (Transistor) 

200 

960 



20 

160 

-15 V 

100 

590 





+50V (Ref) 

Unknown 






+50V (IR) 

200 






-50V (IR) 

+75V 

-75V 

200 







Unless otherwise specified, all voltages are in millivolts peak to peak. 

*In addition to the 4 volts ripple, the d-c level shifts approximately 
-1-1/2 volts. 


The voltage conditions noted during this program are discussed briefly 
below. 

5-4. 1.1 +28 Vdc [Maximum Allowable A-C Peak-to-Peak Ripple 4 volts as 
specified in T.O. 11F1-MA-1-12-1] . The +28 Vdc line was monitored in aircraft 
number 500 with a Tektronix storage oscilloscope, type 564. At times transients 
were observed, but attempts to photograph these transients* using only the 
oscilloscope were unsuccessful. Therefore, the magnetic tape recorder was used 
to monitor these voltages, and photographs were made during subsequent playback 
of the tapes. Figure 5-6 illustrates one condition recorded during a three- 
minute tape run when radar on/off mode switching was taking place. Figure 5-7 
is another recording made during a tape run in an aircraft. In this case, a 
transient of +80 volts was recorded; but because of saturation of the recorder 
amplifiers, the total amplitude of this transient is not shown. This transient 


*The quality of the photographs of the worst-case voltage conditions was such as 
■to preclude reproduction in this report. 



65 

















Sweep speed: 1 sec/cm 


+150 Vdc, 100^ load 

Electric Shop Mock-up Test Stand 

Amplitude: 2 V/cm 


FIGURE 5-5 

COMPARISON OF THE +150 Vdc: 
NO LOAD AND FULL LOAD 



FIGURE 5-6 


TRANSIENTS ON THE +28 Vdc LINE 




is being coupled into the 115-Vac 1600-Hz reference signal. The +28 Vdc and 
115-Vdc 1600-Hz voltages, recorded on magnetic tape during flight, were observed 
on a dual -trace oscilloscope; they indicate that when the transient on the +28 
Vdc occurs, a transient is also injected into the 115-Vdc 1600-Hz reference 
voltage (see Figure 5-8). 

Magnetic-tape recordings of the voltage in the radar shop mock-up also show 
the presence of ripple (4 volts peak to peak); however, although not shown, a 
shift in the d-c reference of 1-1/2 volts was observed when the gyro heat 
circuit in the radar antenna was cycled on and off. 

5*4. 1.2 -i4o Vdc [Maximum Allowable Ripple 200 Millivolts Peak to Peak 
(mV P-P)]. Measurements in the aircraft showed that the allowable maximum ripple 
was exceeded in the 062 unit (Figure 5-9), the 195 unit (Figure 5-10), the 289 
unit (Figure 5-11) and the 389 unit (Figure 5-12). 

First recordings of the -140 Vdc in the radar shop mock-up (with an 
oscilloscope sweep speed of 0.2 milliseconds/cm) appeared to be within specifica- 
tions, as shown in the upper trace of Figure 5-13- However, re-examination at 
a lower sweep speed (5 milliseconds/cm) revealed that the a-c component (see 
Figure 5-14, lower trace) was 800 millivolts peak to peak. This was the result 
of one noise level of 180 mV riding on a 60-Hz signal that is introduced on the 
-140 Vdc in the radar shop mock-up. 

5. 4. 1.3 +150 Vdc [Maximum Allowable Ripple 200 Millivolts Peak to Peak 
(mV P-P)]. Two of the recordings made in the aircraft indicate the P-P limits 
are exceeded. One recording was made in the 289 unit (see Figure 5-11 lower 
trace), and the other in the 389 unit (see Figure 5-15, upper trace). In both 
cases the levels recorded were 300 mV or more. 

The +150 Vdc voltage in the radar shop was found to have a ripple level of 
240 mV P-P (see Figure 5-16, lower trace). When this same voltage was observed 
on the electric shop test stand under both load and no-load conditions, the 
measured noise Increased from 260 mV P-P under no-load (Figure 5-17, upper trace) 
to 500 mV P-P under 100$ load (Figure 5-18, upper trace). 

Figure 5-5 shows the +150 Vdc as observed in the electric-shop test stand. 
The slow oscillation shown was later corrected by replacing a gassy tube in the 
991 unit. This condition was not observed when the equipment was operating 
under 50-percent or 100-percent loads. Figure 5-19 shows the slow recovery time 
of the +150 Vdc caused by the defective tube. 

5. 4. 1.4 +300 Vdc f Maximum Allowable Ripple 300 Millivolts Peak to Peak 
(mV P-P)] . The +300 Vdc line was monitored in several units in the aircraft. 
However, the allowable limits were exceeded at only one unit, P/N 464389. In 
this unit the level was 360 mV P-P, as shown in Figure 5-20, lower trace. 


Magnetic Tape Recording In Aircraft 



FI6URE 5-7 

TRANSIENTS ON THE +28-Vdc LINE 


In-flight Recording 



FIGURE 5-8 

SWITCHING TRANSIENTS IN THE 28-Vdc DISTRIBUTION SYSTEM AND THE 115-V, 1600 Hz REFERENCE, 
MODULATED BY TRANSIENTS GENERATED IN THE 28-Vdc SYSTEM 



Direct Oscilloscope Recording in Aircraft S.N. 500 


-l40 Vdc 



P/N 464791 Unit 
I Amplitude : 50 mV/cm 
"i - Denotes Specific Limits 


Sweep speed: 0.5 ms /cm 


FIGURE 5-10 

VOLTAGE WAVEFORMS OF THE 140-VMc DISTRIBUTION 




\n 






Sweep speed: 2 ms/cm 


+150 Vdc 
P/N 464289 

Amplitude: 200 mV/cm 


FIGURE 5 

VOLTAGE WAVEFORMS IN THE 289 UNIT 


Direct Oscilloscope Recording in Aircraft S.N. 500 


-140 Vdc 

P/N 464389 Unit 

Amplitude: 100 mV/cm 


l 


1 l 

1 ^ 




-140 Vdc 

P/N 464591 Unit 

Amplitude : 100 mV/cm 


Sweep speed: 0.5 ms/cm 


FIGURE 5-12 

VOLTAGE WAVEFORMS OF THE -140-Vdc DISTRIBUTION 



-140-Vdc Reference 





) " 

•>A vj.vv^ i^V WMx y^y,' '■* 


Sweep speed: 0.2 ms/cm 


Radar Shop Mock-Up 
Amplitude: 200 mV /cm 


FIGURE 5-13 

VOLTAGE WAVEFORMS FOR THE -140 Vdc AND THE -140-VdC REFERENCE 


Direct Oscilloscope Recordings In Maintenance Shop 

-140-Vdc Reference 
Radar Shop Mock-up 
Amplitude: 200 mV /cm 


-140 Vdc 

Radar Shop Mock-up 
Amplitude : 200 mV/cm 


FIGURE 5-14 

VOLTAGE WAVEFORMS FOR THE -140 Vdc AND THE -140-Vdc REFERENCE 



Sweep speed: 5 ms/cm 


71 




+150 Vdc 

P/N 464103 Unit 

Amplitude: 100 mV/cm 


{ 

1 

I 







Direct Oscilloscope Recording in Maintenance Shop 



Sweep speed: 0.2 ms/cm 


+150 Vdc, No Load 

Electric Shop Mock-up Stand 
Test Point 

Amplitude: 200 mV/cm 


+150-V Field, No Load 
Pin 8 of AR 203 
Electric Shop Mock-up 
Amplitude: 100 mV/cm 


FIGURE 5-1 7 

VOLTAGE WAVEFORMS OF THE +150-Vdc OUTPUT AND THE +150-V FIELD VOLTAGE 





Direct Oscilloscope Recording in Maintenance Shop 



+150 Vdc 100# Load 

Electric Shop Mock-up Stand 
Test Point 

Amplitude: 200 mv/om 


+150-V Field, 100# Load 
Pin 8 of AR 203 
Amplitude: 100 mV/cm 


Sweep speed: 0.2 ms/cm 


FIGURE 5-18 

VOLTAGE WAVEFORMS OF THE +150 Vdc OUTPl'T AND THE +150 V FIELO VOLTAGE 


73 



Direct Oscilloscope Recording in Maintenance Shop 



No 

Load 


50 $ 100 $ 

Load Load 
On On 
[-•Note slow 
time recovery! 
to +150 V 
level 

Load switch 
activated & 
held until 
next change 



+150 Vdc Load Switching 
Electric Shop Mock-up Test Stand 
Amplitude: 50 v/cm 

Sweep speed: 0.5 seconds/cm 


FIGURE 5-19 


SWITCHING TRANSIENTS AND REGULATOR RECOVERY FOR THE +150-Vdc DISTRIBUTION 


Direct Oscilloscope Recording in Aircraft S.N. 500 



+300 Vdc 
PA 464103 Unit 
Amplitude : 100 mV/cm 



p speed: 0.5 ms /cm 


+300 Vdc 

P/N 464389 Unit 

Amplitude: 200 mV/cm 


FIGURE 5-20 

• itt (ICE WAVE TRIMS Of THE Oil Vdc IN THE 113 ANO 311 UNITS 


— — 




The +300 Vdc line was also monitored at the test points in the electric- 
shop test stand. The ripple present on the +300 Vdc supply under no load was 
900 mV P-P (Figure 5-21, upper trace). When the load was increased to 100-percent, 
the ripple increased to 1.3 volts (Figure 5-22, upper trace). Although in both 
cases the specification was exceeded, since standard test- procedures do not require 
the measurement of ripple voltage, this condition would go undetected and shop main- 
tenance personnel would consider this to be a good unit. 

5.4. 1.5 -250 Vdc [Maximum Allowable Ripple 200 Millivolts Peak 

to Peak (mV P-P)]. The maximum allowable ripple was found to be exceeded in 

only one unit (P/N 464389) of the several units tested in the aircraft. The 
ripple was 700 mV P-P as shown in Figure 5-23, lower trace. The -250 Vdc voltage 
at this point resembles a sawtooth waveform. 

When the -250 Vdc was observed in the radar shop, the ripple was found to 
be 460 mV P-P (see Figure 5-16, upper trace), but the waveform did not resemble 
that seen in the aircraft. In the shop, one noise component of about 300 mV 
P-P appears to be riding on a 1600-Hz signal. 

5-4. 1.6 +100 Vdc Reference [Maximum Allowable Ripple 70 Millivolts 

Peak to Peak (mV P-P)] . The maximum ripple level is exceeded in two of the 

locations monitored on the aircraft. These are shown in Figure 5-24, lower 
trace (292 unit) and Figure 5-25, lower trace (096 unit). 

The noise level of the +100 reference voltage was recorded in the radar 
shop at 280 mV P-P as shown in Figure 5-26, lower trace. (The presence of 
1600 Hz is also indicated. ) 

5. 4. 1.7 -l40 Vdc Reference \ Maximum Allowable Ripple Specification 

100 Millivolts Peak to Peak (mV P-P)] . The -140 Vdc reference voltage was 
monitored in the aircraft and the ripple specification was found to be exceeded 
in several cases. (The source from which the -140V reference is derived (the 
292 unit) had been repaired and tested in the shop mock-up just prior to the 
measurement program.) Figure 5-27, upper trace, shows the -140V reference as 
it appears at the test point on the 292 unit. The ripple (180 mV P-P) exceeds 
the specifications for the unit under normal operating conditions.* 

During the aircraft measurements, it was noted that operation of the equipment 
in various modes produced some unusual waveforms on the -140V reference. When 
the operator changed mode to Spotlighting IR Target in Hand Control, the -140V 
reference experienced deviations as much as 1.9 volts, as shown in Figure 5-28, 
upper trace. When the equipment was switched to Radar Hand Control, the noise 

♦Also note the lower trace (-250 Vdc) in Figure 5-27, which is the source for 
the -140V Reference. No evidence of noise peaks are seen here, indicating 
that the noise on the -140V Reference does not originate at the source (-250 Vdc). 


75 





+300 -V Field, No Load 


Electric Shop 


Pin 11 of AR 103 


Amplitude: 100 mV/cm 


Sweep speed: 0.2 ms/cm 


FIGURE 5-21 


RIPPLE VOLTAGE LEVELS OF THE +300-V4C SYSTEM 


Direct Oscilloscope Recording in Maintenance Shop 


+300 Vdc, 100# Load 


Electric Shop Stand Test Point 


Amplitude: 500 mV/cm 


+300 V Field , 100# Load 


Electric Shop Stand 


Pin 11 of AR 103 


Amplitude: 100 mV /cm 


Sweep speed: 0.2 ms/cm 


FIGURE 5-22 


RIPPLE VOLTAGE LEVELS OF THE +300-¥dc SYSTEM 



Direct Oscilloscope Recording In Aircraft S.N. 500 




Sweep speed: 0.5 ms/cm 


-250 Vdc 

P/N 464103 Unit 

Amplitude: 100 mV/cm 


-250 Vdc 

P/N 464389 Unit 

Amplitude: 500 mV/cm 


FIGURE 5-23 

VOLTAGE WAVEFORMS FOR THE -250-Vdc SYSTEM AT THE 103 AND 389 UNITS 


Direct Oscilloscope Recording In Aircraft S.N. 500 

+300 Vdc 
P/N 464292 Unit 
Amplitude: 100 mV/cm 


+100-Vdc Reference 
P/N 464292 Unit 
Amplitude: 100 mV/cm 

Sweep speed: 0.5 ms/cm 

FIGURE 5-24 

NOISE PRESENT IN THE +300-Vdc AND +100Vdc REFERENCE 





77 






Direct Oscilloscope Recording In Aircraft S.N. 500 


-140 Vdc Reference 
P/N 464096 Unit 
Amplitude: 500 mV /cm 



+100 Vdc Reference 
P/N 464096 Unit 
Amplitude: 100 mV /cm 


Sweep speed: 1 ms /cm 


FIGURE 5-25 

NOISE AND RIPPLE PRESENT IN THE -140-Vdc REFERENCE 


Direct Oscilloscope Recording in Maintenance Shop 


-15 Vdc 

Radar Shop Mock-Up 
Amplitude: 200 mV /cm 




78 




Direct Oscilloscope Recording in Aircraft S.N. 500 





-140 Vdc reference 
P/tt 464292 Unit 
Amplitude: 50 mV/cm 


-250 Vdc 
P/to 464292 Unit 
Amplitude: 100 mV /cm 

Sweep speed: 0.5 ms /cm 

FIGURE 5-27 

NOISE PRESENT IN THE -14-Vdc REFERENCE AND THE 
-250-Vde PRESENT AT THE 292 UNIT 



Direct Oscilloscope Recording in Aircraft S.N. 500 

-l40 Vdc Reference 
P/to 464292 Unit 
Amplitude: 500 mV/cm 


-250 Vdc 

P/N 464292 Unit 

!_ Amplitude: 100 mV/cm 

I Denotes Specific Limits 



Sweep speed: 0.5 ms /cm 


FIGURE 5-28 

VOLTAGE WAVEFORMS FOR THE -140-Vdc REFERENCE AND THE 
250 Vdc AT THE 292 UNIT 

(OPERATING MODE: SPOTLIGHT IR TARGET IN HAND CONTROL) 



79 




In the -140V reference was found to be 480 mV P-P (Figure 5-29> upper trace). 
Switching the equipment to the Visual Identification Mode produced a noise level 
of 720 mV P-P on the -140V reference, as shown in Figure 5-30. 

The upper trace of Figure 5-lb shows the -140V reference as observed in the 
radar shop. This waveform does not resemble the waveforms observed in the 
aircraft. Note the presence of 60 Hz on the -140V reference in the radar shop. 

5. 4. 1.8 +50 Vdc (Transistor ) [Maximum Allowable Ripple 200 
Millivolts Peak to Peak (mV P-P)] . The lower trace of Figure 5-31 shows the 

+50 Vdc in the aircraft. The noise appears like blocks of pulses on the +50 Vdc. 
Also observed but not photographed were spikes with an amplitude of 300 mV P-P. 

The origin of these pulse blocks could not be determined, but they are of concern 
because the +50 Vdc is used in many critical circuits including new circuitry 
being added to the F-106. 

Several good 326 units were observed in the electric -shop mock-up, and 
Figure 5-32, lower trace, shows the +50 Vdc in a typical unit. Noise shown in 
the photograph is about 20 mV P-P, well within specifications. One 326 unit, 
removed from an aircraft for control -stick chatter in the Assist mode, had 
approximately 40 mV P-P noise under no load (see Figure 5-33> lower trace). 

When the unit was placed under load, the +50 Vdc noise level Increased to 150 
mV P-P, as shown in the lower trace of Figure 5-34 (still within the specification). 
However, after repair, the noise level was reduced to approximately 20 mV P-P, 
as shown in Figure 5-32. When installed in the aircraft, the unit operated 
properly. This example Indicates that the noise specification for this unit 
should be reduced to assure proper operation. 

5. 4. 1.9 -50 Vdc (Transistor ) [Maximum Allowable Ripple 200 
Millivolts Peak to Peak (mV P-P)] . The upper'trace of Figure 5-31 shows the 
-50 Vdc voltage as recorded on the aircraft. The noise appears like blocks 

of pulses on the -50 Vdc. The origin of these pulse blocks could not be determined 
but they are of concern because, as with the +50 Vdc, the -50 Vdc is used in many 
critical circuits, including new circuitry being added to the F-106. 

Several good P/N 464326 units were observed in the electric shop mock-up; 
the noise level of a typical good unit is shown in Figure 5-32. The noise level 
of approximately 20 mV P-P is well within specifications. The same unit, 
described in the discussion of +50 Vdc, was found to have a noise level of 
approximately 40 mV P-P under no load (see Figure 5-33 > upper trace), and 160 
mV P-P under full load (see Figure 5-34), both within the specification. Unit 
repair also reduced the noise level of the -50 Vdc to approximately 20 mV P-P, 
as shown in Figure 5-32. When installed in the aircraft, the unit performed 
properly. Again, the noise specifications are not realistic. 


80 


Direct Oscilloscope Recording in Aircraft S.N. 5°0 





Sweep speed: 5 ms/cm 


-140 Vdc Reference 
P/N 464292 Unit 
Amplitude: 200 mV/cm 

+100 Vdc Reference 
P/N 464292 Unit 
Amplitude: 100 mV/cm 


FIGURE 5-29 

VOLTAGE WAVEFORMS FOR THE -140-Vdc REFERENCE AND 
THE -250-Vdc AT THE 292 UNIT 
(OPERATING MODE: RADAR HAND CONTROL) 


Direct Oscilloscope Recording in Aircraft S.N. 500 



-140 Vdc Reference 
P/N 464096 Unit 
Amplitude: 200 mV/cm 


Denotes Specific Limits 


Sweep speed: 5 ms/cm 

FIGURE 5-30 

WAVEFORM FOR THE -140-Vdc REFERENCE AT THE 096 UNIT 
(OPERATING MODE: VISUAL IDENTIFICATION) 


8 ] 




Direct Oscilloscope Recording in Aircraft S.N. 50° 


-50 Vdc (Transistor) 
P/N 464096 Unit 
Amplitude: 100 mV/cm 





Direct Oscilloscope Recording in Maintenance Shop 


NOISE PRESENT IN THE -50 Vdc AND +5 Vdc LINES 



-50 Vdc, No Load 

P/N 464326 Unit (Defective Unit) 

Amplitude : 100 mV/cm 


+50 Vdc, No Load 

P/N 464326 Unit (Defective Unit) 

Amplitude: 100 mv/cm 


Direct Oscilloscope Recording in Maintenance Shop 


-50 Vdc 100$ Load 

P/N 464326 Unit (Defective Unit) 

Amplitude: 100 mv/cm 


+50 Vdc No Load 

P/N 464326 Unit (Defective Unit) 

Amplitude: 100 mv/cm 


Sweep speed: 0.5 ms/cm 


FIGURE 5-34 


NOISE PRESENT ON THE -50 Vdc AND THE +50-Vdc LINES 


83 




r 



5.4.1.10 -15 Vdc [Maximum Allowable Ripple 100 Millivolts Peak 

to Peak (mV P-P)]. No problems were noted with the -15 Vdc voltage when It was 
observed in the aircraft (see Figure 5-35» lower trace). 

The upper trace of Figure 5-26, taken In the radar-shop mock-up, shows the 
noise level to be 590 mV P-P, well over the 100 mV P-P specification. 

5.4.2 Unit Problems 

5-4.2. 1 - 250 -Vdc Power Supply (P/N 464192 ) - One problem with 

the 192 unit is the need to select tubes (i.e., trying several tubes) during main- 
tenance to obtain the proper output voltage. This is a unit deficiency since pro- 
perly designed circuits are not so critical that replacement tubes have to be 
selected. The tube selection as required in this application is expensive, involves 
many man-hours, and requires extensive use of the AGE equipment. 

No adjustments are incorporated in this unit. This is complicated by the 
fact that tubes and other components may change slightly with age. Most well 
designed power supplies contain limited adjustments that can compensate for 
slight aging of components or tubes. 

Another problem of the 192 unit concerns the NE-68s (neon bulbs) used in 
the saturable -reactor and voltage -amplifier circuitry. These components change 
with time and are partly responsible for the tube -selection requirement. Also, 
oscillations in this unit have been found to be caused by a change of value in 
the NE-68 s. 

The operation of the NE-68 regulators have been found to vary over a wide 
range when the unit is adjusted in normal light (as in the shop) and then retested 
in total darkness (as in the aircraft). 

Four samples were tested with the following results: 

• Ionization voltage increased by 7 percent to 20 percent when the light 
source was removed. 

• There was little change in the extinguish voltage or the operate voltage 
due to removal of the light source. 

It is probable that units aligned in the shop's light would completely 
fail to regulate in the aircraft in the absence of light since the voltage 
required for initial ionization would never be reached. 

5. 4. 2. 2 +100V/-140V Reference Supply (P/N 464292 ) - No adjustments 
are provided in the 292 unit to compensate for component aging or difference in 
tube characteristics. This necessitates careful selection of replacement tubes. 

Characteristics of the regulators are such that under specific load conditions 
or MA-1 system modes, high -amplitude oscillations occur. A square-type waveform 
of 2 -milliseconds duration with amplitudes of approximately 3 volts were recorded 
and photographed. 


J 

j 

J 

] 

J 



84 





The +100 and -140V reference voltages provided by this unit are utilized 
for precision bias and excitation voltages in the radar, FC&M, and computer 
subsystems and are extremely Important to weapon- systems accuracy. 

The vibrators (choppers) G1-G2 and G3-G4 must be matched (same manufacturer 
and part number); otherwise problems result. The polarity of chopper connections 
is reversed by one manufacturer (pins i and 4 or 1 and 6 are reversed). This 
in itself does not create a problem so long as two choppers of the same type are 
used for G1 and G2 or G3 and G4. However, the use of one manufacturer's chopper 
in G1 or G3 and a second manufacturer's chopper in G2 or G4 creates a problem 
because of the polarity of the pins. If the maintenance technician is not aware 
of this noninterchangeability, a unit might leave the shop with intermixed 
choppers. This would cause an on-aircraft computer malfunction . 

5.4.2. 3 ±50-Vdc and -15 -Vdc Power Supply (P/N 464326) - Power supply 

P/N 464326 contains two separate but identical channels. One channel develops 
and regulates the +50-Vdc power, and the other develops and regulates the -50 
Vdc power. A divider network, consisting of a resistor and two series zener 
diodes connected between the -50 Vdc line and ground, develops -15 Vdc. Both +50 
Vdc and -50 Vdc supplies contain a full -wave rectifier, a filter circuit, and two 
regulators. The input transformer is common to both supplies and operates on 
the 115-Vac, l600-Hz, single-phase basic power. The regulators are controlled 
by magnetic amplifiers located between the input transformer and the rectifier. 

One malfunction frequently attributed to the 326 unit is "stick chatter 
in the pilot-assist mode." During this program, voltage -output waveforms of 
+50 and -50 Vdc were observed in several 326 units in a bench mock-up at Dover 
AFB. Several units (operating satisfactorily) were observed under different 
load conditions and the waveforms were photographed (Figure 5-32) on one unit 
(Serial Number 348). No change was observed in the waveforms under the different 
load conditions. Next, a unit (Serial Number 37) that was removed from an 
aircraft "for stick chatter in the pilot-assist mode" was observed and photographed 
(Figure 5-33) under the same load conditions as the good unit. 

In a good 326 unit (such as Serial Number 348) approximately 20 millivolts 
of noise (small spikes that occur at each half cycle of the 115 Vac, l600-Hz 
voltage) are present on both +50 and -50 Vdc voltage outputs under no-load 
conditions (Figure 5-32). In the unit (Serial Number 37) removed from the 
aircraft, 40 millivolts of noise (at a higher frequency and not related to 115V 
1600 Hz as it would be in a good unit) were present (Figure 5-33) on each output 
under no-load conditions -- approximately twice the normal noise exhibited by 
a good unit. 

Next a 100-percent load was applied to the +50 Vdc output (no load on 
-50-Vdc output). There was no appreciable change in noise (Figure 5-32) on the 
good unit. However, on the unit removed from the aircraft the -50-Vdc output 


86 





(Figure 5-36, upper trace) increased to 80 millivolts, and the noise on the +50- 
Vdc output (Figure 5-36 lower trace) increased to 70 millivolts. 

The 100-percent load was then applied to the -50-Vdc output (no load on 
the +50 Vdc), and waveforms were observed. No change occurred in the outputs 
of the good unit. The noise on the outputs of the unit removed from the aircraft 
changed to 140 millivolts (Figure 5-34, upper trace) on the -50-Vdc output and 
150 millivolts on the +50 Vdc output (Figure 5-34, lower trace). All of the 
measurements are within the tolerance (200 millivolts peak to peak) specified 
in the checkout and troubleshooting procedures in T.O. 11F1 -MAI -12-1, Section 
XVII, page 17-2. 

The unit removed from the aircraft was repaired ■, a detailed record of repair 
actions was made. The malfunction was found in the -50 Vdc section of the 
unit. The large bursts seen in Figure 5-34 were attributed to a defective 
voltage comparator that consists of two encapsulated transistors. The spikes 
seen in Figure 5-34 were removed by replacing the voltage -amplifier transistor, 
Q10. 

After the repairs were made, the +50 and -50-Vdc waveforms were observed 
to be the same as those observed for a good unit (Figure 5-32). These unit 
repairs corrected the stick-chatter malfunction. Testing this unit to the T.O. 
limits of 200 millivolts of noise on the +50 and -50 Vdc outputs would not have 
uncovered the malfunction. 

Another item that may be affected by problems within the 326 unit is the 
new Multi-Mode Storage Tube (MMST) modification, which uses ±50 Vdc from this 
unit. The MMST appears to be sensitive to noise on the ±50 -Vdc line. Also, 
several display tubes have been removed because of burn spots, and it is possible 
that these failures are related to the unit. 

Other problems observed in the 326 unit are as follows: 

• A common ground connection is not used; instead, brass ground studs are 
connected to the chassis. Corrosion has been observed at these connec- 
tions, and in some cases the leakage of tantalum capacitors 06 and C25 
appears to accelerate this corrosion. 

• Resistors Rl4 and R29 are underrated, and they overheat. 

• Oscillations occur under various load conditions when there is voltage 
imbalance in the voltage comparator Ql-A/B or Q6-A/B. This results in 
control -stick chatter in the pilot -assist mode of operation. 

• The adjustments of variable resistors R8 and R23 must be set at their 
limit for proper output voltage . (ARINC Research recommended corrective 
action for this problem in the 20th Monthly Status Letter dated 
February 1966, Task RI-O5-3), 


87 












« 



One of the major problems In the power subsystem Is caused by the loop arrangement, 
whereby secondary power supplies and filter units are Interrelated and dependent 
on other units for proper performance (see Figure 5-2). These conditions create 
feedback loops, which make It extremely difficult to pinpoint problems. 




The 115-Vac, 1600-Hz source also provides Inputs to this group of units, and j 
supplies tube heater voltages and the power to compensate for under -voltage j 
excursions In the related output voltages through the series clampax circuits. j 



A problem common to all these units is tube -type 6094 (clampax). These 
tubes frequently fail and go undetected In the aircraft when there is no T.O. 
requirement to check them during scheduled maintenance such as the 100 -hour 
check. Failed tubes are not detected in the aircraft, because most tubes operate 
in parallel circuits; if one fails (from low omission, for example), the unit will 
still operate but in a degraded manner. One of these types of failure was 
observed during checkout of units removed from an aircraft. In this case a slow 
oscillation (see Figure 5-5) was observed on the +150 Vdc line, accompanied by 
slow recovery of the +150 Vdc under load switching (see Figure 5-19). This 
problem was corrected by replacing two type-6094 tubes in the P/N 464991 unit. 

One tube was removed for excessive leakage, and the other was gassy. 

Provisions for testing tubes are made in shop procedures but not in any 
of the scheduled -maintenance procedures on the aircraft. Present flight -line 
noise checks do not disclose this type of malfunction. Testing of these tubes 
should be added to the 100-hour check to assist in detecting cases such as the 
one described above. 


J 



One other problem associated with power supplied ‘dncT f lifer units was 
observed recently at Dover AFB. The type-6094 tubes had been replaced in 
one unit, but careful observations of the output revealed that the unit was 
still unstable. Further investigation revealed that the 100-ohm series equalizing 
resistor of one of the type-6094 tubes had changed value to 80 ohms. Replacement 
of the resistor corrected the malfunction. 

5-4.2. 5 Clock Pulse Generator, P/N 464489 - The clock pulse generator 
is not a part of the power subsystem, but this unit contains several voltage - 
test points such as the +100 Vdc reference. During tests of the Fault Detection 
Tester at Dover AFB, noise checks were made at TP-3 on this unit (the test point 
for +100 Vdc reference). However, this test was inconclusive, because a filter 
is located in unit 489 between the voltage source and TP-3. To obtain proper 
noise checks at this test point, the unit should be modified to provide a direct 
connection between the test point and the voltage source. 

5. 4. 2. 6 Units 464092, 464591, and 464791 - The series clampax circuits 
in units 092, 591> and 791/ are designed to compensate for short-term low-voltage 
conditions that may occur in the -140, +150/ and +300 Vdc circuits. The series 


: 

:1 

1 

1 

j 

j , 

:: 

| 

1 


88 


V 


circuits are supplied by the 115-Vac, l600-Hz source. Parallel or shunt clampax 
circuits act to prevent the incoming voltages supplied by field regulator 

circuitry from falling below a specified minimum level. In this configuration, 

. 

* a momentary loss (less than 50 seconds; of the d-c source voltage from the field 

regulator units will cause the series clampax circuits to attempt to correct 
for this loss. 

The series clampax circuit and transformer -rectifier supply (fed by 115-Vac, 
l600-Hz source) were designed to correct for low -voltage conditions but become 
overloaded when required to withstand the full MA-1 system load. The normal 
result in this situation is an immediate catastrophic failure of components in 
l ■ • the 115-Vac, l600-Hz transformer and rectifier-filter circuits. A secondary 

result of this momentary condition is a short in the power transformer, which 
causes a power dump of the 115-Vac, l600-Hz source and consequent loss of the 
entire power subsystem. The loss of the transformer, T-l, in unit 591 is typical 
of this type of failure, and transformer T-l is known to be a high-failure item. 

The addition of a properly designed overload-protection circuit would prevent 
aborts due to these short-term under-voltage conditions and prevent the catastro- 
phic component failures that are now being experienced. 

5. 5 Findings Related to A-C Voltages 

s ■ . The basic a-c voltages generated for the MA-1 system are 115 volts, 3 phase, 

400 Hz; and 115 volts, 1 phase, 1600 Hz. Frequency control of these voltages is 
dependent on the Convair frequency-control system. Voltage regulation of the 
400 -Hz and l600-Hz sources is accomplished in the 892 unit (or in the interchange- 
able 992 unit) by controlling the field voltages of the 31056-002 MA-1 generator. 

The regulator circuits have a closed-loop configuration, with the sensing voltage 
being picked off the a-c power bus. Field power is derived from the output 
voltage except during initial turn-on, when the field is excited by 28 Vdc. 

The regulator units also contain circuitry to sense over- or undervoltage conditions 
and remove power from the MA-1 systems when these extremes occur. 

When the system is being operated on the aircraft with ground power or in 
the system shop mock-up, the regulator sections of units 892 to 992 are not 
utilized. To test the condition of the complete field regulator unit in the 
aircraft installation, it is necessary to operate the system on aircraft -engine 
power. The electric-shop mock-up has provisions for connecting the test stand 
to an ECU-10 ground-power unit, to a short-life ground power unit (P/N 586-3OO), 
or to the motor-driven constant-speed-drive unit and aircraft-generator test stand. 
The short-life ground-power unit is configured with aircraft generators (P/N 
464089) to permit shop maintenance of the field regulators. The constant -speed- 
drive unit is configured with aircraft generators of the same type currently in 
use on the aircraft. The major differences between the two generators are in the 



! 




i 






28-Vdc section; however, there are also slight differences in field voltage and 
current requirements in the a-c section. These differences, noted during tests 
with a P/N 464892 regulator, were demonstrated by the fact that the voltage - 
adjustment controls required resetting to maintain the desired 116 volts during 
switching from the short-life ground-power unit to the constant -speed-drive unit. 

The compatibility between shop test and aircraft Installation conditions 
related to adjustment of the regulators could be enhanced by using the same type 
of generator in both locations. However, this will not completely solve the 
problem, because of differences in the voltage losses in the aircraft wiring. 

The Technical Order procedures call for a shop setting of 116 volts ±2 volts 
and a maximum deviation of 4 volts from a no-load to a full-load condition. 
Measurement of a-c voltages on an aircraft (using the 892 unit) following adjust- 
ment in the shop to Technical Order specifications results in a voltage reading 
approximately 2 volts lower than the shop setting. (The voltage difference is 
of less concern when the 992 unit is used, because thlB unit has wider control 
range. ) 

Investigation into the cause of the recorded voltage differences between 
the shop and aircraft installations indicated that additional voltage drop (IR 
loss) in the aircraft wiring and differences in regulator-sensing pick -off points 
were contributing factors. The a-c regulator is located some distance from the 
generator. In the case of the l600-Hz supply — which is a single-phase, two 
wire "floating" system -- a two-wire field supply is also required; this doubles 
the wiring losses. The wiring used from the regulator to the generator for the 
1600-Hz field power is number 16 wire. The cable is approximately 35 feet long, 
for a total wire length of 70 feet. The full-load field current for this system 
is rated at 7.6 amperes. The IR loss for number 16 wire at 7.6 amperes was 
computed to be 30 millivolts per foot — a total voltage drop of 2.1 volts rated 
generator load. The MA-1 system load is approximately 60 percent less than full 
rated load; therefore, the expected voltage drop in the aircraft would be 1.35 
volts. This represents a major portion of the recorded 2-volt difference between 
shop and aircraft measurements. 

The sensing-voltage pick-off point for the a-c regulators from the aircraft 
power bus is at the 162 unit. The current required by the sensing circuit is 
relatively small; the wiring is thus susceptible to noise and transient pick-up. 
The sensing-wire routing is not direct from the pick-off point to the regulator; 
it Is routed to the P/N 174 rack through the 062 unit, back out through the 
P/N 174 rack and then to the P/N 074 rack, and eventually to the regulator unit. 

An apparent design deficiency also exists. The pick-off points for the 792 
and 692 units are made at about the midpoint of the sensing lines (P/N 174 rack). 
This practice of sharing the sensing lines with other circuitry, especially 
circuits of variable loading, is undesirable. The wire size between Terminal 
Board 16 and the 1600-Hz a-c bus is number 18, which is believed to be too small. 



90 



The result of this arrangement Is amplitude modulation of the 115-volt, l600-Hz 
MA-1 power during load changes on the 792 and 692 d-c regulators. Amplitude 
variations of the 1600-Hz supply were noted during this program. However, to 
measure and record properly the total effect of the sensing-lead loading and 
noise pick-up, It would he necessary to modify the aircraft wiring. This was 
not attempted during the program. 

The 115-volt, 400-Hz system is a 3-phase, 4-wire "Y" connected system; the 
neutral point is connected to aircraft ground near the generator. The terminal 
strip for the MA-1 generator and the ground is located above the generator on 
the airframe. This area is subject to vibration. A number of cases of fluctua- 
ting 400 -Hz MA-1 voltage were verified as loose ground connections at terminal 
strips TSN-6 and TSN-8. Access to the terminal -strip area with the engine 
installed is difficult. Therefore, it is desirable to check all connections on 
this terminal strip whenever the engine is removed. 

The 115-volt, 1600-Hz reference voltage is developed in the 491 unit. It 
is derived from the l600-Hz basic source and referenced to the +300 Vdc supply. 

5.5.1 Voltage Problems 

The a-c voltages of the power subsystem were observed and photographed in 
both the aircraft and shop environments. The conditions shown in the photographs 
taken on the aircraft represent an optimized F-106 power subsystem. Extensive 
work was performed on the subsystem in this aircraft in preparation for the 
field evaluation tests of the Fault Detection Tester. 

The general findings concerning the various a-c voltages of the power 
subsystem are discussed in the following paragraphs. A later section covers 
the units associated with the a-c voltages. 

5. 5. 1.1 26 Vac, 400 Hz - The 26-Vac, 400-Hz, phase-B waveform as 
observed on the aircraft is shown in Figure 5-37- The 26-Vac return was 
photographed on the aircraft at the 289 unit and is shown in Figure 5-38. 

Note the presence of spikes with a peak-to-peak amplitude of 2 volts. 

5. 5. 1.2 115 Volt, 400 Hz - Figure 5-39 shows the Phase-A waveform 
as it appears on a magnetic- tape recording taken in the aircraft. Figure 5-40 
shows the waveform in the shop mock-up, taken from the magnetic -tape recording. 

The lower trace of Figure 5-4l shows Phase A as it appears in the radar shop 
mock-up with regulation provided by an AF/ECU-10M power unit. 

The Phase-B waveform is shown in Figure 5-42 as it appears on a magnetic- 
tape recording on the aircraft. Figure 5-43 is from a magnetic tape run in the 
shop mock-up. The upper trace in Figure 5-41 shows Phase B as it appears in the 
radar shop mock-up with regulation provided by an AF/ECU-10M power unit. 

Phase C is shown in Figure 5-44 as the waveform appears on a magnetic tape 
run on the aircraft. Figure 5-45 is from a magnetic tape run in the shop mock-up. 
The upper trace of Figure 5-46 shows Phase C as it appears in the radar shop 


91 


Direct Oscilloscope Recording in Aircraft S.N. 500 



Sweep speed: 0.5 ms/cm 


26 Vac, 400 Hz, Phase B 
P/N 464289 
Amplitude : 20V/cm 


Note: Upper trace only, 


FIGURE 5-37 

26-Vdc WAVEFORM SUPPLIED TO THE 289 UNIT 



FIGURE 5-38 


26 Vac RETURN LINE 





Direct Oscilloscope Recording in Maintenance Shop 



115 Vac, 400 Hz, Phase B 
Radar Shop Mock-up Test Stand 
Amplitude : 100 V/ cm 

115 Vac, 400 Hz, Phase A 
Radar Shop Mock-up Test Stand 
Amplitude: 100 V/cm 


Sweep speed: 0.5 ms/cm 


FIGURE 5-41 

115 Vac AT RADAR TEST STAND 


Magnetic Tape Recording in Aircraft S.N. 500 



115 Vac, 400 Hz, Phase B 
P/N 464489 Unit 
Test Point TP-14 


FIGURE 5-42 

WAVEFORM FOR 400 Hz. PHASE B ON AIRCRAFT 





Magnetic Tape Recording in Maintenance Shop 



115 Vac, 400 Hz 
Phase B 

Shop Test Stand 




FIGURE 5-43 

WAVEFORM FOR 40G Hz, PHASE C IN SHOP 


Magnetic Tape Recording in Aircraft S.N. 500 



115 Vac, 4 00 Hz 
Phase C 

P/N 464489 Unit 
Test Point TP-15 


FIGURE 5-44 

WAVEFORM FOR 400 Hz. PHASE C IN AIRCRAFT 

95 





Magnetic Tape Recording in Maintenance Shop 



115 Vac, 400 Hz 
Phase C 

Shop Test Stand 


FIGURE 5-45 

WAVEFORM FOR 400 Hz. PHASE C IN SHOP 


Direct Oscilloscope Recordings in Maintenance Shop 



I 



115 Vac,. 400 Hz, Phase C 
Radar Shop Mock-up 
Amplitude: 100 V/cm 

115 Vac, 400 Hz, Phase A 
Radar Shop Mock-up 
Amplitude: 100 V/cm 


Sweep speed: 0.5 ms/cm 


FIGURE 5-46 


115 Vic AT RADAR TEST STAND 






7 | 

• » 


1 




! 



mock-up with regulation provided by the AF/ECU-IOM power unit. A phase-loading 
difference of approximately 15 percent exists between Phase A and Phase C for 
the 115-volt, 400-Hz source. A high-resistance connection from neutral to ground 
at the generator or terminal strip will therefore result in abnormal differences 
between phase voltages. 

5. 5. 1.3 55/115-Vac Reference - The waveforms of the voltages as they 
appear in the aircraft are shown in Figure 5-47. Figure 5-48 shows the voltage 
waveforms as they appear in the radar shop. No changes in waveform were observed 
as a result of changing the loads by switching the radar. 

5. 5. 1.4 115 Vac, 1600 Hz Reference - Two photographs. Figures 5-49 
and 5-50, were taken of a magnetic-tape recording in the computer stand, they 
show a point where the peak-to-peak amplitude dropped to 150 volts for several 
cycles. Figure 5-51, taken in the radar shop, also shows this ripple. 

Figure 5-52 shows the waveform as it was recorded by magnetic tape in the 
aircraft. Figure 5-53 is a photograph of a magnetic tape recording of the shop 
mock-up waveform. 


5. 5. 1.5 115 Vac, l600 Hz Warm - The upper trace of Figure 5-54 shows 
the voltage waveforms (regulation provided by the AF/ECU-IOM power unit) with 
the radar transmitter turned off. Note the waveform distortion that occurs in 
the upper trace of Figure 5-55 (same point where Figure 5-54 was taken) when 
one radar transmitter is turned on: 

The l600-Hz voltage waveform in a 326 unit, removed from an aircraft for 
control-stick chatter in the assist mode, is recorded in Figures 5-56 and 5-57- 
In this case the 326 unit was found to have oscillation problems; the results of 
these oscillations may be seen in the photographs. 

5. 5. 1.6 115 Vac, l600 Hz Return - The lower trace of Figure 5-54 
shows the waveform, radar transmitter off, in the shop mock-up. However, with 
the transmitter on, harmonic distortion occurs, as shown in Figure 5-55* 

5.5.2 Unit Problems 

5. 5.2.1 P/N 464491 Reference Regulator Unit - The 491 unit develops 
a 55/115 volt, 1600 Hz reference (to +300 Vdc) voltage for use by computer and 
radar resolver circuits. These circuits are extremely critical to both noise 
and out -of -tolerance levels. A review of maintenance history of the 491 unit 
disclosed that the most frequent complaint to result from failure of this unit 
was "computer stop". It was also found that this unit has both a high bench- 
check-serviceable rate and a high repeat -write -up rate. Investigation revealed 
that an unwanted 40- to 100-Hz modulation component was present in both serviceable 
and unserviceable units. Comparison tests of 491 unite in the shop and in the 






97 


Direct Oscilloscope Recording In Aircraft S.N. 500 



Sweep speed: 0.2 ms/cm 


55-Vac, l600-Hz Reference 
P/N 464491 Unit 
Amplitude: 20 V/cm 

115-Vac, l600-Hz Reference 
P/N 464491 Unit 
Amplitude : 50 V/cm 


FIGURE 5-47 

A C INPUTS TO THE 491 UNIT ON AIRCRAFT 



Direct Oscilloscope Recording In Maintenance Shop 



Sweep speed: 0.2 ms/cm 


115-Vac, l600-Hz Reference 
Radar Shop Mock-up 
Amplitude: 100 V/cm 

55- Vac, l600-Hz Reference 
Radar Shop Mock-up 
Amplitude : 100 V/cm 


J 


FIGURE 5-48 


A C INPUTS TO THE 491 UNIT IN SHOP 



Magnetic Tape Recording in Maintenance Shop 


■ ' « c 





Sweep speed: 5 ms/cm 


115- Vac, l600-Hz Reference 
Computer Stand; 486101 Unit 
Amplitude: 100 V/cm 


FIGURE 5-49 

MODULATION OF 1600 Hz REFERENCE 







Magnetic Tape Recording In Maintenance Shop 



Sweep speed: 5 ms/cm 


115-Vac, 1600-Hz Reference 

Computer Stand; 486101 Unit 

Recorder Monitored In Space 
Normally occupied by the 
P/N 464296 Unit 

Amplitude: 100 V/cm 


FIGURE 5-50 

MODULATION GF 1600-Hz REFERENCE 


99 





Direct Oscilloscope Recording in Maintenance Shop 


115 -Vac, 1600-Hz Reference 
Radar Shop 
Amplitude: 20 V/cm 


A C RIPPLE ON 1600-Hz REFERENCE 


Magnetic Tape Recording in Aircraft S.N. 500 


115 -Vac, 1600-Hz Reference 
P/N 464489 Unit 
Test Point TP-16 


WAVEFORM FOR 1600-Hz REFERENCE IN AIRCRAFT 




Magnetic Tape Recording in Maintenance Shop 



115 -Vac, l600-Hz Reference 
Shop Mock-up, 

Test Point TP- 16 


FIGURE 5-53 

WAVEFORM FOR 1600-Hz REFERENCE IN SHOP 


Direct Oscilloscope Reloading in Maintenance Shop 


115 -Vac, l600-Hz Warm Line 
Radar Shop Mock-up 
Amplitude : 100 V/cm 


FIGURE 5-54 

WARM AND RETURN LINES: RADAR OFF 


Sweep speed: 0.2 ms/cm 


115 -Vac, l600-Hz Return Line 
Radar Shop Mock-up 
Amplitude: 100 V/cm 

Radar Off 


\ 

\ ■ 


A' 

/ • \ 

A . 




/( 

\ 

| 

/ 

\ 

- • \ 

■ v 







Direct Oscilloscope Recording in Maintenance Shop 



115- Vac, 1600- Hz Warm Line 
Radar Shop Mock-up 
Amplitude : 100 V/cm 

115 Vac, 1600 Hz Return Line 
Radar Shop Mock-up 
Amplitude: 100 V/cm 


Sweep speed: 0.2 ms/cm 


FIGURE 5-55 

WARM AND RETURN LINES: RADAR ON 


Direct Oscilloscope Recording in Maintenance Shop 



Sweep speed: 0.5 ms/cm 


115 Vac, 1600 Hz 

P/N 464326 Unit, Pin 1-FL1 

Amplitude : 100 V/cm 

115 Vac, 1600 Hz 

P/N 464326 Unit, Input at Tl, 

Pin 3-FL1 

Amplitude: 100 V/cm 

Note: No Load on ±50 Vdc outputs, 


FIGURE 5-56 


DISTORTION ON 1600-Hz LINE 




Direct Oscilloscope Recording in Maintenance Shop 





Sweep speed: 0.5 ms/ cm 


115 Vac, 1600 Hz 

P/N 464326 Unit, Pin 1, FL1 

Amplitude : 100 V/cm 

115 Vac, 1600 Hz 

P/N 464326 Unit, Input at Tl, 

Pin 3-FL1 

Amplitude: 100 V/cm 

Note: 100# load on -50 Vdc out- 
puts; no load on +50 Vdc outputs. 






FIGURE 5-57 

DISTORTION ON 1600 Hz LINE 

aircraft indicate that the aircraft equipment will tolerate the presence of a 
modulation amplitude of one volt or less. Units with a modulation level in excess 
of one volt would not perform properly in the aircraft. 

A number of the units that failed to operate in the aircraft were made usable 
by replacement of the magnetic amplifiers (AR1 and AR2). This did not eliminate 
the problem of modulation but reduced it to an aircraft -acceptable level. However, 
reducing the modulation amplitude to an acceptable level is not considered to be 
the solution to the problem, although it has reduced the unit repeat-write -up rate 
and increased the number of available, useable 491 units at Dover AFB. 

To eliminate the problem it will be necessary to perform a complete analysis 
of the associated circuitry to define the combination of components and conditions 
that are the cause of the problem. 

5- 5-2.2 P/N 464892 Unit - A test was conducted by ARINC Research 
engineers at Dover AFB to compare the regulation characteristics of the P/N 464892 
unit and the P/N 464992 unit. Two units of each type were tested on the LS-440 
load bank using a 056 generator with a constant speed drive as the source and a 
T-121 digital meter to monitor the output levels. The test results, presented 
in Table 5-6, show that the two 892 units tested did not perform as well as the two 
992 unltB, and support frequent reports of the poor regulation characteristics 
of the 892 units. An in-flight test of the 892 unit using the recording equipment 
was requested near the end of the contract period by SAAMA. Time constraints 
and limited aircraft availability made it impossible to perform this test. 



103 







TABLE 5-6 








COMPARISON OF REGULATED VOLTAGES 




Unit 

0 % Load 

25# Load 

50$ Load 

75$ Load 

IOO56 Load 

400 

Hz 

1600 

Hz 

400 

Hz 

1600 

Hz 

400 

Hz 

1600 

Hz 

400 

Hz 

1600 

Hz 

400 

Hz 

1600 

Hz 

464892 

S/N 21147 
S/N 23055 

117.0 

116.5 

116.0 

115.5 

116.5 

115.0 

114.0 

113.5 

115.0 

112.9 

112.0 

112.3 

113.8 

111.5 

110.0 

111.0 

112.5 

110.5 

109.0 

109.8 

464992 

S/N 193 
S/N 204 

116.5 

116.5 

116.5 

116.5 

115.2 

115.3 

115.5 

115-5 

114.2 

114.4 

114.5 

114.5 

113.7 

113.5 

113.2 

113.5 

113.0 

112.7 

113.0 

113.0 


For test purposes several 892 units were adjusted to the proper regulation 
level In the shop, Installed In an aircraft and rechecked. In general, the 
level dropped from the shop setting of 116 volts to 115 volts. This one-volt 
change alone would have little or no Impact on the power subsystem. However, 
this variation added to the poor regulation capabilities of the 892 unit frequently 
results In a system malfunction due to low voltage. The 892 units that have been 
removed from an aircraft as a result of a low voltage problem have been successfully 
reinstalled and used by adjusting for regulation levels of 117 volts as opposed 
to the prescribed levels of 116 volts. This test demonstrates the significance 
of a small difference between the test operation and actual flight. A change 
In the test configuration, a change in the shop regulation level, or the addition 
of an on-alrcraft adjustment capability would greatly improve this situation. 

5- 5- 2. 3 P/N 464992 Regulator Unit - The most frequent operator complaint 
associated with the 992 unit is an out-of-tolerance voltage condition. This 
condition is normally caused by variation in component operating characteristics 
due to aging. Problems of this type are corrected by component substitution, 
since this unit has no provisions for adjustment. The addition of adjustment 
capability would reduce maintenance time, the number of parts replaced, and the 
number of units returned to the special repair activity. 

5-5.3 Effects of Harmonics on the 289 Unit 

The Stable Element, P/N 464289, in the F-106 aircraft exhibited a high 
failure rate during the course of the ARINC Research investigations under Contracts 
AF09( 603) -48024, AF09(603) -60655, and F09(603)67-A-0003-0001. This difficulty 
was aggravated by the 289-unit repair limitations. The rate of arrival of repair- 
able units exceeded the rate of repair at the WRAMA SRA, resulting in a permanent 
overhaul and repair backlog. 


104 






To Improve these conditions a number of tasks were Initiated by WRAMA. One 
task assigned to ARINC Research was to define the possible effects of power 
deviations, particularly harmonics, on the reliability of the 289 unit. Information 
from Aircraft Power Specifications was correlated with the power actually experi- 
enced in a typical F-106. 

Harmonic content specifications for aircraft systems have changed since the 
F-106 was designed. Changes to the aircraft electrical load have also taken 
place. Conclusions must therefore be based on what constitutes a reasonable 
amount of harmonic distortion for this power subsystem and this stable -reference 
equipment. 

Military -Specification MIL-E- 7894 , 28 April 1952, was in force at the time 
of the original aircraft design. This was superseded by specification MIL-E-7894A, 
17 May 1955- The latter document was superseded by Military Standard MIL-STD-704, 

6 October 1959, which was in turn superseded by MIL-STD-704A, 9 August 1966 . 

That portion of each of the three documents that applies to harmonic distortion 
is quoted below: 

(1) From MIL-E-7894A, 17 May 1955: "3.2. 1.2 . 5 Waveform — The phase 

voltage waveform shall be within the following limits: 

(a) Amplitude factor: 1.4l ± 0.14 

(b) Harmonic content: The value of any harmonic shall not exceed 
5 percent of the fundamental." 

(2) From MIL-STD-704, 6 October 1959: "5.1.3. 5 Waveform — The voltage 

waveform shall be within the following limits: 

(a) Crest factor: 1.4l ± 0.1 

(b) Total harmonic content: 4 percent of the fundamental (rms) with 
linear loads, or 5 percent of the fundamental (rms) with non- 
linear loads, when measured with a distortion meter as distortion 
of the fundamental frequency. 

(c) Individual harmonic content: 3 percent of the fundamental (rms) 
with linear loads, or 4 percent of the fundamental with non-linear 
loads, when measured with a harmonic analyzer." 

(3) From MIL-STD-704A, 9 August 1966 : " 5 . 1.3* 5 Waveform — The voltage 

waveform shall be within the following limits: 

(a) Crest Factor: 1.4l + 0.15 (see 7 . 6 . 8 ) (7.6.8 Crest Factor . The 
Crest Factor limits specified in this standard assume that the 
crest factor limits at the terminal of electric power sources do 
not exceed 1.4l ± 0.10 and are degraded to 1.4l ± O .15 by the 
character of the loads. ) 



105 


(b) Total Harmonic Content: 8 percent of the fundamental (rms) when 
measured with a distortion meter as distortion of the fundamental 
frequency. 

(c) Individual Harmonic Content: 5 percent of the fundamental (rms) 
when measured with a harmonic analyzer. 

(d) Deviation Factor: In any event, the waveform shall not deviate 
from corresponding points of the fundamental by more than 5 percent 
of peak value of the fundamental." 

Note : MIL-STD-704A has been In effect for one year. During that time there 

has been an Increasing sentiment among electronic equipment manufacturers 
and users of airborne electronic equipment for a return to the more 
stringent requirements of MIL -STD -704. 

The Military Specification and the two Military Standards Indicate that the 
harmonic distortion in the F-106 aircraft should not exceed five percent. In 
addition, Kearfott Test Instruction E-1220, supplied by the equipment manufacturer, 
Kearfott Corp. (now Systems Division, Aerospace Group of General Precision Inc.), 
specifies that the a-c voltage waveform shall contain less than five-percent 
harmonic distortion. Power distortion below the indicated level should permit 
the Stable Element to perform within its accuracy specifications and should not 
degrade performance. 

Four a-c power waveforms were photographed between the 409 unit and the 
289 unit on Aircraft S/N 500, operating on ground power. These four waveforms 
were analyzed to establish the coefficients of the Fourier Series equation, 
which has the form: 


y = f(x) = A q + A.^ sin x + B-^ cos x + Ag sin 2x + Bg cos 2x 

+ A^ sin 3x + B^ cos 3x + . . . + A n sin nx + B n cos nx 


(1) 


From this general equation, the magnitudes of the coefficients allow quanti- 
fication of the harmonic content in each waveform. The phase B' and C' power 
is used to drive the gyro spin motors. Phase A' is used in the same manner, but 
this bus also serves as the 400-Hz, 115-volt return, and it is the center point 
of the aircraft three-phase, four-wire "Y" generators. Figures 5-58 and 5-59 
indicate the presence of some voltage distortion in phase B' and C' power. This 
distortion was defined as comprising odd-order harmonics. When the lower portion 
is superimposed on the upper portion, it is seen that the curves have symmertical 
positive and negative loops, excluding the possibility of even harmonics. 

Figure 5-58 shows the phase B', 115-volt, 400-Hz power supplied from the 
409 unit to the 289 unit at plug J28901, pin B. This photograph was reduced to 
a 35-millimeter slide and projected to approximately 20" by 25". Data points were 


106 




obtained from the projection, and a computer analysis of these resulted in the 
following Fourier Series equation: 

y = f(x) = 115.00 sin x + 0.79 cos x - 0.26 sin ?,x -0.11 cos 3x (2) 

+2.99 sin 5x - 0.77 cox 5x -1.39 sin 7x + 0.98 cos 7x 

- 0.20 sin 9x -0.70 cos 9x - O.58 sin llx + 0.49 cos llx 

The coefficent numbers are the rms voltage magnitudes of the components of the 
fundamental and the harmonics when the amplitude of the fundamental is 115 volts. 
This equation can also be expressed as only a sine function — plus phase angle — 
for the fundamental and each harmonic. In this case the equation becomes. 

y = f(x) = 115.00 sin (x + 0°24') + 0.28 sin (3x + 203°02') (3) 

+ 3.18 sin (5x + 345° 32 ' ) + 1.59 sin (7x + l44°58') 

+ 0.73 sin (9x + 254°29') + 0.75 sin (llx + 139°52') 

From Equation 3, each harmonic can be related as a percentage of the fundamental: 

Third Harmonic = 0.23$ Ninth Harmonic = 0.63$ 

Fifth Harmonic = 2.68$ Eleventh Harmonic = 0.65$ 

Seventh Harmonic = 1.47$ 

Each of the harmonics is below 5 percent. The total harmonic content is defined 
as the rms value of the Individual harmonics; it equals 3-2 percent. 

Figure 5-59 is the phase C', 115-volt, 400-Hz power supplied from the 409 
unit to the 289 unit at plug J28901, pin C. As in the previous case, a 35mm slide 
was projected and data points obtained. Computer analysis resulted in the following 
equation: 

y = f(x) = 111.1 sin x + 1.56 cos x - 0.62 sin 3x - 0.47 cox 3x (4) 

- 2.59 sin 5x - 1.10 cos 5x - 2.14 sin 7x - O.59 cos 7x 

- O.71 sin 9x - 0.88 cox 9x + 0.13 sin llx + 1.15 cox llx 

Equation 3 expressed as sine function plus phase angle is: 

y = f (x ) = 115.00 sin (x + 0° 47 1 ) + 0.78 sin (3x 217°l4') (5) 

+ 2.28 sin (5x + 336° 58' ) + 2.22 sin (7x + l85“26') 

+ 1.12 sin (9x + 308°55') + 1.15 sin (llx + 96°28') 

From Equation 4, each harmonic can be related as a percentage of the fundamental: 

Third Harmonic = 0.68$ Ninth Harmonic = 0.98$ 

Fifth Harmonic = 2.45$ Elevent Harmonic =1.0$ 

Seventh Harmonic = 1.93$ Total Harmonic Distortion = 3.3# 


107 


Each of the harmonics and the total distortion are, again, below the 5-percent 
level. 


f 

A 

i ’ 
t -» 

f: 

B i 

h 





\ - . 




i 



Figure 5-37 showed the voltage waveform for the 26-volt, 400-Hz 
the 289 unit. This voltage is developed from the aircraft 115-volt, 
phase B through a step-down transformer in the 409 unit. This power 
the 289 unit as the fixed field for the gimbal motors. 


input to 
400 Hz, 
is used in 


A 35mm slide of the waveform was projected; data points were obtained, and 
the computer analysis was performed. Careful examination of the waveform revealed 
that it does not have perfect symmetry about the horizontal axis (unequal positive 
and negative loops). Therefore, even-order harmonics do exist in the waveform. 

In this case the computer analysis resulted In a close approximation of the original 
curve (by including the first 15 harmonics). 

Since all harmonics through the fifteenth are present in the waveform, and 
there are thus 30 terms in the solution, it is simpler to list the coefficients 
of the Fourier Series in tabular form than to write the actual equation. The 
coefficients for the sin nx and the cos nx terms, along with the distortion for 
each harmonic are shown In Table 5-7- The largest harmonic coefficient was found 
to have an amplitude of less than one-half volt. This represents Just over two- 
percent distortion. The total distortion is the rms value of all harmonics; it 
equals 2.37 percent. 


J 

] 

.J 

J 

J 

J 

J 


TABLE 5-7 

PHASE-B VOLTAGE DISTORTION: 26 V, 400 Hz 

Harmonic 

"n" 

Coefficient 
sin nx 

"An" 

Coefficient 
cos nx 
"Bn" 

Distortion 

1 

26.00 

O.I 85 


2 

-0.155 

0.053 

O.OO 65 

3 

- 0.066 

0.139 

0.0059 

4 

0.045 

O.O 83 

0.0036 

5 

0.463 

-0.254 

0.0202 

6 

0.011 

- 0.002 

0.0004 

7 

- 0.120 

-0.025 

0.0047 

8 

- 0.026 

-O.O 87 

0.0035 

9 

-0.041 

0.075 

0.0033 

10 

-0.063 

0.023 

0.0026 

11 

C .021 

O.O 65 

0.0026 

12 

- 0.011 

-0.029 

0.0012 

13 

0.010 

-0.042 

0.0016 

14 

0.011 

-0.007 

0.0016 


D 


n 

H 

















The l600-Hz power is developed directly from the aircraft generator and Is 
supplied to the stable element through the 409 unit. In the stable element it 
is supplied to the gyroscope heaterB and accelerometers. 

The 115-volt, l600-Hz generator is arranged in the aircraft so that the 
center point is electrically at ground through the load. The 1600-Hz power is 
therefore carried by a twisted pair of shielded wires. One is designated the 
"signal" lead and the other the "return" lead (Figures 5-60 and 5-6l). Although 
the total signal is 115 volts, the scope was referenced to aircraft ground and 
therefore shows 58 volts rms in each photo. The two waveforms closely resemble 
each other, and the five points of inflection near the maximum (and minimum) are 
of interest. 

A 35mm slide of the waveform shown in Figure 5-60 was projected, and data 
points obtained for use in computer analysis. This wave is unsymmetrical about 
the horizontal axis and thus contains even-order harmonics. The fourier Series 
was developed through the twentieth harmonic, resulting in a close approximation 
of the original waveform. Table 5-8 lists of the coefficients of the Fourier 
Series for each harmonic and indicates the relative distortion. 


TABLE 5-8 

RELATIVE DISTORTION: 115 V. 1600 Hz 


Coefficient 

Coefficient 


Harmonic 

sin nx 
"A " 

cox nx 
"B " 

Distortion 


n 

n 


1 

115.00 

-4.87 

— 

2 

-0.262 

0.128 

.0025 

3 

-7.840 

1.290 

.0680 

4 

-0.220 

0.094 

.0024 

5 

-1.408 

3.810 

.0353 

6 

-0.422 

0.107 

.0038 

7 

-1.290 

-0.645 

.0125 

8 

-0.290 

0.023 

.0025 

9 

-0.045 

-0.428 

.0037 

10 

-0.084 

0.247 

.0023 

11 

0.128 

0.168 

.0018 

12 

-0.049 

-0.066 

.0007 

13 

0.107 

-0.066 

.0009 

14 

-0.090 

-0.121 

.0013 

15 

-0. 565 

O.276 

.0055 

16 

0.126 

0.275 

.0026 

17 

-0.034 

0.212 

.0019 

18 

-0.187 

-0.746 

.0067 

19 

-0.183 

-0.169 

.0022 

20 

0.186 

0.568 

.0052 


109 



FIGURE 5-58 

PHASE-B' WAVEFORM 

115 V, 400 Hz 
Plug J28901, Pin B 
Voltage : 100 V/cm 
Time: 0.5 ms/cm 

FIGURE 5-59 
PHASE-C 1 WAVEFORM 

115 V, 400 Hz 
Plug J28901, Pin C 
Voltage : 100 V/cm 
Time: 0.5 ms/cm 

FIGURE 5-60 
SIGNAL WAVEFORM 

115 V, 1600 Hz 
Plug J28901, Pin Z 
Voltage: 100 V/cm 
Time: 0.2 ms/cm 

FIGURE 5-61 

RETURN WAVEFORM 

115 V, 1600 Hz 
Plug J28901, Pin U 
Voltage: 100 V/cm 
Time: 0.2 ms/cm 

FIGURE 5-62 
PHASE-A 1 WAVEFORM 

115 V, 400 Hz 
Voltage: 1 V/cm 
Time: 0.2 ms/cm 










% 



t 


I 



fc. 


Table 5-8 shows that the third, fifth, and seventh harmonics represent the 
major contributors to the total distortion, each being greater than one percent. 

The total harmonic distortion for this curve equals 7.88 percent. 

To verify the analysis, tests were conducted on the system mock-up in the 
A&E shop at. Dover Air Force Base. The three-phase, 115-volt, 400-Hz power for the 
system mock-up was monitored on a Tektronix Model 564 oscilloscope and a Hewlett 

Packard Model 330D distortion analyzer. 
The waveforms appeared to contain 
slightly less distortion than those 
used to perform the analysis. Measure- 
ments were taken while the generator 
was at minimum load and also while the 
UHF and computer subsystems were applied 
as loads. The results of total harmonic 
distortion are recorded in Table 5-9. 
These results are in close agreement 
with those of the waveform analysis 
conducted during this program. 

Additionally, both lines of the l600~Hz, 115-volt signal were monitored at 
two points in the mock-up. The total harmonic distortion at the input to the 
stable element was found to be 3.8 percent on the signal side at plug J28901, 
pin Z, and 3*5 percent on the return line at plug J28901, pin U. At the power 
jacks to the mock-up, the l600-Hz signal read 4.6 percent total harmonic distortion 
on the signal side and 3-9 percent on the return line. Following these measure- 
ments, the radar equipment was turned off, and the distortion level at the signal 
jack to the mock-up was reduced to 4.1 percent. 

All measurements taken on the system mock-up were below the established level 
of five percent. In-flight recordings were made of the aircraft power on 
14 August 1967 . The flight altitude was 12,000 feet, and the mission profile was 
Radar Lead Collision. The mission was recorded on Code 1 (no malfunctions, and 
no maintena' e required. ) 

The 115-volt, 400-Hz power was recorded at the 489 unit. Figures 5 - 63 , 

5-64, and 5-65 show the phase A, B, and C waveforms as recorded. Visual inspec- 
tion of these waveforms indicates that they are no worse than those analyzed 
earlier in this report. However, a peak-to-peak amplitude variation was present, 
and these amplitude variations prevent Fourier Analysis since the waveforms are 
nonperiodic. The worst amplitude variation recorded was approximately 10 percent, 
as shown in Figure 5-66 (Phase C). 

The l600-Hz, 115-volt unregulated waveforms for the signal side and the 
return line are shown in Figures 5-67 and 5-68. The high levels of distortion 



111 



FIGURE 5-63 



HARMONIC DISTORTION: AIRCRAFT PHASE A 


115 Volts, 400 Hz 


HARMONIC DISTORTION: AIRCRAFT PHASE B 


115 Volts, 400 Hz 


HARMONIC DISTORTION: AIRCRAFT PHASE C 


115 Volts, 400 Hz 



I’j’U . ,.tl_ t ,, V 

1 . mkt* • x 

&*, 













1' [ \ 1 















— \ 



f ] *? ' Tv. 1 i\ 

















FIGURE 5 66 






HARMONIC DISTORTION: AIRCRAFT PHASE C 
(WORST CASE OF AMPLITUDE VARIATION] 


115 Volts, 400 Hz 



FIGURE 5 67 
HARMONIC DISTORTION: 
AIRCRAFT UNREGULATED SIGNAL SIDE 


115 Volts, 1600 Hz 



FIGURE 5 68 

HARMONIC DISTORTION: 
AIRCRAFT UNREGULATED RETURN LINE 

115 Volts, 1600 Hz 






discussed earlier (Figures 5-60 and 5-6l) did not occur during the in-flight 
recording, however, amplitude variations were present. 

The three-phase, 115-volt, 400-Hz aircraft power was monitored on aircraft 
S/N 500 on 10 October 1967. The amplitude variations noted on aircraft S/N 502 
(one of serial aircraft instrumented) in flight were not present on aircraft 
S/N 500. Photographs of the aircraft power showed a typically small amount of 
harmonic distortion (less than 4 percent). The aircraft became unavailable for 
testing before a complete set of photographs could be obtained. Therefore, the 
absence of the amplitude variations cannot be documented. 

The worst case of harmonic distortion recorded by photographic analysis 
or actual measurements was the 115-volt, l600-Hz power experienced by aircraft 
S/N 500 while operating under ground power at Dover Air Force Base. Since this was 
the only incident in which harmonic distortion was found to exceed the five-percent 
limit specified in the Kearfott Test Instruction, it is concluded that this was 
a unique condition associated with one specific generator. The manufacturer's 
specification indicates that the equipment can be expected to perform without 
degradation on power being generated for the F-106 equipment. 

The amplitude variation noted on aircraft S/W 502 is of concern. The three- 
phase, 400-Hz, 115-volt power is used in the 289 unit to power the gyroscope spin 
motors. It is assumed that these motors are of the synchronous type, and their 
speed is dependent only on input power frequency. Therefore, this amplitude 
variation should have no significant effect on the performance of the 289 unit. 

The effect of the amplitude variation noted on the l600-Hz power cannot be 
adequately determined, since the exact nature of the accelerometer and its circuitry 
are unknown at this time. However, these variations occurred on a flight that 
was scored as Code 1. In addition, the x- and y-axis signals ere demodulated 
in the 209 unit and became a d-c level. It is therefore concluded that the 
accelerometer circuitry is insensitive to these variations. 

5.6 Recommendations 

ARINC Research recommends the following corrections to problems that have 
been discussed in some detail earlier in this section of the report: 

(1) The clampax units (P/Ns 092, 591, 791* 891, and 991) should be added 
to the list of units to be checked during the 100-hour periodic 
inspection. 

(2) A resistance check of the ground-strap connection should be added to 
the 100-hour periodic inspection. The total rack-to-alrframe resistance 
should not exceed 0.01 ohms. 

(3) A requirement to clean and tighten the generator field and output 
connectors and the connections to the associated terminal strip should 

be added to the major Inspection procedures and to the routine associated 
with an engine removal or change. 


114 


(4) Replacing the l600-Hz generator field wiring with a larger wire size 
should be considered. 

( 5 ) The sensing -voltage pickup point for the l600-Hz regulator should be 
relocated to eliminate the requirement to share lines with the 692 and 
792 generators. 

(6) The Fault Detection Tester, Model MPM- 54 , should be used for noise-level 
tests on the power subsystem under actual on-alrcraft operating 
conditions. 

( 7 ) Test points should be incorporated into the P/N 486111 power subsystem 
stand to provide for monitoring the generator field-voltage characteristics 
for all MA-1 generators. 

(8) Availability of either the P/N 31056-002 and 464689-151 generator on 
the short-life ground -power unit or on the CSD-generator test stand 
for test and adjustment of the MA-1 field-voltage regulators should be 
assured. 

(9) The remote-load bank (P/N LS-440) used with the power-subsystem test 
stand for regulator checks should be checked to verify that the desired 
load is being selected. 

(10) The Technical Order procedures for maintenance and troubleshooting of 
the P/N 464992 a-c regulator unit should be updated and expanded to 
Include in-circuit voltages and waveforms. 

(11) Filter reactors (P/Ns 035 and 135) should be incorporated into the shop 
CSD-generator test stand to provide a more realistic generator test. 

(12) The load-bank test capability, LS-440 and 486111, should be modified 
so that 100 percent load test of the l 600 -Hz power is less than the 
present 63 . 0 amperes (or 60 percent more than the actual on-aircraft 
load). Preliminary findings indicate that a load corresponding to 
75 percent of present test full load would be adequate. This change 
could be accomplished by a simple wiring modification in the load 
bank. 

(13) A one-time test -inspection of all 035 and 135 units should be made. 

When all units are restored to the best possible condition, a preventa- 
tive maintenance schedule to prevent degraded performance should be 
considered. 

(14) A design review shall be undertaken for both the 892 and 992 units 
to stabilize the amplifier circuits. At the same time consideration 
should be given to relocating the voltage -adjustment controls, R2 and 
R20, to a front panel of the 892 unit. 

(15) Maintenance personnel should be advised of the problems that can be 
caused by using a magnetic amplifier of varied characteristics (produced 
by different manufacturers) when repairing the 692 , 792 , and 892 units. 


115 




4 



j 


% 


\ 




f 



(16) A detailed investigation into the exact cause of the low-frequency 
oscillation and modulation that occurs in the 491 unit should be under- 
taken. Preliminary studies indicate a resonant condition that can be 
corrected with a minor change of component values. 

(17) Adjustment provisions should be added to the 192 unit to compensate 
for component aging. 

(18) The neon voltage -reference diodes, V13, Vl4, V15, and Vl6, should be 
replaced with suitable solid-state devices. 

(19) The 19 2 unit should be modified to include decoupling circuitry in the 
-140-V bias-voltage input to the cathode of Vll-B (pin 8). 

(20) An adjustment should be added to the 292 unit to compensate for 
component aging and differences in tube characteristics. 

(21) Maintenance personnel should be certain that the 292 -unit vibrator 
units (Gl- G2 and G3- G4) are used in matched-0 pairs from the same 
manufacturer and with the same part number. 

(22) There is a design flaw m the voltage comparator and adjustment circuit 
of the 326 unit. It is recommended that, in addition to implementing 
the ARINC Research modification (Monthly Status Letter, February 1966), 
the following steps be taken: 

• Resistors Rl4 and R29 should be replaced with resistors of similar 
value, but with higher wattage rating. 

• Individual ground connections in the unit should be converted to 
one common ground point. 

• Tantalum capacitors C6 and C12 should be replaced with new solid- 
state components to eliminate the present problems caused by 
leaking capacitors. 

• The specified maximum-noise-level tolerance should be reduced from 
the present 200 mV to 50 mV for shop test of the unit. 

(23) A modification of the 292 unit to connect the screen grids of V7 - V8 
to the screen grids of V9 - V10 would reduce the oscillation tendencies 
resulting from unbalanced tubes. 

(24) Shop tests of the 292 unit should include monitoring both output voltages 
simultaneously on a dual trace scope, while the technician switches 
loads to all other d-c supplies in the test stand. 

(25) Front -panel adjustments should be provided on the 792 and 692 units 
to enable optimizing the output level in the aircraft. 

(26) Adequate overload protection should be added to the 115 V, l600-Hz 
input lines to the 092, 591, and 791 unit to prevent complete loss of 
the power subsystem and unnecessary component failures in the event of 
intermittent loss of the -140 or +300 Vdc generator power. 


I 

.1 

:i 

1 

j 

1 

j 



116 


(27) Additional decoupling circuits should be added to certain bias lines 
to reduce the interactions now present. The lines are the +300 Vdc 
line in the 791 unit and the -140 Vdc line in the 092, 591> 891, and 
991 units. 

(28) Procedures should be expanded to provide additional in-circuit voltage 
data for the clampax units to enable maintenance personnel to better 
determine the conditions of the 6094 tubes in the regulators. 

(29) The test-stand configuration should be changed to allow monitoring 

of unit test points without interaction with the basic voltage source. 

Note: The maintenance personnel at Dover Air Force Base have, on their 
own initiative, implemented many of these recommendations (other than unit modi- 
fications) in a local test program. The results at the time of this writing are 
inconclusive; however, they are encouraging. Early in the test program it was 
noted that there was a major reduction in the rate of aircraft aborts due to 
power dumps. There were only one-fifth as many aborts due to power dump during 
the month after the program began than there were during the preceding month. 

And this trend is continuing well into the second month of the test. The gains 
to be expected from this reduction are numerous, with the most important one 
being the increase in the number of successful missions. 


Wan 


fiyc ediA'Z Thpa - £ 


/V ? 





6. HUMAN FACTORS 

6.1 Introduction 

The results of previous ARINC Research work had Indicated that equipment 
performance is sometimes adversely affected by maintenance. For instance, it 
was found that up to 20 percent of the unreliability of certain electronic 
equipment installed in the B-58 aircraft could be attributed directly to previous 
maintenance activity rather than to the inherent unreliability of the hardware.* 

WRAMA's reasoning that hardware performance (viz., reliability) Is affected 
by the user and his maintenance procedures is supported by recent analyses of 
commercial aircraft electronics: 

"New data on avionic equipment failures indicate that 
the user and his maintenance procedures may have as 
great an effect on reliability as the manufacturer. 

This conclusion is based on widely different failure 
rates experienced by 21 airlines each using identical 
equipment. . . ."** 

On the basis of analyses performed by ARINC Research in a previous contract,! 
WRNEW Judged that equipment performance could be improved by modifying the 
maintenance situation. Accordingly, they established the following task: 

To develop modifications to procedures and techniques 
associated with the existing maintenance system, based 
on a detailed analysis of the problems associated with 
the existing system, and directed toward an overall 
improvement in maintenance efficiency and effectiveness. 

The statement of the project task gave rise to three general questions, 
which served as guidelines throughout the study: 

(1) To what degree does the maintenance system influence the effectiveness 
of the F-106 system as a whole? 

(2) How is the influence exerted? 

(3) How can the maintenance subsystem be modified to make the entire 
F-106 system more effective? 


* ARINC Research Publication 513-01 -5-672, B-58 Aircraft Avionic Subsystems 
Reliability and Maintainability Program , October 19bb. 

** Philip J. Klass, "New Data Yield Clues to Reliability," Aviation Week and 
Space Technology, 17 February 1967. 

t Final Engineeri ng Report, Reliabi l ity and Mai nt ainability Improvement Program 
for the F-lOb Avionics, ARlNC Research Publication 516-01-2-039, 19 July 1966. 


119 



Within the framework of these three critical questions, there were specific 
problems : 

(1) How can the number of flights rated as successful be Increased? 

(2) How can the "bench-checked-OK" rate be reduced? 

(3) How can aircraft availability be increased? 

These problems are discussed in detail in Section 6.2. 

Tasks were defined as follows: 

(1) Data acquisition 

(2) Data analysis 

(3) Formulation of conclusions and recommendations 

(4) Design for implementation and validation 

Data elements were defined in terms of origin and use. Most of the 
required data were already at hand, as a result of previous and current ARINC 
Research work. 

The data-acquisition effort is discussed in Section 6.3. 

Data analysis was divided into two broad categories: 

(1) Correlations between factors influencing total F-106 system effectiveness 

(2) Analysis of maintenance -information system 

Most of the quantitative analyses performed were of the first type. 

The data analysis is described in detail in Section 6.4. 

This study was concerned primarily with the problems of the F-106 maintenance 
system as it is now constituted. During the investigation, however, it was 
inevitable that investigators would encounter factors that Influence the subsystem 
under study but appear to be external to it. These factors nevertheless contribute 
to total system effectiveness and therefore should be studied. On example of such 
a factor is the Fault Detection Tester, an item of test equipment that is currently 
undergoing final acceptance evaluation by the Air Force. Another example is the 
pressure for good performance (of both men and equipment), which influences 
maintenance reporting. These and other such factors are discussed in Section 

6.5. 


120 




6.2 Problem Definition and Approach 


6.2.1 Problems Defined In Terms of Objectives 

6.2. 1.1 Increase In Probability of Next-Flight Success 

The probability of next-flight success in a given situation is the probability 
that the next flight follow ng that situation will be rated successful by whatever 
criterion is used for determining success. This probability is determined by 
computing the percentage of following flights that are rated successful. 

This definition of next-flight success contains two important concepts. 

First, there must be a specific criterion of success. It may be meaningful to 
call a flight a success even through the target was not successfully intercepted, 
if the radar functioned properly throughout the flight (i.e., the next-flight 
success probability of some subsystem can be treated separately from the next- 
flight success probability of the system as a whole). 

Second, success can be examined as a function of the state of the system 
prior to flight (which gives rise to the name next-flight success). Thus, if 
the state of a system is determined by its previous flight, the probability of 
next-flight success is related to previous flight performance. In other words, 
a sequence of flights can be considered to be a Markov chain, with each transition 
probability determined by the current state. 

The effects of maintenance on the transition probabilities of this Markov 
chain are not all clear. For instance, when maintenance involves the substitution 
of one or more components, is the system, in fact, still the same, with the same 
rules for determining transition probabilities? Because the number of possible 
states is large, an enormous number of observations would have to be made to 
determine a transition matrix. However, it is clear that next-flight success 
(according to whatever criteria are chosen) is a vital measure of system 
performance. First, as a measure of hardware performance, the probability of 
next-flight success for each subsystem is related to how well that subsystem is 
performing its function in the overall system, and is thus an important tool for 
the system analyst in determining overall system effectiveness and in analyzing 
possible improvements. Second, as a measure of maintenance performance, next- 
flight success as a function of maintenance state gives an indication of the 
effectiveness of maintenance. 

The primary problem of this study, then, was: How can the probability of 
next-flight success be Increased by means other than improving the intrinsic 
reliability of the equipment? 


121 





T 


6.2. 1.2 Decrease In "Bench-Checked OK" Rate 

The second major problem confronted In the study was: How can the number 
of units "bench -checked OK" be reduced without (l) modifying test equipment, 

(2) degrading the spares supply, (3) reducing system reliability, or (4) signifi- 
cantly reducing system availability? 

It is clear that what is sought is a way to reduce unnecessary shop work 
by reducing the number of "good" units processed through the shop. This reduction 
must be accomplished without performing trade-offs against equipment effectiveness. 
Furthermore, it has been demonstrated that the major cause of the high "bench- 
check OK" rate is the high rate of unit removals on the flight line, rather than 
faulty shop diagnosis. Thus the objective is to find a method of reducing the 
number of "good" units sent to the shop for maintenance without reducing equipment 
reliability. 

Within the equipment-effectiveness constraint defined above, the following 
four areas can be explored to meet this objective. 

(1) Can a flight-line maintenance strategy be devised that significantly 
reduces the total number of removals per repair action? 

(2) Can a replacement strategy be devised in which good units are reinstalled 
in the aircraft, rather than being sent to the shop? 

(3) Can the number of maintenance events in which no unit is removed be 
increased? 

(4) Can the total number of maintenance events be reduced? 

6. 2. 1.3 Increase in Total Aircraft Availability 

Since the effectiveness of the F-106 system is directly proportional to its 
availability to fly against a target, an increase in its availability must result 
in a corresponding increase in its overall effectiveness. In addition to ensuring 
that the system will be in good operating condition when flown, it is also 
desirable to increase the fraction of "up" time (availability), time during which 
the system is available to be flown. The question to be answered, then, is: 

Can the fraction of aircraft availability be increased without significantly 
increasing cost? 


6.2.2 Problems Defined in Terms of Elements Studied 

As well as defining problems in terms of objectives, it is possible to define 
them in terms of the elements studied. The basis for this study was the analysis 
of information transmission. It was found that the information system "rewards" 
some reports more than others (at least under certain circumstances), thus intro- 
ducing bias into the information recorded. The emphasis is placed on what the 
system seems to reward when its content is used as a measure of personnel 



J 

I 

J 

J 

J 

j 

J 

J 

.] 

j 


- 

.-i 

: 


122 




» 


performance, not directly on the "truth" of the data. The argument here Is not 
that maintenance personnel do not tell the truth, but rather that they are 
placed In an environment In which telling the truth Is at cross purposes with 
what the system rewards. 

Analysis of information suggests the following questions: 

(1) What Is the effect of failure information on the performance of 
maintenance? 

(2) What is the effect of the maintenance information system on the 
performance of maintenance? 

6.2.3 General Assumptions Limiting Objectives 

Certain assumptions were made that limited the objectives of this study, or 
at least provided guidelines within which the solutions to the problems were 
sought. The first assumption was that the quick -turn-around philosophy of flight- 
line maintenance is to be preserved. The P-106 electronics system is designed 
for "remove and replace" flight -line maintenance. Any changes in maintenance 
strategies must increase effectiveness of maintenance without reducing aircraft 
availability. The second assumption is that the objective should be primarily 
to increase operational -system effectiveness, rather than to concentrate on more 
efficient utilization of support -system manpower. Third it is assumed that 
recommendations for changes in maintenance strategy must consider the limits of 
the physical resources available (spare parts, equipment, and personnel). 

6.3 Data Acquisition 

This study was started with data already acquired by ARINC Research for 
previous work on the P-106 system. From the outset, it was known that certain 
data were missing. Therefore, a new data-acquisition effort was included in the 
early planning. 

6.3.1 Previously Acquired Data 

ARINC Research field engineers had acquired maintenance information on all 
maintenance events occurring during the period October 1965 to February 1966 
at Dover Air Force Base and Selfridge Air Force Base. Table 6-1 lists all of 
the variables on which data were acquired and which this study considered. These 
data include: (l) a pilot identification code, (2) a coded description of what 
the pilot said was wrong, (3) a code for the maintenance supervisor (and also 
for his senior helper, at Dover), (4) a statement of whether the malfunction was 
verified, (5) an identification of which units were involved in the action, 

(6) a coded description of what the maintenance man said he found, and (7) an 
indication of the shop disposition of each unit. 


123 


TABLE 6-1 

VARIABLES CONSIDERED IN FIRST DATA SAMPLE 


Variable 

Description or Explanation 

Base 

Dover, Self ridge 

Squadron 

71st, 94th, 95th FIS 

Pilot Number 

Locally assigned; not necessarily unique 

Aircraft Number 

Last three digits of serial number 

Aircraft Report 

Sequentially assigned by event within aircraft number 

Date 

Day, month, year 

Cumulative Plight Hours 

On this aircraft 

Mission Type 

Assigned Code* 

Mission Success 

Pilot's Rating cf System Readiness: 1, 2, or 3 

Mission Evaluation 

Aborted (where) or other 

Passes Attempted 

Number of intercepts attempted during mission 

Passes Completed 

Number attempted minus number of symptoms reported 

Report Type 

Symptoms classified In 7 general areas 

Reason for Report 

Assigned code 

Malfunction Number 

Consecutively within aircraft report number 

Symptom Code 

From AFTO Form 76-3 as reported by ADCR-66-28 Code 

When Discovered 

Taken from AFM 66-1 codes 

Severity of Complaint 

ARINC Research field engineer's rating: 1, 2, 3> or 4 

Verification of Complaint 

Assigned code 

Action Taken 

From T.O. IF -106-06 code 

Maintenance Concept 

Assigned code 

Method of Troubleshooting 

Assigned code 

Next-Flight Success 

Assigned code 

AFCS Subsystem Downtime 

Total maintenance delay time in tenths of an hour 

Work Unit Code 

From T.O. 1F-106-06 code for each unit involved 

Flight -Line Action for Unit 

Assigned code 

Unit Serial Number 

Last four digits of serial number 

Unit Report Number 

Consecutively within serial number 

Unit Cumulative Hours 

Estimated total operating hours 

How Malfunctioned (Unit) 

Symptom reported to shop on 210/211 form 

Reason for Action (Unit) 

Assigned code 

When Discovered (Unit) 

From T.O. 1F-106-06 code 

Action Taken/Disposition (Unit) 

Assigned Code 

Method of Fault Isolation (Unit) 

Assigned code 

Unit Active Repair Time 

Man-hours, for shop, in tenths of an hour 

Flight -Line Personnel 1 

Name code (Dover only) of maintenance supervisor for this 
action 

Flight-Line Personnel 2 

Name code (Dover only) of maintenance man No. 2 on this 
action 

Squadron -Flight 

A, B, or C according to work shift 


•These codes, along with a detailed description of data -collection procedures, are presented In 
ARINC Research Publication 518-01-2-639 Reliability and Maintainability Improvement Program 
for F-106 Avionics (prepared for WRAMA, July l$bb). 











6.3.2 New-Data Acquisition 

The new data were acquired at Dover Air Force Base for the period April to 
May 1967, with emphasis on the symptom report, verification report, action 
taken, and unlt(s) replaced. These data were evaluated to determine whether 
the symptom was actually reproduced during maintenance, the action taken was 
appropriate, and the unit(s) installed could have corrected the malfunction as 
reported. 

In addition to the hardware aspects of the maintenance reporting, ARINC 
Research desired information on the human influences in the reporting system. 

An attempt was made to isolate the factors that make some pilots better symptom 
reporters, since analysis had shown that the pilot was the greatest single factor 
contributing to next-flight success (see Section 6.4). Pilots were rated by 
debriefers according to cooperative spirit, ability to describe system malfunc- 
tions accurately, and willingness to assume responsiblity for pilot error. 

Pilots were rated "high", "medium", or "low" by six debriefers, and these ratings 
were converted into a numerical rating for each pilot. 

6.4 Analysis of Data 

6.4.1 Correlation of Factors Influencing System Effectiveness 

Various factors that Influence system effectivensss were believed to be 
related to each other. Statistical tests were performed to prove or disprove 
these hypotheses. The results obtained are described below. 

The identity of the pilot was related to the number of symptoms reported 
after a mission. This relationship was established by a chi-square test on the 
distribution of number of symptoms by pilot Identity. For pilots P^, P 2 , ..., 

P i’ P n’ the number of flights f^ was multiplied by the average number of 

symptoms reported per flight (by all pilots), and the resulting expected number 
of symptom reports (n^) was compared with the actual number (n^) in a contingency 
table : 


^( n i- n i )2 


Pilot 

P 1 

P 2 

• * • 

P n 

K i 





n i 





( H i _n i) 2 






The hypothesis of independence of pilot identity and symptom frequency was 
rejected at the 0.01 significance level, indicating either that some pilots are 
reporting too many symptoms or that others are reporting too few. In either 
case, it is clear that symptom reporting is not uniform. 


125 






'1 




! 





I 


The specific symptoms reported were strongly related to the pilot; different 
pilots appear to concentrate on different symptoms. This fact was determined 
by an information analysis of the table of symptom vs. pilot: 


n 

= X A ij 
j=i 

n 

X* ■ 1 4 u 

J=1 


The amount of information in the distribution by symptom (X's) plus the amount of 
information in the distribution by pilot (Y's) minus the amount of information in 
the joint distribution (A's) gives the influence of the pilot on the symptom. 

This number divided by the amount of information in the distribution of symptoms 
gives the degree of influence exercised by the pilot. (For a full discussion 
of this method of analysis, see Final Engineering Report, Reliability and 
Maintainability Improvement Program for the F-106 Avionics , ARINC Research 
Publication 518-01-2-639- ) 

The identify of the pilot was found to be related to the probability of 
malfunction verification (at the 0.01 significance level). 

The probability of malfunction verification is a function of the identity 
of the pilot who reported the symptom, more than of the symptom itself. 



m m 

V i-I A n yj-Z*u 

i=l m 


The pilot, the symptom reported, and the aircraft were all found to be 
related to next-flight success, with the greatest correlation being to the 
pilot. 


Table 6-2 shows the relationship of several variables that were thought to 
be possible predictors of system performance and several proposed measures of 
performance. It is significant that the variable that has the greatest correla- 
tion with each of the measures is the pilot. 


- 


U 

L 


L 

j 

L 

l 

£ 

£ 




126 




« » 


TABLE 6-2 

PREDICTOR VARIABLES VERSUS SYSTEM PERFORMANCE 

Predictor 

Variable 

Performance Measure 

Total 

System 

Downtime 

Malfunction 

Verification 

Next- 

Misslon 

Success 

Pilot 

35# 

51# 

29# 

Symptom Reported 

28# 

30# 

20# 

Maintenance Man 

29# 

13# 

9# 

Aircraft Number 

25 # 

24$ 

23# 

Repaired Element 

18 # 

N/A 

12$ 


The percentage shown is that part of the variance in the 
i criterion variable that can be accounted for by knowledge 

of the predictor variable. (Correlations between the 
predictor variables account for the fact that the sum of 
the individual predictor variances exceeds ] 0 u$. ) 

6.4.2 Analysis of Maintenance Actions 

This study was focused primarily on maintenance effectiveness (probability 
j| of successful repair). However, efficiency (manpower, equipment, and time 

utilization) was not ignored. To reach conclusions regarding profitable alloca- 
tion of effort, it was necessary to study the physical performance of maintenance. 

6.4.3 Analysis of Pilot's Symptom Reporting 

Debriefers 1 ratings of pilots were found to be related to symptom frequency 
and next-flight success, but the relationship was not obvious. This correlation 
can be useful only if the cause -effect mechanism can be discovered. The goal 
of future analysis in this area, then, should be to determine what makes some 
pilots better symptom reporters than others and how pilots can be trained to 
report better. 

6. 5 Peripheral Influences 

There are peripheral influences, not properly part of the maintenance system, 
that have significant effects on motivation in the performance of the maintenance 
system. There are also items of equipment (or modifications to equipment) which 
are not now part of the system but which, when incorporated into the system, can 
have a significant effect on performance. 

6 . 5.1 Unofficial Ratings 

Throughout the Air Force, there are official rating systems: pilots are 
rated on performance of technical (piloting) skills and on various military 
characteristics; officers in supervisory positions are rated by criteria of 
supervisory effectiveness as well as military characteristics; enlisted men 


127 









1 


at every level are rated by measures of performance. The pressure to achieve 
Is certainly felt at every level. 

In addition to these official ratings, other, unofficial ratings are being 
made all the time. Possibly, these should be properly called Judgments, rather 
than ratings; however, each man is constantly being Judged by both his peers and 
his superiors. When the pressures created by these Judgments are at cross- 
purposes with the objectives of the system, then the system suffers. For example, 
when a maintenance man feels pressure to work fast, he may tend to Judge reported 
malfunctions as "cleared" when in fact they are still present. A goal of super- 
visory personnel at all levels, then, should be to assure that pressures do not 
work against the interests of the system. 

6.5.2 Outside Demands 

An important Influence on the performance and reporting of maintenance is 
the demands placed on the maintenance system (and on the F-106 system) by the 
Air Defense Command. The level of effort and the level of reported performance 
are related directly to those demanded by higher -command levels. Thus when a 
maintenance officer is told that he must have a certain level of availability, 
he makes sure that his records show aircraft availability to be as required. 

This may involve demanding more or better work from his men, or it may involve 
some data manipulation to shown an increase. It is this latter situation that 
may prove harmful. When information "fudging" is encouraged to meet some criteria 
for acceptable performance, something is wrong with the information system. 

6 . 5.3 Spare-Parts Allocation 

During this study it was learned that the spare-parts-allocation policy is 
actually working against the philosophy of remove -and -replace for quick turn- 
around. The number of spares of each part (or component) allotted to each 
installation is determined by the past history of failures of that part*. However, 
items that are "bench-checked OK" are not recorded as failures for purposes of 
spares allocation. Thus the fact that the unit was removed from the aircraft 
(requiring an immediate replacement) and processed by the shop (requring some 
unit downtime) is not accounted for in the allotment of spares of that unit type. 
The result of this policy is that maintenance personnel feel it necessary to report 
a discrepancy on each unit removed so that their future spares supply will be 
adequate for the remove -and-replace maintenance they must perform. 

6.5.4 New Test Equipment 

During the Investigation, information was obtained on the Fault Detection 
Tester (FDT), a device used to test the functions of the MA-1 on the ground by 

*Air Force Manual 67 -I, Volume II, Chapter II, describes in detail the procedure 
for determining stock levels. 


128 



programmed simulated inputs and measurements of outputs. If the Fault Detection 
Tester proves itself to be reliable and practical to use, it will provide a 
check on the state of the system that will eliminate the need for the pilot 
in the information loop. This will not mean that pilot information will be 
completely ignored; however, the pilot may be influenced to report more accurately, 
knowing that he is subject to check by the FET. Similarly, although the main- 
tenance man will still have to verify the symptoms to repair the malfunctions, 
there will be more Incentive for him to repair only the real malfunctions. This 
observation applies also, of course, to any other test equipment that provides 
symptom information. 

6.6 Conclusions and Recommendations 

In Section 6.2, three objectives were stated: 

(1) An increase in the probability of next-flight success 

(2) A decrease in "bench-checked-OKs" 

(3) An Increase in total aircraft availability 

The following material answers the four questions raised in Section 6.2. 1.2 
with regard to these three objectives. Recommendations are made, and a plan for 
validating them is presented. Pilot/maintenance man motivation is also discussed. 

6.6.1 Reduction of Number of Removals 

Can a flight -line maintenance strategy be devised that significantly reduces 
the number of removals per repair action? It is ARINC Research's Judgment that 
it is not possible to reduce significantly the number of removals per repair 
action within the other constraints imposed. Such reduction would require the 
following: 

• Large increases in troubleshooting time 

• Extensive revision of troubleshooting methods 

• Many additional test equipments 

This is not to say, however, that the total number of removals cannot be reduced 
(see below). 

6.6.2 Replacement of Good Units 

Can a replacement strategy be devised in which good units are reinstalled 
in the aircraft rather than sent to the shop? Such a strategy should make 
significant reductions in shop checks and at the same time increase system 
reliability. If several units are removed from an aircraft before the malfunction 
is cleared, those units not responsible for the malfunction should be reinstalled 
in the aircraft. This procedure would require little additional flight-line 
maintenance time. It would reduce the number of wasted shop checks by a large 



T 


fraction, and it would result in better system performance because of the 
greater degree of system integrity preserved. 

6.6.3 Reduction of Repair Actions 

Can the percentage of maintenance events in which no unit is removed be 
increased? Here it is necessary to distinguish between a repair action and a 
maintenance event. A maintenance event occurs whenever a maintenance man 
investigates a symptom report; a repair action occurs whenever a maintenance 
man attempts to correct a malfunction. It has been noted both by ARINC Research 
and by Air Force maintenance officers that the percentage of symptoms that are 
verified is much higher than might be expected. This high verification rate is 
due to excessive confidence in the symptom reports of the pilots, coupled with 
pressure on the maintenance man to "fix" the equipment following each symptom 
report. In general the pilots' symptom reports are not infallible, and the 
maintenance men should be encouraged to be completely objective, rather than 
"proficient". Of course, this approach may be different to implement, because 
of the problem of quantifying the aforementioned confidence in symptom reporting, 
but it can be recommended that symptoms which are reported as "intermittent" 
should not be pursued unless they are obviously verified or are repeat write- 
ups, or they affect the safety of flight. This would entail a change in main- 
tenance philosophy in that management must be prepared to accept a larger number 
of repeat write-ups (other than safety). 

6.6.4 Reduction of Maintenance Events 

Can the total number of maintenance events be reduced? One approach to 
this problem is to increase the probability of next-flight success. Another 
approach is to reduce the number of spurious symptom reports. Through the 
institution of a team concept in the assignment of pilots, maintenance men, 
and aircraft, both of these improvements will be effected. First, the maintenance 
men will do better work because they will be more deeply involved in the perform- 
ance of the aircraft. Second, the pilots will submit better reports because 
they will be more closely concerned with the performance of the maintenance men. 
Ideally, this team consist of one pilot, one crew chief, and one aircraft. 

However, operational constraints make this impractical. Nevertheless, the Air 
Force can approach this ideal by employing the smallest groups possible and 
minimizing the cross-assignments from group to group. This concept is discussed 
further below. 

6 . 6.5 Delineation of Performance Goals 

A possible solution to the problem of inaccurate symptom reporting by the 
pilot is to delineate c3 early the objective of each flight. When a pilot is on 
a training flight, he should not be Judged by the standards applied to evaluation 
flights. If most flights were considered training flights, pilots would not feel 



t 

L 


[ 


— 

L 

i 


L 

l 

l 

[ 

1 

I 



130 


the same pressure to perform or place the blame; there would be less emphasis 
on ''blame." Likewise, the delineation of realistic goals for maintenance would 
tend to Increase the effort spent on legitimate repair work, and decrease that 
expended on "cover up" work. Thus the maintenance man's objective should be 
to assure that the equipment Is performing properly rather than to "clear a 
complaint. " 


6.6.6 Use of Test Equipment 

The use of test equipment. In addition to aiding the maintenance man in 
troubleshooting and repair, can also serve as a motivation for the pilot and 
the maintenance man to report more accurately. 

During this study, ARINC Research personnel visited Aircraft Armaments, 

Inc. ( AAI ) , where they were briefed on that company's Fault Detection Tester 
(FDT ) . From a human-factors point of view, the equipment design Is quite good. 

The displays are well arranged; controls are convenient; troubleshooting 
communication channels are quite effective. The primary concern regarding the 
Implementation of the FDT as a useful aid to maintenance Is Its plausibility. 

If any test equipment is to be effective, the user must have confidence in It. 
During early acceptance tests at Dover AFB, the FDT has experienced reliability 
problems (such as power-supply failures) that may create lack of confidence in 
its capabilities, in which case its usefulness will be limited. 

6 . 6.7 Validation of Recommendations 

The steps recommended above lend themsleves readily to statistical validation. 
In a short test program, the efficacy of each recommendation can be tested. 

6.6.7 . 1 Troubleshooting Strategies 

The Air Force can verify that the expense of redeveloping the troubleshooting 
strategies for flight-line maintenance would far outweigh the benefits to be 
gained. The value of the improvement must be measured in terms of man-hours 
saved, plus the additional reliability to be achieved by making fewer changes 
in equipment integrity. ARINC Research believes that changes in the trouble- 
shooting strategies could not significantly reduce the number of removals. 

Appendix B describes a method for determining the optimum troubleshooting 
strategy for a particular symptom. The information required for this method 
includes the expected cost of checking each unit and the probability that, the 
given symptom, the unit is the malfunctioning one. Even under the assumption 
that unit checkout costs are equal, the information requirement for computing 
troubleshooting strategies is substantial. Under present conditions of manpower 
utilization, it does not seem worthwhile to expend the money and effort that 
would be required for a small improvement. Of course, this Judgment is subject 
to review in the light of new analysis techniques or new requirements. 


131 


6. 6. 7. 2 Replacement Strategy 

The policy of reinstalling good units in the aircraft after the malfunctioning 
units have been identified would reduce shop time (which would be significantly 
greater than the extra time spent at the flight line ) and would simultaneously 
result in better performance. The change in performance, as well as the time 
saving can be determined by a simple experiment. At a base where other conditions 
can be kept equal, the maintenance jobs are divided into two sets. One set 
is completed by substituting spares for all units removed from the planes. The 
other set is completed by reinstalling all good units in the aircraft after the 
faulty ones have been found. Then, with a large enough sample of maintenance 
actions to produce statistical significance, the following are determined: 

• The extra cost, in man-hours, of the replacement strategy 

• The saving in shop man-hours 

• The change in reliability of the equipment worked on 

6. 6. 7. 3 Verification Requirements 

The institution of more realistic criteria for verification and a more 
realistic attitude about symptom reports will significantly reduce the number 
of events in which units are removed. This is especially true for symptoms 
reported as "intermittent." This change would result in substantial savings 
of flight-line and shop time; it is also believed that it would not adversely 
affect equipment reliability. This conclusion could easily be verified by 
observing the next-flight success of maintained and unmaintained equipments 
following "intermittent" symptom reports. Some sample (random) of "Intermittent" 
symptoms would be checked out with rigorous criteria, and if they were not 
verified, no maintenance would be performed. The observations would be easy 
to test by a chi-square test for significance in the following form: 


Event 

Number of 
Occurrences 

Number of 
Next -Plight 
Failures 

Intermittent failure not 
reproduced, but some repair 
action taken 



Intermittent failure not 
reproduced, no repair attempted 




6 . 6 . 7 . 4 Team Assignments 

To validate the recommendation concerning team assignment of planes and 
men, such a system would be inaugurated for some of the men and planes at a 
base, while other assignments would be made on a random basis. At the end of 
the test period, the effectiveness of the two approaches would be compared 
according to the criteria of next-flight success, mission success, maintenance 


132 




man-hours, etc. It is crucial that this validation experiment he carried out 
in a situation in which the influence of other factors, such as mission type 
and flying conditions, will not affect the comparison. 


133 




/VO 



7 SPECIAL TASKS 

In this program WRAMA Engineering assigned five special tasks to ARINC 
Research. The results of these tasks were reported in detail to WRAMA in attach- 
ments to the Monthly Status Letters. They are summarized briefly below. 

7.1 Evaluation of the General Electric Rapid Tune Test Set 

7.1.1 Task Definition 

ARINC Research was directed to evaluate the proposed General Electric Rapid 
Tune Test Set to define its advantages and disadvantages as compared with current 
test methods. A secondary consideration was to consider possible Improvements 
through changes in the production model of the teBt set. 

7.1.2 Task Summary 

General Electric designed and constructed an engineering model of a test 
set for use in conjunction with the Rapid Tune units (P/n's 464541, 464641, and 
t 464741) of the F-106 radar system. 

The test set was partially evaluated in an earlier test program by the Air 
Force. ARINC Research completed the evaluation by comparing the results obtained 
with the tester with those obtained through current test methods. Major consid- 
erations for the ARINC Research evaluation were the following: 

• Impact on the operational system 

• Compatibility with the system 

• Test accuracy and fault-detection capability 

• Tester reliability and maintainability 

7.1.3 Conclusions 

The proposed test set, with minor modifications as noted below, will generally 
fulfill the requirements for field-base repair and alignment of the three General 
Electric units (P/N's 464541, 464641, and 464741) employed in the post-Group II 
configuration of the F-106 MA-1 system. 

Specific conclusions are as follows: 

• The tester will greatly relieve the present work load on the Radar- IR 
test stand. 

• Unit standardization will be significantly improved since each unit will 
be checked against a standard unit rather than against units in another 
system. This testing approach provides uniformity of performance through 
the elimination of interface variables. 


135 



. A greater degree of accuracy can be obtained with the GE teeter 
because of Its more accurate read-out. 

• Test and alignment times are reduced through the use of the GE tester. 

• Malfunctions or marginal operation that can reduce system performance 
but are not readily detectable through current methods are easily 
detected by use of the GE tester. 

7.1.4 Recommendations 

The following recommendations are made: 

• The present method of providing additional cooling air to the 641 unit 
during test should be modified. 

• The present method for mounting the 64l unit should be modified to allow 
360° rotation of the unit for better maintenance access. 

. Procedures should be instituted to permit use of the extender boards 
supplied with the test stand. 

• The test cable used with the 541 unit should be lengthened. 

• The necessary wiring should be installed and procedures instituted to 
permit measurement of Threshold No. 1 and No. 2 voltages, and to elimi- 
nate the present requirement for an external voltmeter. 

• The module mounting screw in the tester should be changed to permit 
connector alignment prior to screw engagement. 

• The procedures should be modified to ensure that the 74l unit will 
provide the proper filament voltage to the 641 unit in the aircraft. 

(A similar change may be required for the F-101/MG-13 system.) 

7 .2 Reliability Investigation of IFF Control Switch 

7.2.1 Task Definition 

WRAMA, WRNEW directed ARINC Research to investigate the failure of the 
control-switch in the P/N 464555 unit. 

7.2.2 Task Summary 

Maintenance records at Dover Air Force Base were reviewed. During an 
eight-month period three switch failures had occurred. In all three instances, 
the failure was the result of rotating the switch beyond its limits in a counter- 
clockwise direction. 

Two conditions are believed to contribute to this type of switch damage: 

• The decals for an adjacent switch are misleading to untrained personnel. 

• No mechanical stop is provided to prevent the switch from being rotated 
beyond its limit in the counterclockwise direction. 


I 

J 

d 

j 














136 


7.2.3 Recommendations 


Figure 7-1 shows the front panel in its present configuration with the 
"OUT" decal for the toggle switch in a horizontal plane. Because of its proxim- 
ity, the "OUT" position appears to be related to the IFF switch rather than the 
toggle switch. 


This confusion can be eliminated, as shown in Figure 7-2, by displaying 
the "OUT" decal in a vertical plane which will disassociate the position from 
the IFF switch completely. 


An additional safeguard can be provided by mounting a mechanical stop pin 
to the front panel, as shown in Figure 7-2, which will prevent rotating the 
switch beyond its intended limits. 



PRESENT CONFIGURATION 
OF FRONT PANEL 


Stop Pin 



FIGURE 7 2 

RECOMMENDED CONFIGURATION 
OF FRONT PANEL 


The decal for the toggle switch can be changed with a minimum of cost and 
should provide the desired results. The addition of the stop pin is a slightly 
more complex modification; however, it will eliminate the possibility of damaging 
the switch by rotating it beyond its limits. 


137 






7.3 Evaluation of the Fault Detection Tester (FDT) and IRAM Computer 

7.3.1 Task Definition 

ARINC Research was directed to review available technical information to 
define the extent to which the FDT or the IRAM Computer will compensate for 
deficiencies in the Short System Ground Check (SSGC) and self test as defined in 
earlier reports.* 

7.3.2 Task Summary 

One of the tasks assigned ARINC Research under Contract AF 09(603)-60655 
was to "determine the effectiveness of the aircraft Short System Ground Check 
(SSGC) and the Automatic Flight Control System (AFCS) self test to identify a 
system and/or a unit failure . " 

The original task was directed to the tests associated with the Automatic 
Flight Control Group (AFCG), Stable Coordinate Reference Group (SCRG), Air Data 
Computer Group (ADCG), and Stability Augmentation System equipment. The task 
assigned under the current contract was to define which of the previously reported 
deficiencies will be eliminated through the use of the Aircraft Armament Inc. (AAl) 
Fault Detection Tester (FDT) and through the installation of the Improved Relia- 
bility and Maintainability (IRAM) Computer developed by Hughes Aircraft Company. 

The technical documents used for reference in this task were as follows: 

• Hughes Aircraft Company, Report FD 30678-907, 15 March 1966 

• ECP HUG (MA-llASQ-25) 1447, 22 February 1966 

• ARINC Research Problem Report MI-04-5, part of the Twenty-Second F-106 
Status Letter, 5 January 1966 

« Fault Detection Tester, AN/MPM-54, Aircraft Armaments, Inc. ER-3272 
(no date indicated) 

• ADC Final Test Report, Project ADC/73AD/64-19, 23 February 1965 

7.3.3 Task Findings 

7. 3. 3.1 Short System Ground Check (SSGC) 

The Short System Ground Check is designed primarily for rapid GO-NO-GO test 
of the F-106 AWCIS. In the event of a malfunction, indicated by failure of a 
test, a fault indication will isolate the malfunction to a subsystem or to an 
area within a subsystem. 

The capabilities of the SSGC were defined in relation to the particular 
equipment groups under test. In each of these areas the statements of deficiencies 
are quoted below from the earlier study. Each quotation is followed by a descrip- 
tion of the most recent findings. 

* ARINC Research Publication 518-01-2-639, Final Engineering Report, Reliability 
and Maintainability Improvement Program for the F-106 Avionics. 10 July 1986. 

ARINC Research report on Task MT-04-5 was included in the 22nd F-106 Status Letter. 


138 





* • 






Stable Coordinate Reference Group (SCRG) 

(1) "The SSGC will not detect marginal equipment performance or gyro 
precession (drift) malfunctions." 

More accurate measurement will be possible with the IRAM Computer. 

This will Improve the capability to detect marginal conditions. 

(2) "Malfunctions cannot be isolated to a specific unit." 

Neither the FDT, nor the new computer change will improve the 
malfunction-isolation capability. 

Automatic Flight Control Group (AFCG) 

(1) "The SSGC is not considered to be a valid AFCS test without the use 
of the Mobile Radiation Test Set (MART cart)." 

Maintenance personnel at the bases under ARINC Research surveillance 
do not, as a general practice, use the MART cart. This factor is not 
expected to change with availability of the new Computer or the FDT. 

(2) "The SSGC will not detect marginal equipment operation in the AFCG." 

More accurate measurement will be possible with the IRAM Computer. 

This increased capability will improve the capability to detect 
marginal conditions. 

Air Data Computer Group (ADCG) 

(1) "Air data from the transducers are simulated by the SSGC. As a 
result, faults in the transducers will not be detected." 

Transducer faults can be detected by use of the FDT since the 
transducers' outputs are checked. 

(2) "Improper positioning of analog computer servo mechanisms could not 
be detected by the SSGC." 

The IRAM Computer will provide the capability to detect this type of 
malfunction. 

(3) "Computer malfunctions Invalidate the SSGC for the ADCG approximately 
50 percent of the time . " 

The expected high reliability of the new computer will greatly improve 
this situation. 

7. 3. 3. 2 Self Test 

No change in the self-test features were indicated in the technical material 
used in this task. 


139 





7 .4 Investigation of Ground Loops 

7.4.1 Task Definition 

ARINC Research was directed to investigate the feasibility of improving 
the reliability of the stable reference equipment by modification of the ground- 
ing and shielding circuitry. 

7.4.2 Task Summary 

Several problems associated with the 464009, 464109, and 464309 units of 
the Stable Control Reference Group (SCRG) were Identified during earlier con- 
tract activities. Initial investigations indicated that interference was being 
introduced into the system as a result of the grounding and shielding techniques 
being used. 

Laboratory tests conducted by ARINC Research showed that connecting the 
signal ground to the chassis ground significantly reduced the interference pres- 
ent on the input and output signals. On the basis of these findings, field 
tests were designed for the three units to determine the feasibility of solving 
the problem by using this grounding technique. The 464009 unit was selected as 
the first unit to be tested since the high-gain amplifiers used in this unit 
were more susceptible to noise than the lower-gain amplifier used in the remain- 
ing units . 


7.4.3 Conclusions 

A variety of changes were tested. The best changes from the viewpoint of 
cost and complexity of the modification were found to be the following: 

• To connect instrument ground to chassis ground 

• To connect d-c ground to chassis ground 

The overall reduction in Interference gained through these wiring changes 
eliminated the requirement for modifications to the remaining two units 
(P/N's 464109 and 464309). 

7.4.4 Recommendations 

It is recommended that the P/N 464009 unit be modified as follows: 

(1) Locate the Electrical Components Bracket Assembly, P/N 333016. 

(2) Locate the mounting bracket, P/N 333017, and identify the end 
containing six insulated terminal studs. 

(3) Locate the insulated terminal studs identified as E3, E4, and E6. 

(4) Connect and solder a suitable length of No. 22 awg black, 19-strand 
Teflon-wire (MIL-W-I 6878 D Type E or equivalent) between terminal 
studs E4 and e 6 located in Step 3. 



I 

i 

J 

I 

I 

I 

1 

I 

I 

I 

I 

I 


140 



f . 

** . 


t ' ! 


4 • 


(5) Connect and solder a suitable length of No. 22 awg black, 19-strand 
Teflon-wire (MIL-W-I 6878 D Type E or equivalent) between terminal 
studs E3 and E 6 located in Step 3 . This completes the modification. 

7.5 Investigation of Rotary-Joint Failures 

7.5.1 Task Definition 

ARINC Research was directed to determine if rotary-joint failures were a 
problem at the P-106 bases under ARINC Research surveillance. 

7.5.2 Task Summary 

At the request of WRAMA engineering a search of maintenance records was 
conducted at Dover and Tyndall Air Force Bases by ARINC Research personnel. 

The findings of this search indicate that there was no rotary- joint-failure 
problem at either of the two bases. On the basis of this finding, WRAMA 
directed that the task be closed. 







« 






APPENDIX A 

PART FAILURE RATES USED IN RELIABILITY PREDICTIONS 
FOR F-106 ELECTRONIC SYSTEMS 

1 . Sources of Data 

The part failure rates used in the reliability predictions for the F-106 
equipments are presented in Table A-l. All the malfunction rates, except as 
noted, are based on information contained in ARINC Research Report "Prediction 
of Field Reliability for Airborne Electronic Systems," 31 December 1962*. Part 
failure data presented in that report were accumulated by ARINC Research Corpora- 
tion during surveillance of B-52 aircraft at Walker Air Force Base for a period 
of nine months ending in March 1962. Equipments represented in the B-52 study 
were the AN/ARC-3 2 *, AN/ARC-65, AN/ARA-25, AN/APX-25, AN/APN-89, AN/APN-89A, 
AN/ASB-4, and AN/ASB-4A. The part data obtained from other sources, as referenced 
in Table A-l, are from other airborne equipments studied, or were already modified 
for the airborne environment. Only the basic part failure rates from the Martin 
Handbook required special modification for F-106 predictions. 

2 . Adaptation of B-52 Data to F-106 Equipments 

A portion of the given part reliability data accumulated in the B-52 program 
is in Table A-2. The diversity of B-52 systems (as to types and applications) 
from which the malfunction rates were derived provide these rates with "built-in" 
averaging factors, with respect to stress and maintenance conditions, similar to 
those of the F-106 systems. Therefore, the B-52 data are the most logical choice 
of available data for application to the F-106 systems. 

In the derivation of the part reliability data in Table A-2, it was assumed 
that in any one maintenance action one part alone was the actual cause of equip- 
ment malfunction. Any other part failures reported were considered secondary 
failures or mishandling accidents during maintenance operations. Thus the data 
were corrected for clustering effects (l.e., multiple parts actions in any one 
maintenance action). In addition, the failure-rate data of Table A-2 reflect the 
effect of adjustments performed during the correction of equipment malfunctions. 

In Table A-2 a distinction is made between part failure rate and part mal- 
function rate, which is the sum of the part failure rate (Column 9) and the part 
adjustment rate (Column 10). This gives a more realistic rate for adjustable 
types of parts than otherwise would be obtained. The instantaneous malfunction 
rates of the various parts were assumed to be constant. Justification for this 
assumption is rather tenuous for some part types, with respect to the physics- 
of-fallure implications associated with constant malfunction rates; this is 

* ARINC Research Publication 203-1-344, Prediction of Field Reliability for 
Airborne Electronic Systems , 31 December l9b2. 

A -3 


1 


TABLE A-1 


PART MALFUNCTION RATES FOR RELIABILITY 

PREDICTIONS OF F-106 ELECTRONIC SYSTEM 


Malfunctions 


Malfunctions 

Part Category 

Per Hour < 

Part Category 

Per Hour r 


(Multiply by 10” D ) 

(Multiply by 10” ) 

Accelerometer 

280.00 (1) 

Electron Tubes 
( continued) 


Bearing 

50.00 

Subminiature, 

33.25 

Capacitor, Fixed 


Amplifier 

Ceramic 

0.54 

Subminiature, 
Rectifier & Gas 

51.44 

Electrolytic 

11.00 

Filter (Band Pass, 
Harmonic RF Inter- 


Glass 

0.70 

2.75 

Mica 

0.34 

ference ) 


Paper 

0.95 

Filter, Mechanical 

30.00 (1) 

Plastic Film 

43.60 


Tantalum 

21.93 

Gear Assembly 

90.00 (1) 

Capacitor, Variable 

4.74 

Gyro 

490.00 (l) 

Chopper 

50.00 (1) 

Heater 

15.53 

Clutch Magnetic 

60.00 (1) 

Inductor 

5.15 

Connectors 


Meter, Electrical 

12.19 

Coaxial Plug 

7.19 

Motors 


Coaxial Receptacle 

6.04 

A-c Blower 

142.33 

Other Type, 

0.02/Pin 

D-c Bxower 

108.28 

Receptacle 


Other, A-c 

75.82 

Other Type, Plug 

0.02/Pin +1.15 

Other, D-c 

7.26 

Counter, Mechanical 

2.44 

Motor-Generator 

216.78 

Crystal, Quartz 

3.04 

Relays 


Diode 


Rotary 

206.34 

Regulator and 

17.70 (2) 

Switching 

146.87 

Rectifier ( 1A) 

Low Current ( 1A) 

10.20 (2) 

Time Delay 

82.40 

Rectifier, 

12.71 

Electron Tubes 


Selenium 


Cathode Ray 

705.97 

Resistor, Fixed 


Klystron 

1,191.75 

Wlrewound 

9.15 

Klystron, 

Adjustable 

1,334.76 

Other 

0.79 

Magnetron 

2,783.02 

Resistor, 


TR 

423.10 

Variable and 
Potentiometers 


Miniature, 

Amplifier 

42.15 

Carbon 

297.60 

Miniature, 

100.38 

Composition 


Rectifier & Gas 


Carbon Film 

536.85 









TABLE A-1 (CiitiimJ) 


Part Category 

Malfunctions 

Per Hour c 

(Multiply by 10 ) 

Part Category 

Malfunctions 

Per Hour r 

(Multiply by 10“ ) 

Resistor, 


Synchros 


Variable and 




Potentiometers 


Resolver 

103.28 

(continued) 


Other 

26.10 

Wirewound 

336.19 

Thermister 

60.00 (1) 

Wirewound, 

613.35 



Infinite 


Transformers 


Resolution 






High Voltage 

31.10 

Socket 

0.43 (3) 

Power and Filament 2.27 

Solenoid 


Other (IF, Audio 

1.34 



etc . ) 


Axial 

30.87 



Rotary 

227.00 

Transistors 


Switch 


High Power ( 5w) 

20.30 (2) 



Low Power ( 5w) 

7.50 (2) 

Cam 

1,448.26 



Commutator Type 

1,568.57 

Wave Guide 


Micro switch 

73.32 

Fixed 

110.00 (1) 

Pushbutton 

56.52 

Flexible 

284.00 (1) 

Rotary 

8.40 

Rotary Joints 

127.00 (4) 

Sliding Action 

73.32 



Thermostat 

2.72 



Toggle 

4.09 




(l) "Reliability Application and Analysis Guide," The Martin Company, July 1961. 


(2) "Semiconductor Reliability," final report under Contract NObsr-87664, ARINC 
Research Publication 239-01-4-383, 31 July 1963. 

(3) "Operational Reliability Estimates and Part Failure Rates for Naval Avionics 
Equipments," ARINC Research Publication 202-1-331, 15 November 1962. 

(4) "A Preliminary Study of Microwave and Transmitting Tubes, Semiconductors, 
Relays, and Other Parts," ARINC Research Publication 123-6- 189, 

30 September i960. 

Notes: (a) A gear assembly is considered to be equivalent to six gears or less 

and associated parts. 

(b) Fuses will not be included in part counts. 

(c) Lamps will not be included in part counts. 
















TABLE A-2* 

PART RELIABILITY DATA FOR AIRBORNE ELECTRONIC SYSTEMS 


Net* 

Reference* 


Me tor 


Motor h Pump, AC 
Servo or Set, *C 
Serve or Set, DC 
Tlaittf or Clock, AC 
Timing or Clock, DC 


Rotary 

Switching 

Switching, dry circuit 
Tin* Delay 


NOTES FOR TABLE A- 2 


The adjustment data for this 
category has been presented on 
a separate line since it per- 
tains only to plugs. For the 
malfunction rate of plugs, the 
adjustment rate must be added 
to the failure rate; the mal- 
function rate of receptacles 
Is equal to the failure rate 
alone. 


No replacements or repairs were 
observed for this part type. As 
a conservative^ estimate of the 
failure rate, ^i, the upper 50 $ 
confidence limit on zero observa- 
tions may be used. This value Is 
computed by the equation 


Letters (a) and (b) refer 
respectively to parts groups 
1 and 2 as defined in Table A-2 


Refer to Section 3, Part B of 
the report proper for optional 
procedure for modifying the re- 
placement/repair/failure values 
(columns 4,5,7 and 9) of this 
part type to account for abnormal 
stress and operating conditions. 
See page 6 of the report for 
remarks concerning the optional 
nature of these modifications. 


Observed part hours 


The corresponding confidence 
statement is P(0 4 k, s ■■ 


The 'part hours' shown on this line 
are pin and socket hours for the 
plugs and receptacles. Correspond- 
ingly, the value in column 9 repre- 
sents the rate of plug and recep- 
tacle failure per pin and socket 
hour. To obtain the estimated 
failure rate for a plug, therefore, 
the number of pins is multiplied 
by 0.02; similarly, for a receptacle, 
the number of sockets is multiplied 
by 0.02. 


•Abstracted from ARINC Research 
Corporation Publication 203-1-344 
Prediction of Field Reliability 
for Airborne Electronics Systems, 




Ocaerved Data 


2 

5 4 S 6 

Part Category 

Part Type 

Part Hour*, Ti Part* Part* Part* 

(Millions) Replaced Repaired Adjusted 



On-Off Cycling Rat* - 0.15 per Hour 









especially true for part types that exhibit strong mechanical wearout character- 
istics. This should not, in general, cancel the value of the failure rates, 
except in chance circumstances where a severely unfavorable combination of parts 
should appear. 

In developing the part malfunction rates for the P-106 equipments, data for 
a number of similar parts, with similar malfunction rates, were combined. 

Regarding the part types that exhibited no malfunctions in the B-52 sur- 
veillance, when data on these were used for compiling malfunction rates for 
application to P-106 equipment predictions. Note 2 of Table A-2 was taken into 
consideration. 

To account for the deleterious effect of different equipment on-off cycling 
rates the B-52 data were modified by the factor 1 + 8N, where N is the number of 
on-off cycles per hour. The factor for the P-106 calculated to be a value of 
5 (N = 0.5 cycles per hour). 

3 . Use of Other Part-Reliability Data Sources 

Part malfunction rates listed in Table A-l, which are not based on B-52 
data, are representative of airborne environments. However, no attempt was made 
to modify the original data, to account for clustering and cycling effects, as 
was done for the B-52. Factors were used to transform the values for airborne 
environment use when the data were not originally generated as such. 


Operate hours and removal data, from which the semiconductor malfunction 
rates were derived, are summarized in the following: 


Semiconductor Category 

Operate Hours 
(in millions) 

Number of 
Removals 

Removals per 
Million Hours 

High-power transistors (P 5w) 

9.05 

184 

20.3 

Low-power transistors (P 5w) 

78.2 

586 

7.5 

High- current diodes ( 1 amp) 

66.4 

1,146 

17.2 

Low- current diodes ( 1 amp) 

66.4 

708 

10.7 


As indicated above, the malfunction rates used for semiconductors were 
listed in the source information as removal rates. Thus it is possible that 
these rates might be somewhat pessimistic, or high. 


Available information regarding the remaining parts listed in Table A-l is 
as follows: 


Part Type 

Operate 

Hours 

Number of 
Failures 
or Removals 

Removal or 
Malfunction 
Rate 

(xlO' 6 ) 

Source 


Accelerometer 

* 

* 

280.00 

Reference 1 
Table A-l 

of 

Bearing 

* 

* 

50.00 

Reference 1 
Table A-l 

of 

Chopper 

* 

* 

50.00 

Reference 1 
Table A-l 

of 

Clutch. Magnetic 

* 

* 

60.00 

Reference 1 
Table A-l 

of 

Gear Assembly 

* 

* 

90.00 

Reference 1 
Table A-l 

of 

Gyro 

* 

* 

490.00 

Reference 1 
Table A-l 

of 

Wave Guide, Fixed 

* 

* 

110.00 

Reference 1 
Table A-l 

of 

Wave Guide, Flexible 

* 

* 

284.50 

Reference 1 
Table A-l 

of 

Rotary Joints 

55,000 

7 

126.70 

Reference 4 
Table A-l 

of 

Sockets 

4,639.083 

2 

0.43 

Reference 3 
Table A-l 

of 


♦Unknown. Malfunction rates were determined by multiplying generic failure 
rate by a factor of 100. Both the generic rates and multiplying factors 
are given in Reference 1 of Table A-l. 


A -8 



r 


A 






APPENDIX B 

AN OPTIMUM STRATEGY FOR 
SINGLE -UNIT SEQUENTIAL TESTS 

1. Introduction 

In section 6. 6. 7.1 of the text It was concluded that revision of trouble- 
shooting strategies would not be worthwhile under present operational constraints, 
because the cost would far outweigh the benefits to be realized. This appendix 
describes a method for determining the optimum troubleshooting strategy. 

2. Problem 

When a symptom of a failure Is reported to a maintenance technician, he 
must repair the failed equipment by replacing a faulty line-replaceable unit 
(LRU). Either he removes a suspect LRU from the equipment and replaces It with 
an LRU from the spares supply, sending the removed LRU to the shop for repair, 
or he removes the suspect LRU and immediately takes it to the shop for checkout 
and repair, so that it can be reinstalled In the plane. In either case, the 
technician needs a method that tells him in which order LRUs should be removed 
to try to clear the symptom. Such a method now exists; it Is a branching strategy, 
in which the technician selects a starting point by referring to the symptom report 
and follows the procedure through successive decisions until the malfunction Is 
corrected. 

With each removal, check, and replacement, there can be associated a cost, 
equipment utilization, manpower, and related factors. The history of failures 
with the same symptom permits an estimate of the possibility that any particular 
LRU is the bad one. The assumption is made, for purposes of this analysis, that 
the failure is in a single LRU. It Is desired to order the actions of the main- 
tenance man so that the repair can be accomplished at the lowest possible cost. 

3. Solution 

For any given symptom and any LRU, there is a probability that the LRU is 
causing the failure. If there are n LRUs in the equipment, P^, Pg . . . P must 
be known, where P. = probability that LRU number 1 is bad. C. — the cost of 
removing, testing, and replacing the i 1 LRU — must also be known. 

It Is proposed that to minimize cost, the first LRU chosen for a check 

°i 

should be the one with the lowest ratio -s— ; the next LRU chosen should have the 

C i 1 

next lowest ratio -k— ; etc. 

*1 


Suppose there are n LRUs (b^, b g . . . b n ). Let one strategy be an ordering 
of the LRUs for checking: S^^ = b^, b ig , b^ • • • b ln (there are n.' strategies). 
Now, for each strategy S^, there correspond two strategies S ^ and S^' , defined 
as follows: 

S* (the inverse of S ± ) = b lg , b^, b ± g, b li; , b 15 . . . b ln 
S^ (the complementary part of S^) = b^* b^, tl 15 • • • b ln 

Let 

X^ = the expected cost of S^ 

* 

X*^ = the expected cost of S^^ 

X = the expected cost of in S^ 

(i.e., that part of X^ due to ) 


The costs of strategies can then be computed as follows : 
Suppose that in S^, 

C 11 ^ C 12 
P il P 12 


Then 


X^ = C 11 + (l - P^) C lg + X^' (see below for determination of X) 


That is, b^ is always checked, at cost b lg is checked a fraction (l - P^) 

of the time, at cost C^ g ; etc. 

X i* = C i2 + d - P i2> C il + X i 


Then 


* 

X 4 - X 


i " A i “ C i2 + f 1 “ P i2^ C il " C il _ ^ “ P ll) C i2 


~ ^ - Cj 0 + C. . — P. n C . t — C. . “ C. r, + P. . c. 


■ P il C i2 “ P i2 C il 


Prom the assumption that 


C il . C i2 , 


il 12 


B-4 


it follows that 


C il P i2 < C 12 P il 


S, > 0 


Thus it is shown that, for any given strategy after removal of the first 
two LRUs, the lower cost is achieved by choosing first the LRU with the lower 

^ ratio. The best strategy, then, will choose first the LRU with lowest 

ratio. The same argument is applied to subsequent choices. 

The expected cost of is given by 

S i = C il + (1 - P il> C i2 + (1 - P il + P i2> C i 3 ' • * * 


o — 

d " X P ik } C 1J + •