(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Study of capacitors for static inverters and converters final report"

A 



t,5--/S&9 ^ 



NASA CR54248 
R63SCW-31-5 



NASA 



STUDY OF CAPACITORS 
FOR STATIC INVERTERS AND CONVERTERS 

By J. F. Scoville 



PREPARED FOR 
NATIONAL AERONAUTICS AND SPACE ADMINISTRATION 



CONTRACT NAS3-2788 




GENERAL® ELECTRIC 



NOTICE 

This report was prepared as an aooount of Government-sponsored work. 
Neither the United States nor the national Aeronautics and Space 
Administration (RASA)) not any person acting on behalf of NASA: 

A) Makes any warranty or representation! expressed or implied, with 
respect to the accuracy j completeness, or usefulness of the 
information contained in this report, or that the use of any 
information, apparatus, method, or process disclosed in this report 
may not infringe privately-owned right; or 

B) Assumes any liabilities with respect to the use of, or for damages 
resulting from the use of any information, apparatus, method or 
process disclosed in this report. 

As used above, "person acting on behalf of NASA" includes any employee 
or contractor of NASA, or employee of such contractor, to the extent 
that such employee or contractor of NASA or employee of such contractor 
prepares, disseminates, or provides access to, any information pursuant 
to his employment or contract With NASA, or his employment with such 
contractor. 



Copies of this report can be obtained from: 



National Aeronautics and Space Administration 
Office of Scientific and Technical Information 
Washington, D.C., 20546 
Attention: AFSS-A 



T.I.S. R63SCW-31-5 



m0A 



FINAL REPORT 

STUDY OF CAPACITORS 
FOR STATIC INVERTERS AND CONVERTERS 

By J. F. Scoville 

PREPARED FOR 
NATIONAL AERONAUTICS AND SPACE ADMINISTRATION 

OCTOBER 30, 1964 

CONTRACT N AS3-2788 

TECHNICAL MANAGEMENT 
NASA-LEWIS RESEARCH CENTER 
CLEVELAND, OHIO 

AUXILIARY POWER GENERATION OFFICE 
FRANCIS GOURASH 



GENERAL ill ELECTRIC 



FOR USE OF G-E EMPLOYEES ONLY 

generalHelectric 



TECHNICAL INFORMATION SERIES 

Title Page 



AUTHOR 



J. F. Scoville 



SUBJECT CLASSIFICATION 

CAPACITOR 



NO. 

R63SCW-31-5 



DATE 

18 Oct. 



1964 



TITLE 



Study of Capacitor for Static Inverters and 
Converters . 



abstract Objectives of this study of capacitors for applica- 
ti on in aerospace static inverters and converters were to 
determine state-ofrthe-att capacitors, their AG charac- 
teristics, safe operating Voltages and temperature consis- 
tent with reliability and weight objectives. Analysis of 
operation and losses of commutating capacitors. Capaci- 
tance change and dissipation factors for polycarbonate, 
tnetallized paper and tantalum capacitors are presented 



G.E. CLASS 



1 



GOV. CLASS, 



None 



REPRODUCIBLE COPY FILED AT 

Library - Specialty Control 
Department 



NO. PAGES 
70 



CONCLUSIONS 

Polycarbonate film dielectric capacitors are considered 
state-of-the-art. Low dissipation factors and capaci- 
tance stability of polycarbonate capacitors were obtained 
in tests over an ambient temperature range from cycles. 
Appreciable size and weight reductions of commutating 
and load filter capacitors in aerospace static inverters 
and converters may be realized with the use of poly- 
carbonate capacitors. 



By cutting out this rectangle and folding on the center line, the above information can be fitted 

into a standard card file. 

For list of contents — drawings, photos, etc. and for distribution see next page (FN-610-2). 



information prepareo fq B National Aeronautics and Space Administration 
J. F. Scoville and L. G. Evelsizer 



TESTS MADE BY. 



COUNTERSIGNED. 



H. S. Sechrist // -c> J[&u^*^Cd\ v s P ec ^ alt y Control Department 



niveous Industrial Electronics t nmmoi< Waynesboro, Virginia 



FN4WM-A U-49) 



GENERAL ELECTRIC COMPANY 

TECHNICAL INFORMATION SERIES 

CONTENTS PAGE 



CONTENTS OF REPORT 

NO. PAGES TEXT 70 

no. charts See Page ii 

DRAWING NOS. 



photo nos. See Page ii 



For Normal TIS Distribution 
Specialty Control Department 

V. J. Wattenberger 

W. . Hansen 

D. L. Plette 

P. M. Tabor 

H. S. Sechrist 

H. W. Gayek 

D. F. Rogers 



nMIM C1.4M 



TABLE OF CONTENTS 



SUMMARY 




INTRODUCTION 


1.0 


Scope of Study 


1.1 


Discussion of Capacitor Functions 


1.2 


Application Considerations 


1.3 


Seleetloff of 'G^jfHitfitor Values for 




Industry Siar^ty 


2,0 


Capacitor Survey 


2.1 


Results of Survey 


2.2 


Selection of Capacitor Types and 




Values for Experimental Tests 


2.3 


Tubular Capacitor Physical 




Comparisons 


3.0 


Capacitor Experimental Tests 


3.1 


Capacitance and Dissipation Factor 




Test and Results 


3.2 


Comrautating Capacitor Test and 




Results 


3.3 


Life Test and Results 


4..0 


Conclusions 


5*0 


Recommendations 


Report Distribution List 


Appendix A 




Appendix 5 





Report No. R63SCW-31-5 
Page i 



Page No. 



I 

5 
8 



10 
1Q 
13' 

H 

17 

17 

29 

38 
43 
43 

45 
50 
59 



■■ "' : -:■*%*& y'T"W' 1 



Report No. 
Page ii 



R63SCW-31-5 



LIST OF FIGURES 



Figure 1 


Figure 2 


Figure 3 


Figure 4 


Figure 5 


Figure 6 


Figure 7 


Figure 8 


Figure 9 


Figure 10 


Figure 11 


Figure 12 


Figure 13 


Figure H 


Figure 15 


Figure 16 


Figure 17 


Figure 18 


Figure Al 


Table I 


Table Bl 


Table B2 


Table B3 


Table B4 


Table B5 


Table B6 


Table R7 


Table B8 



A.C. Filter Configuration 

Elementary Diagram of Impedance Bridge 

Calorimeter and Test Equipment 

Teat Specimen Mounting in Calorimeter 

Average. Percent Capacitance Versus Frequency in 25*C Ambient 

A-N^g* *er%enVM*sipation Factor Versus Frequency in 

2^; ; Aflfci«nk;;: _•.;.'_ 

Pereent Capacitance Versus Frequency and Temperature 

Percent Dissipation Factor Versus Frequency and Temperature 

Percent Dissipation Factor Versus Extended Frequency Range 

in 25°C Ambient 

Average Dissipation Factor for Tantalum Capacitors 

Dissipation Factor for 2.5 MFD Metallized Polycarbonate 

Capacitor 

Commutating Capacitor Waveforms 

Commutating Capacitor Waveform 

Commutating Capacitor Waveform 

RMS Current and Voltage- for Commutation Pulse 

Mounting of Life Test Capacitors 

Life Test Monitoring Equipment 

Elementary for Capacitor Life test 

Peak Current and Voltage Waveforms 

Tabulation of Tubular Capacitor Sizes and Weights 

25°C Test Data for Metallized Polycarbonate Capacitors 

25°C Test Data for Polycarbonate/Foil Capacitors 

25°C Test Data for Metallized Paper Capacitors 

Capacitor Test Data Versus Temperature (Metallized 

Polycarbonate) 

Capacitor Test Data Versus Temperature (Polycarbonate/ 

Foil) 

Capacitor Test Data Versus Temperature (Metallized Paper) 

Capacitor Test Data Ver sus Temperature (Tantalum) 

Capacitor Test Data Before and After Life Test 



Report No. R63SCW-31-5 
Page i 



SUMMARY 



This is the final report for the "Study of Capacitors for Static 
Inverters and Converters". The objectives of this study were: (l) To 
determine state-of-the-art capacitors for use in aerospace static 
inverters and converters: and (2) Determine A.C. characteristics of 
these capacitors to facilitate selection of ratings and limits for such 
factors as safe-operating voltage, temperature rise, size, weight and 
reliability. 

Investigations were limited to those capacitors suitable for use in 
static inverters and converters operating in a space environment with 
these ratings: 

Input: 25 to 105 volts, D.C, 

Output: 0.1 to 10.0 kilowatts, 115/200 volts, 3 phase, 4-00 cps 

An industry survey, conducted in the early part of this study, indicated 
that state-of-the-art capacitors are polycarbonate for film capacitors 
and tantalum for electrolytic capacitors. Metallized polycarbonate, 
polycarbonate/foil, and metallized paper capacitors were tested to 
determine capacitance and dissipation factor variations versus temperature 
and frequency. Measurements were taken in -55°C to +85°C ambients with 
excitation frequencies from 0.4. to 10,0 kilocycles and a few measurements 
were taken with excitation frequencies extending to 80 kilocycles in a 
25°C ambient. 

In general, polycarbonate capacitors exhibit capacitance and dissipation 
factor value variations of less than half that for metallized paper 
capacitors over the temperature and frequency range of the tests. 

The small dissipation factors (i.e. power losses) of polycarbonate 
capacitors facilitates smaller size and weight capacitors with lower hot 
spot temperatures than metallized paper capacitors in some A.C. applica- 
tions. 



Report No. R63SCW-31-5 
Page 2 



INTRODUCTION 



Lack of adequate alternating current data and characteristics of paper 
and film dielectric capacitors contribute to the difficulty of properly 
applying capacitors to aerospace static inverters and converters. 
Improper application of capacitors in these equipments could result in 
appreciable-penalties in weight and reliability factors, both of which 
are at a premium in equipment operating in space. 

A need for the "Study of Capacitors for Static Inverters and Converters" 
was influenced by the stringent requirements imposed on capacitors by 
the operating nature of the equipment in a space environment and by the 
general lack of capacitor A.C. characteristics and data. 

In static inverters, appreciable heat generation within capacitors is 
caused by alternating currents from ripple, commutation and output 
voltages. Heat transfer is generally limited to conduction across the 
capaeitor mounting surface to radiator systems on space vehicles. 

The purpose of this study is to obtain the alternating current charac- 
teristics and data of statar-of-the-art capacitors, for application in 
aerospace static inverters and converters. These characteristics will 
facilitate proper selection of ratings and limits for such factors as 
safe operating voltage, temperature rise, size, weight and reliability. 

State-of-the-art metallized polycarbonate and extended foil polycarbonate 
capacitors from several manufacturers were evaluated for capacitance and 
dissipation factor variation with temperature and frequency* Evaluation 
of A.C. characteristics for a few metallized paper capacitors was included 
for reference purposes. Similarly, a very limited quantity of state-of- 
the-art tantalum capacitors was evaluated, because of availability 
of adequate A.C. characteristics and data for this type of capacitor. 

Capacitor alternating current characteristics were determined by 
impedance bridge measurements. Some of these capacitors were also tested 
in a calorimeter to measure power losses. Results from these calorimeter 
tests were utilized as correction factors for the capacitor dissipation 
factor obtained by the impedance bridge. This method of testing, although 
more time consuming, is considered more accurate for determining dissipa- 
tion factors that were often below the resolution of most commercial 
capacitor bridges. 



Report No. R63SCW-31-5 
Page 3 



It is hoped that the commutating capacitor analyses of operation and 
losses contained in this report will provide capacitor manufacturers 
with a better understanding of this type of application. Capacitance 
and dissipation factor test data presented in this report is expected to 
facilitate the proper capacitor selection by equipment designers for 
A.C. application in aerospace inverter s and converters. 



Report No. R63SCW-31-5 
Page U 



1.0 Scope of Study 



The lack of adequate alternating current characteristics and 
data has caused considerable difficulty for equipment designers 
in selection of capacitors for application in static inverters 
and converters. 

Objectives of this study were: (l) To determine state-of-the- 
art capacitors for aerospace inverter applications; and (2) to 
determine state-of-the-art capacitor A.C. characteristics to 
facilitate selection for the inverter applications. 

In accomplishing these objectives, this study included a brief 
.analysis of commutation capacitors, conduction of an industry 
survey for state-of-the-art capacitors and testing of selected 
capacitor types. Testing of capacitors was limited to capacitance 
and dissipation factor variations with temperature and frequency 
and a life test. 

This study was not primarily concerned with capacitor D.C. charac- 
teristics. However, one supplier of polycarbonate film lists typical 
values of insulation resistance in 25°C ambient of 250,000 megohm- 
microfarads and 20,000 megohm-microfarads in 125°C ambient. This 
manufacturer also lists a typical dielectric absorption of 0.2$ for 
polycarbonate as tested in accordance with MIL-C-19978B . Specific 
capacitor application requirements such as vibration, shock, 
radiation affects and hermetic seals, were not considered within 
the scope of this study, but these requirements were used as guidelines 
in the industry survey. 

1.1 Discussion of Capacitor Functions 

Capacitors in aerospace static inverters and converters provide these 
basic functions: (l) D.C. Filter; (2) A.C. Filter; and (3) 
Commutation. Generally, static inverters that employ silicon controlled 
rectifiers as static switches require all three capacitor functions and 
inverters using transistors as static switches require only the filter 
functions . 

Commutating capacitors in conjunction with commutating reactors 
are used for electrical energy storage to effect commutation of load 
current from one silicon controlled rectifier to another. The 
stored electrical energy in these commutating components furnish 
sufficient back biasing voltage to a rectifier that has been conduct- 
ing to turn it off when a second rectifier is turned on. The analysis 
contained in Appendix A explains this functional operation in more 
detail. 



Report No. R63SCW-31-5 
Page 5 

Alternating current filter capacitors are normally required with 
reactors in the output circuits of static inverters to shape 
voltage and current waveforms from quasi square-wave, caused by 
the alternate static switching, to sinusoidal waveforms. 
Generally, these inductive-capacitive (L-C) filters employ 
capacitors connected in series and in parallel with the load as 
shown in Figure 1. 

Direct current filter capacitors are usually employed in input 
circuits of static inverters and output circuits of converters. 
Their main purpose is storage of electrical energy and release 
thereof for reducing or regulating source voltage variations 
from pulse loading. 

1.2 Application Considerations 

Environments 

Potential environments that capacitors may encounter in aerospace 
static inverter and converter applications are major considera- 
tions in the selection of capacitor materials. Environmental 
conditions, assumed to be representative for a variety of equip- 
ment design applications, selected for a state-of-the-art survey 
in this study are: 

a) Heat Sink Ambient Temperature Range: -55°C to +85°C. 

b) Capacitor heat transfer by conduction to heat sink. 

c) Shocks of 35 g's in half sine wave shocks for 0.008 seconds. 

d) Vibration: Sinusoidal 

Frequency (CPS) Force or Displacement 

5-20 0.3 inches double amplitude 
20-100 5 g's 

100-500 10 g's 

500-2000 15 g's 

e) Radiation: _ 

5 x 10" i<J NVT Fast Neutrons/cm 

5 x 10" 7 RADS (carbon) Gamma Particles 

f) Hermetically sealed construction to protect capacitor from 
effects of sublimation from temperature-pressure conditions. 



Report No. R63SCW-31-5 
Page 6 



tt 



L 
-TYWV 




^g 



1 

L 

A 
D 

J 



'V 



FIGURE I 



A.C. FILTER CONFIGURATION 



C s - Series Capacitor 
Cp - Parallel Capacitor 



Report No. R63SCW-31-5 
Page 7 

Use of these environment considerations in the survey was 
utilized to facilitate the usefulness of this study to equip- 
ment designers of aerospace static inverters and converters. 

Power Source 

Types of power sources considered for the inverters and 
©(wverters.are: a) Batteries, b) Solar cells, c) Thermionic 
converters, d) Fuel cells, and e) Rotating D.C. generators. 

Power source characteristics, particularly voltage regulation, 
ripple and transient voltages are important considerations in 
selection of capacitors for equipment designs. 

Source voltage range used is 25 to 105 volts D.C. However, with 
present and near future power sources it appears that source 
voltage ranges from 25 to 35 volts, 50 to 65 volts and 90 to 105 
volts offer a representative coverage within the range from 25 
to 105 volts. These three (3) discrete source voltage ranges 
were used for the capacitor survey. 

A second power source consideration is ripple. A peak-to-peak 
ripple of plus and minus 10 per cent of the maximum steady state 
voltage was chosen as being representative for various applica- 
tions of D.C. input filter capacitors. Maximum frequency of the 
ripple voltage was considered to be 25 kilocycles and is representa- 
tive of other unfiltered static inverter equipment operating from 
the same power source. 

The third power source consideration is voltage transients. 
Removal of loads near the end of a transmission line can cause 
transient overvoltage conditions. In this study, transient over- 
voltage of 150 per cent for durations up to 100 milliseconds are 
considered representative for some aerospace static inverter appli- 
cations of D.C. filter capacitors. 

Load 

Load characteristics, such as transient overvoltage from load 
removal and overcurrents are also important considerations in 
the selection of A.C. filter capacitors for equipment designs. 
Transient overvoltage of 125 per cent of rated voltage for a 
duration of 5 cycles of the rated 400 cycle base was chosen to 
be representative of equipment applications. Overcurrents of 



Report No. R63SCW-31-5 
Page 8 

twice-rated load current for 5 seconds was also considered 
representative of equipment applications. However, output 
indue tive-capacitiv© (L-C) filter configurations and power 
factor of the load during overourrent conditions have a 
relationship to the peak currants and voltages of the capacitors. 
Two (2) rated rms voltage values were selected for the capacitor 
eurvey. These are 135 and 270 volts rms, 420 cycles. Dielectric 
▼oltage rating for these capacitors were chosen to be 600 and 
1000 volts respectively to accommodate peak tmnsient voltage 
condition*. The 12.0 cycle, rating was chosen to provide adequate 
margin from overheating during inverter operation at 4.00 + 20 
cps . ~ 

Three (3) discrete output ranges within the 0.1 to 10.0 kilowatt 
output power range for the static inverters considered were: 
(1) 0.1 to 0,5 kilowatts? (2) 0.5 to 2.0 kilowatts; and (3) 2.0 
to 10.0 kilowatts. 

1.3 Selection of Capaoitor Values for an Industry Survey 

Commutatlng Capacitors 

Analysis of the operation of commutatlng capacitors and design 
experience with static inverters has shown that a relationship 
exists between the value of capacitor and characteristics of 
the static switching device. The characteristics of the switching 
device that enter this relationship to capacitor value are magni- 
tude of load current flowing through the switching device at the 
end of each half cycle (i.e., start of commutation) and device 
turn-off time. 

Four (4.) silicon controlled rectifiers, with current ratings that 
are capable of handling the inverter power output range in this 
study, were chosen for sizing the oommutating capacitors. Turn- 
off times for these silicon controlled rectifiers are either vendor 
guaranteed or test selected by vendor. Values of current flowing 
through the silicon controlled rectifiers at the end of the half 
cycle are based on safe thermal limits for these devices for 180 
degree conduction angles. 



are: 



The design equations used for sizing the commutatlng capacitors 

C = 5.22 It -. j I = I.37 I n E 

o off P o max. 

^in. E min. 



Report No. R63SCW-31-5 
Page 9 

Where I is current flowing through the silicon controlled 
rectifier at the end of the half cycle, t of f is turn-off 
time for the rectifier, E and E^^ are maximum and 
minimum steady state source voltages respectively and I p 
is peak amperes capacitor current. 

These equations are derived in Appendix A of this report. 

Commutating capacitors, selected from the analysis, shown in 
Appendix A, for the survey were: 

Capacitance (mfd) D.C. Voltages 
5 35, 65 and 105 

15 35, 65 and 105 

50 35, 65 and 105 

These capacitance values, when used singly or in multiples, 
are capable of covering the inverter output power range for 
this study. 

A.C. Filter Capacitors 

Inverter output voltage is 115/200 volts, 4.00 cycles per second. 
Filter capacitors connected in parallel with the load, as shown 
in Figure 1, have voltage ratings consistent with the load 
voltage. Voltage ratings of filter capacitors connected in series 
with the load or reactors are dependent on inverter waveforms, 
filter design and load currents. 

The following capacitors were selected for the survey: 

Capacitance (mfd) RMS Voltage 

1-3 135 and 270 

8-10 135 and 270 

50-60 135 and 270 

These capacitance values, used singly or in multiples will 
accommodate many equipment designs within the output power 
range of this study. 

D.C. Filter Capacitors 

Discrete source voltage levels, establish voltage ratings of the 
input filter capacitors. Two (2) capacitance ranges were 
selected for the survey: 



Report No. R63SCW-31-5 
Page 10 

Capacitance (ufd) Voltages 

1000-1500 35,65 and 105 

10,000-15,000 35,65 and 105 

These capacitance values, or multiples thereof, will accommodate 
many equipment designs within the inverter and converter ratings 
for this study. 

2.0 Capacitor Survey 

Filter and commutating capacitor specifications, similar to the 
one contained in Appendix A for commutation capacitors , were 
submitted to a large number of capacitor manufacturers as a part 
of the industry survey. Requests to these manufacturers to 
furnish technical and cost proposals to these specifications 
resulted in formal proposals from six (6) manufacturers and 
incomplete proposals from five (5) additional manufacturers. 

From the responses to the survey and visitations to five (5) 
capacitor manufacturers, it was apparent that polycarbonate 
capacitors were considered state-of-the-art for film capacitors 
and tantalum for electrolytics . 



2.1 Results of Survey 



Commutating Capacitors 

The following four (4) types of capacitors were recommended by 
capacitor manufacturers for the commutating capacitor applications: 

A) Metallized polycarbonate film 

B) Polycarbonate film and foil 

C) Paper-Mylar* and foil 

D) Paper and foil 

The evaluation criteria variations encountered in the survey 
were: 



^Trademark of the E. I. Dupont Company 



Report No. R63SCW-31-5 
Page 11 

1) Volume to capacitance ratio of: 

0.22 to 1.85 in 3 /ufd for the 35 and 65 Volt ratings 
1.05 to 4.8 in 3 / ufd for the 105 Volt rating 

2) Weight to capacitance ratio of: 

0.015 to 0.18 pounds/ufd for the 35 and 65 Volt ratings 
0.058 to 0.34 pounds/ufd for the 105 Volt rating 

3) Volume to energy storage ratio of: 

44.5 to 411 in 3 / joule for the 35 and 65 Volt ratings 
85 to 374 in 3 /joule for the 105 Volt rating 

A. C. Filter Capacitors 

Capacitor manufacturers recommended the following four (4) types 
of capacitors for the A.C. filter capacitor applications: 

A) Metallized polycarbonate film 

B) Polycarbonate film and foil 

C) Paper-Mylar and foil 

D) Paper and foil 

Evaluation criteria variations encountered were: 

1) Volume to capacitance ratio: 

0.21 to 5.35 in 3 /ufd for the 135 V rms ratings 
0.5 to 7.65 in 3 /ufd for the 270 V rms rating 

2) Weight to capacitance ratio: 

0.018 to 0.33 pounds/ufd for the 135 V rms ratings 
0.035 to 0.535 pounds/ufd for the 270 V rms rating 

3) Volume to energy storage ratio: 

10.3 to 29.6 in 3 / joule for the 135 V rms ratings 
6.5 to 15.6 in 3 /joule for the 270 V rms rating 



Report No. R63SCW-31-A 
Page 12 



D.C. Filter Capacitance 



Recommendations for the D.C. filter capacitor applications 
included three (3) types: 

A) Sintered tantalum 

B) Tantalum foil 

C) Polycarbonate film 

Reservations concerning the specified peak to peak ripple 
voltage of 10 per cent of the D.C. voltage ratings accompanied 
many of the manufacturer ' s recommendations for usage of tantalum 
electrolytic capacitors. 

Slightly different evaluation criteria are used for the D.C. 
filter capacitors. The volume and weight to volt-capacitance 
ratios were favored because the volume to capacitance for 
electrolytics varies approximately with voltage. Film capacitors 
have broad discrete steps of volume and weight to volt-capacitance 
ratio. 

The volume to voltage capacitance ratios determined from the 
survey for these types of capacitors were: 

Type Cubic Inches/Volt-Microfarad 

125°C, Tantalum porous anode 3.0-4.0 x 10" 5 
125°C, Tantalum foil 11.0-13.0 x 1Q~ 5 

Polycarbonate film 100-185 x 10" 5 

Only 125°C tantalum capacitors were considered in the survey to 
facilitate temperature derating for reliability purposes. 

The effective series resistance (ESR) , in 25°C ambient and with 
120 cycle/second voltage, is within the same order of magnitude 
(i.e., 1 to 10 ohms) for both types of tantalum capacitors. The 
ESR of polycarbonate film capacitors is at least one order of 
magnitude smaller than the tantalum electrolytics. 

Weight to voltage-capacitance ratio for these three types of 
capacitors are: 



Report No. R63SCW-31-4 
Page 13 

Type Pounds/Volt-Microfarad 

Sintered Tantalum 2.5-10.0 x 10~£ 

Tantalum Foil 5.0-16.8 x 10"? 

Polycarbonate Film 87.5-265 x 10"° 

Volume tp energy storage ratio variations are: 

--- Tgpe - Cubic Inches/joule 

Sintered Tantalum 0.8 to 1.8 

Tantalum Foil 0.75 to 4.06 

Polycarbonate Film 44.5 to 120.0 

2.2 Selection of Capacitors for Experimental Testing 

Capacitor selection for experimental tests was based on the 
following considerations: 

a) Low volume and weight to capacitance ratios 

b) Low dissipation or power factor 

c) Adequate temperature capability to enable temperature 
derating for reliability purposes 

Although the capacitor survey was conducted with specifications 
for several ranges of capacitances and voltages, economies could 
be realized by selecting commutating and A.C. filter capacitance 
values of 1 to 5 microfarads for the experimental tests. Packaging 
of larger capacitance values will be influenced by the power 
factor of the capacitor, frequency, voltage waveform, materials 
and physical environmental requirements. 

Commutating 'and A.C. Filter - 

Metallized polycarbonate film capacitors having capacitance 
values of 1, 2, 2.5 «ad 3 microfarads with D.C. voltage ratings 
of 200, 300 and 400 volts were selected from four (4) manufac- 
turers. 

Polycarbonate film and foil capacitors having capacitance values 
from 1, 2, 3 and 5 microfarads with D.C. voltage ratings of 150, 
200, and 400 volts were selected from four (4) manufacturers. 



"" ■■'■ ^'^i^-^'^^ " y ^K*.?Vi.t.?J 



Report No. R63SCW-31-5 
Page U 

Metallized paper capacitors with capacitance values of 2 to 3 
microfarads and D.C. voltage ratings of 200 and 400 volts were 
selected from two (2) manufacturers. 

Other film type capacitors that were not selected for experi- 
mental test evaluation are: 

A) Teflon* —These capacitors have adequate temperature capabi- 
lities and low dissipation factors but the size, weight and 
cost penalti«s compared with polycarbonate or paper capaci- 
tors offset any advantages. 

B) Mylar — Capacitors of this type are comparable to the size 
and weight of polycarbonate capacitors but have appreciable 
power factors between 85°C and 125°C 

C) Polystyrene — These capacitors have characteristics that are 
comparable or better than polycarbonate capacitors except 
for the temperature limitation of 85°C. 

D.C. Filter 

Since tantalum electrolytic capacitors are considered state-of- 
the-art and A.C. characteristics for these capacitors are generally 
known, only a very limited quantity of sintered and foil types 
were obtained for reference evaluation. 

Sintered tantalum capacitors selected have ratings of 22 micro- 
farads, 100 volts D.C, 125°C. 

Tantalum foil electrolytic capacitors were selected with ratings 
of 36 microfarads, 150 volts D.C, 125°C 

Voltage ratings greater than 100 volts D.C. for sintered tantalum 
electrolytios are available by series element arrangements with, 
associated penalties in size and weight. 

Aluminum electrolytics were not considered because of the 6$°C and 
85°G temperature rating limitations. 

2.3 Tubular Film Capacitor Comparisons 

A comparison of tubular capacitor sizes and weights by type is 
presented in Table I. In the 400 VDC ratings^ the ratio of the 



♦Trademark of the E. I. DuPont Company 



-'*^J^ : -K.,-;r^.:. L ,;- 



Report No. R63SCW_3i_5 
Page 15 



Tabulation of Tubular Film Capacitor Sizes and Weights 



Capacitor 








Identification 


Rating 




V ]/Cap 


No. 


400 VDC 200 VDC 
1 uF 


Type 
MPC 


(in/uF) 


1A-1E 


0.91 


2A-2E 


2 uF 


MFC 


0.94 


3A-3E 


2.5 uF 


MPC 


1.03 


5A-5E 


2 uF 


MPC 


.84 



Average W g t/C a p 

Vol./uF/Tvpe (Xj_£) 



0.93 



.063 
.062 
.059 
.053 





7A-7E 
8A-8E 


2 uF 
2 uF 


MP 
MP 


.89 
.71 


0.80 


.064 
.052 


9A-9E 


1 uF 


PCF 


2.06 


2.06 


.145 





12A-12E 
13A-13E 
14A-14E 
15A-15E 



3 uF 


MPC 


0.35 


3 uF 


MPC 


0.28 


3 uF 


MPC 


0.56* 


3 uF 


MPC 


0.24 



0.29 



.023 
.018 
.036 
.017 





4A-4E 
16A-16E 


2 uF 

3 uF 


MP 
MP 


0.22 
0.21 


0.21 


.018 
.016 




10A-10E 


2 uF 


PCF 


0.93 


0.93. 


.069 



MPC - Metallized Polycarbonate 
Key: MP - Metallized Paper 

PCF - Polycarbonate/Foil 



"Omitted in the Average Voi/uF/Type 
figure 



TABLE I 



Report No. R63SCW-31-4 
Page 16 

average volume/capacitance of metallized polycarbonate to 
metallized paper capacitors is 116 percent and similarly the 
ratio for the 200 VDC ratings is 138 percent. This 22 percent 
increase in volume/capacitance ratio from the 400 VDC to the 
200 VDC ratings may be attributed to a minimum polycarbonate 
film thickness availability limitation. 

The volume/capacitance ratio between polycarbonate/foil and 
metallized paper capacitors is 250 percent for the 400 VDC 
ratings and 450 percent for the 200 VDC ratings. From these 
large ratios, it appears that the foil thickness may approach 
the polycarbonate film thickness. 

The pounds/ capacitance for metallized polycarbonate is approxi- 
mately the same as for metallized paper capacitors. Larger 
physical size of the polycarbonate oapacitors, compared with 
metallized paper, account for the larger pounds/capacitance 
figure. 



Report No. R63SCW-31-5 
Page 17 

3.0 Capacitor Experimental Tests and Results 

Three (3) types of capacitor tests were conducted during this 
study: (l) Capacitance and dissipation factor; (2) Commutating • 
capacitor; and (3) Life. 

Capacitance and dissipation factor test data were obtained as 
a function of frequency and temperature for metallized polycar- 
bonate, metallized paper, polycarbonate/foil and tantalum 
electrolytic capacitors. Commutating capacitor tests were 
conducted with a metallized polycarbonate capacitor and an analysis 
of the capacitor power losses was made utilizing data obtained 
from the capacitor and dissipation factor tests. A life test of 
the three types of capacitors was conducted for 1000 hours in an 
85°C ambient and 370 hours in 125°C ambient. 

3.1 Capacitance and Diasipation Factor Tests and Results 

Capacitance and dissipation factor values for polycarbonate and 
paper capacitors were determined for a temperature range from 
-55 C to +85°C and frequency range from 0.4. to 10 kilocycles with 
sinusoidal waveforms. 

Teats 

Capacitance values were measured with an impedance bridge 
constructed for these tests. This bridge was constructed with 
components and interconnections to facilitate a minimum of stray 
and unknown capacitances and inductances over the frequency range. 
An elementary diagram of this type bridge is shown in Figure 2. 
Capacitance values measured with this bridge compared favorably 
with values obtained with a General Radio Model 716-C and Sprague 
Model 1W2 capacitance bridges. 

Capacitor dissipation factors measured with the impedance bridge 
were considered comparative data that required adjustment with 
data obtained from calorimeter heat loss measurements. Adjustment 
of these dissipation factor data was necessary because unknown 
and stray capacitances of the bridge were still appreciable in 
the frequency range of interest. 



Report No. R63SCW-31-5 
Page 18 



Elementary Diagram of Impedance Bridge 




Rl = 99.7 OHMS; CS - .01 MFD., GEN. RADIO STD. K09L 

CX - Capacitor under test 

RC - Adjustable resistance (switch, fixed resistors and a potentiometer) 

RD - Adjustable resistance (switch, fixed resistor and a potentiometer) 

AMP. - Battery operated, single stage, transistor amplifier 

M - Meter, Harmonic Wave Analyzer - used as null detector 

Dissipation Factor = Rex - RD = Rd(gs)oJ 

x cx x cs 

Where R is the effective resistance of ca[ .citor under test. 



ex 



Capacitances Rl - X cx 



RC X 



CX = CS (RC) 
Rl 



cs 



Figure 2 



Report No. R63SCW-31-5 
Page 19 

Use of the calorimeter was made in measuring the rate of heat 
being generated from a capacitor immersed in a fluid and observing 
the rate of temperature rise. Calibration of the calorimeter is 
accomplished by mounting a known value of resistance to the 
capacitor surface within the calorimeter and measure the direct 
current flowing in the resistor while observing the rate of 
temperature rise of the calorimeter fluid. Caution was exercised 
to measure the rate of temperature rise after the thermal inertia 
had 1>cen evereome. , 

Calculation of the heat input to the calorimeter may be obtained 
from calibration with known heat input rates 

I 2 R = Watts = °C/minute 

After calibrating the calorimeter, the capacitor is energized 
from a variable frequency power supply with sinusoidal voltage. 
Capacitor dissipation factor (D.F.) is obtained from measurements 
of the RMS voltage across the capacitor and rate of temperature 
rise of the calorimeter fluid and by use of the following calcu- 
lations: 

D.F. * Watts ( A°C/At/calibratior. A_°C 

At 



VA 



E 2 6iC 



Shown diagramatically 




E = Capacitor rms voltage 
C = Capacitance* 



Note! Since the angle alpha (oc) is generally 
small, the dissipation factor is 
essentially equal to the capacitor 
power factor. 



Watts 



Care was exercised to minimize the calorimeter heat loss rate 
by maintaining the calorimeter fluid temperature within + 2°C of 
the calorimeter external ambient (24- - 26°C within the enclosure). 
Maintaining the calorimeter fluid temperature relation to the 
ambient was accomplished by replacement of the fluid between tests. 



Report No. R63SCW-31-5 
Page 20 

The calorimeter fluid quantity was kept constant during the tests 
by utilizing weight measurements when replenishing the fluid. 

A picture of the test equipment used is shown in Figure 3. Figure 
U shows the mounting of a capacitor, calibrating resistors and 
thermometer to the top of a commercially available vacuum bottle, 
which was used as a calorimeter. The calorimeter shown in Figure 3 
is in an enclosure to prevent room air currents from striking the 
external surfaces of- the calorimeter that could alter the heat 
loss rate from the calorimeter. 

Testing of the tantalum electrolytic capacitors was limited to 
bridge measurements in ambient temperatures from -55°C to +85°C 
and frequencies from 120 to 1200 cps. 



Report No. R63SCW-31-5 
Page 23 



Results 



Capacitance and dissipation factor test data in 25°C ambient for 
twenty (20) metallized paper, forty-five (45) metallized 
polycarbonate and twenty-four (24) polycarbonate/foil capacitors 
are contained in Tables Bl through B3 in Appendix B of this report. 

The average of the percent capacitance change versus frequency, data 
from Tables Bl through B3 is shown in 'Figure 5. At "10 kilocycles 
the average capacitance variation of metallized paper is 1.68 
percent compared to a 2.95 percent average for metallized 
polycarbonate and a 3.85 percent average for polycarbonate/foil 
capacitors. The lower capacitance variation with frequency of 
metallized paper capacitors in 25°C ambient is attributed to the 
longer established quality control of materials and techniques 
in comparison to the relatively new polycarbonate capacitors. 

The average of the percent dissipation factor versus frequency 
data from Tables Bl through B3 is shown in Figure 6. In general, 
the average of the dissipation factor for polycarbonate/foil 
capacitors is one-half the value of metallized polycarbonate and 
one quarter the value of metallized paper capacitors. 

Capacitance and dissipation factor test data versus temperature 
and frequency in Tables B4 through B6_are shown in Figures 7 and 
8j respectively. Capacitance change _"vith temperature and 
frequency shown in Figure 7 for metallized paper capacitors is 
12 percent, for polycarbonate/foil, it is 5.5 percent and for 
metallized polycarbonate^ it is 3 percent. 

Dissipation factor variation with temperature and frequency varies 
from 0.03 to 0.46 percent for polycarbonate/foil, 0.55 to 1.23 
percent for metallized polycarbonate and 0.16 to 1.93 percent for 
metallized paper capacitors as shown in Figure 8. 

Some test data as indicated are excluded from the presentation 
in Figure 8 because these data exhibited excessive variation and 
were not considered representative. 

Dissipation factor test data were obtained from two (2) metallized 
paper, £rxr (4) polycarbonate/foil and three (3) metallized 
polycarbonate capacitors over an extended frequency range in 25°C 
ambient. These data, obtained with calorimeter measurements are 
shown in Figure 9- Generally, the dissipation factor for all three 
(3) types of capacitors increase 100 percent between 10 and 50 
kilocycles. 



'«™^w>w^f«MS!!(»j»« w «jp w », 



W^^K-VMS,-**,,^^. Xji&wms*^...,,;-^ 



C 
0) 
•fl 

E 

< 

u 

o 

in 



' >*'■'■ 
y 
c 

s 
cr 
v 

h 
«> 


03 

u 

> 

be 

d 

,d 
U 

u 

c 

rt 
+■> 

•H 

u 
ttt 
a 
rt 
U 

+j 

d 

0) 

u 
u 
<u 
(1 

4> 

h 

> 




V. I 




A A A A A A A A 



*!"*':*"■.' .4.-' :• j** 



u. 



I 

U 

Eh 
t3 



1 


a 


1 


a) 


1 


3 




a* 




<D 




k 




in 


1 


W 


1 







CO 




h 


; 


a) 




> 




CD 


>- ; 


t»D 


z ; 




•* 




B. : 




r 


ri 


O i 

u ; 


o 


a . 


CO 


u 

3 


o 


u 


-p 


_j 


•H 


1 ' 


o 




a 


z r 


{X 




cd 




o 


1 


+» 




a 




CD 




O 


j 


f-i 


. 


(1) 


! 


3-, 




-^■^■^:r_. l -^*^j; i v;...-v ! ..j ftf , 




S 



S 
I 



u 

P>4 



o 



^5? C 55 so -* <T 

fe <M rl r4 c? r? 



co v© -=t 
• • • 














r : 










:-r. 


'. ' ! ' 




TTTT 






ii : T 


-T!T 








■:i;| : : ; i 






ii. 


- : ■ : 


;■■:: 


,IIX 


::.:'■ -.: - r ■ 








-i 
















-V- 


















: :)•:■ 














■ i 










i 






■ l~^~^ 


~-ij. 






T*f4i 










'S^r 






:. 1: 


S- 










\ 


L dft 


! 

1 ' 




Z ' : 


,- 




; 




^ 














\ : '^\ 






•■ 1 


3 








.__ 




i: 5)S 




i : 








; ; : J . 


£> 






N 












■;-.| : 












\ 








i 

ro 
I 










p • 








!::■!..■< 










-iS 






^ 


: 


<_ 




: N- : i 




t^t h 








■ r 


















:-~l 






:-.'.):': 






N 


9' 




.. > j 




i ^ -. 








: : !.. 










•I-;: 












ii. '■': 














"■.■!■' 1 

V ;■ -t 1 




: * : 

K 














L: 


*i 








^Si 










:.". 












V' 








] . 






\ 




'«7 : - 






-: : 












»ii 










i ' L 




'■ 




' j 


. I. . 

1 


::* 
















.... . 












1, 










\ 








" 


1 

L - 


■ 




•S 


: :: 






.::■ 


-V 


— • 


. . i... 


n^ 










:..S 


% 


" ,":- 


^i 


Al\i 






00 

-- -m N •— 

O <U 

a Qo ' 


- 
■ 


. 
















"■.": 


■ ■ -i 

-i . 

... .j. 




■^ 


\ 








\ 


\ ■ • 


';' - ;" 




c 

• ft 




: ■ 














'.:. :". 


V ; 


.:-: 


CT!" 


'-7: 


:;!■ 






; 


:\ 










s 


v ; :. 


"^ \ 




< 

U 




« 


fa 

■■ 1 

■ j - 


■ 




:- 






.....:. 






: : !' 






ttH« 


^ 


■sr 




■ : ] : 


i_i-i-i 


j.. ■ , 


.,. v f 


3 


x< 




■ ■ t "; 


; v' 


i - ; ■ 


1 


(VJ 


— . _ 


■1 




-'■■;. 


- I-:. 




*, .".: .•; 


:.:xr; 


•E~ l 






! ■.".... 


:-■!;. 




-:■ 








Ki~ 


■ -^ 








-; : .Tz 


.•__;-;; 


V m 


C 














i: 






:.. i...^^ 


U"' : T 








. ": r 


7 : "-" 


V^ 


BT. -- 




":::: 




&*T 




^srfw-l 


>, 


h ■■ • 




1 
















l'"l 






Ivl 


: ri- 


]-:!:■■ 




l J " 'i 








: T: ? 


.:"■;*- 




r—r- 


~ : t 


Ihj. 


S'l 




i V 


u 


















:' i - ; 










■ \y\ 




: i- 










rH" 










iL "!' 


i n/*v 




c 




■■:" 












:\i'~ 










'■'■; 












.; ■ „ : 














4) 
3 


, 


-— 






:■■ 


















!■•* 


'i:. 1 -;,!- 












-■■■ 










Tit'. 


JTV : ) , 


t' "T~ T v^ 


(T 


1 




'.':.<]'.'■ 




















r J:i 


J .i:-- 


:vf4:i 


ilz:. 


i.; .. 










t ' : ; 


i'if 




- i: : 




p.mx,k w 


0> 


-— t — 

..._(.,;;■ 






.'.1 '■: 


-j'..-- 




-::.•• 


K:.: 


■:;- 


}".':. 




i-.ir. 




rlL: 


i:; ;:;. 






r!:' 


'.•rr 


-;*'.. 


!;•■" 


l-: r - : 










J Ji - ; -; 


feT 7 ^ 


i; ; v';* ^~ 


fa 






f.-.i 




-■;. 




:V. : . 


iiii 




..'- ; 


; ■ r]_ 




Itii- 


ft""" 




i\h 


-,'~- 


3:n.jj 


-in 


Ht:- 


;; : r 


S!i_ 


:V 1 


p.t 


lT.fi. 




tep^i" 


-O 


,„4--'. 








: ! 5 


31 






; \i'.. 




' r, W 


:::?. 


:.:-■: 


;■:■ ! 










: :-: : 




:r.. ; 


;.| . ; 




ill' 


, , .-, 


\\r. 






LiU' '- 


T) 

C 




" ■ '■''"' i ."j.i i ,','.' 




J 












• m"- 


: 'f 






V 1 




1 * ' 


i:H ^ 


IS 


L : tl i: 




— r— 


1—1 
•H 



fa 

-»» 

V 

*j 

C 

o 

!- 

y 
o 

fa 

to <; < < 
X o ,-< o^ 


u 
a 
a 

ft 

0) 
S 

• H 

1— 1 
1—1 

flj 
+> 


- 

- 

': 

V '. 

■M 

c 


43 

flj 

y J 
i— i 



ft i 
-a 














'.:'; 




"■I 




:,, t 


■ •■',\ 


.iil'll 


\' l >' 1:1 

1.',- ! 


||l 


w 




;ii: 






;:;; 






i i 




\' 


. , . , 




, ; ' . I., 






> 












•" 




|' 


1' . . 










ii" ;• 


Hi' 




y 

fa 

C 

o 






' ■ ' 




a 


^ 


1 .' i ' 


"; ' 


: '::; 




1 : !' 


|; ; 


,: ! ; 


■:;||i 


.1-1 

■»-> — -, ■ 




ti : : 


i : , . 




: : 












\..| 


i'ii 'I 


'l ■'. 


' I ' li '■ 








i.* ' : 


!: ': 




i: : 


i'\' 


1 'j 




\ ! 1 


! : ' ! ; i 


tr ■ — 


■.:. J Ji*. 


0) (_. 




*- r .H 


U ' ; 


;;.: 


';; 


: hh • 


:: ! 




lii ; 


:- i; ! 


it; 




.)..... 


:■■■ T : l ; 


03 








•i." 1 




: Ui T 




i: . ■ 


; t ! t 






i:ii | 


1 1 j , : 


,::.{: ' I ! 


Q -~ i~ 



N 

•H 
■M 

< w£ 

i— l PJ i— i 




j-,:.- 


'.• j'r 


::!• 








; -!l: 


';" 


i : ■ i- 






l;l Mt 




• ' SL i ■'• 








; . ;;-i 


1 : n 








i-ii'l 






.t r| , 


i;H 




tt'ji;" 


'■]]• f ! ■'* 






^Hi 


•■> 




ilHfl 






t'-ri 


IHi 


■i i i.r 


n - 








;.:,: !;; J.j ;i j 


*" '",■',■' 


^i.ri 


SS 






iJ4^ 




■ i * t. 


sn- 




iitL 


4ftf 


B\^ 




■•■ ; l» n 


U h~ >s 


i;H 


l^r 








Per 
T 




-"> 




' ; 'V 




■rrt 


T * ji 


:ti : 


;-l., 












; i'|".. jriv 




•i t^ 


; '-. ■ 


' i ,- 


; -i : + 


if!; 




!l! ! 

1 1 












f ■ 


' ' . '"•!', 


■ ■ i 




rli"' 


., , ^ 


:.;■-' 


' ; n 


: :.; 


t I ; " 


.1 . . 






m;.i 


! 


lull 


.!: i v . 






■;■. 
ii 







i! It 


r; ; ; 


:.;lv 


™ r T', 


^i"i 


i'ii' 


'!"■: 


!,;;; 


: ; l; 


:-!:! 


"TH" 

:;tt 


W- 


J. "' 








Jill. 


?J: 


"i.L 1 . 1 . 


■14+!- 


:■: 


il't 


'■!' 




it 


'■I, 


,..j.j 


ni;i : 


j- -it -.. 


::: .i : i :: i 






>» a s 
a> flj 




•jjp: 


X^.-"' 


:r£: 


4'bi 




i ; ;i 


i . ' ; 




-. : .:- ; 


:: r 




; . ^x' "-■■ 


■f . • 


.:i-ri:-:-.'S 




'■V 

•A- 

it: 




. 


j "-.-7 


.TC 


-d«- 




i~h- 

'.iii 
_ 


itci 

4-i-m- 

—IX;. 


-j-1- 


l,'.r; 


':? 


i"i if 


.-i.^t- 




t-'iri'rr 


Lj:: 






i'~." 


•j+T!- 


trr- 




'■"' .-J... .1 _:..:. 


■■*■— U ' --i 


1. „ : 
.,.'..: 1-.! : l ■ l ■ i " ' ■ i i ■ ■ r ■ i i " 


'in* 


tH r 




^ 


V 


.... 


Tiv 


Tffi 


an 


ns 


m 


■md 


xuJ 


m 


# 


rrn 




mil 


_ 



1 o 

9 O 



o 

s 



I 



fa 



o 



S 



O 






o 



v\ 


c 


• 


» 


CM 


CXI 






4 






H 

# 



Report No. R63SCW-3"! -5 
Page 29 

Commutating capacitors are often subjected to ripple frequencies 
in this range in static inverter. Such ripple frequencies can 
be caused by minor resonances between inductive and capacitive 
components . 

Capacitance and dissipation factor test data for sintered and 
foil tantalytic capacitors versus frequency and temperature are 
shown in Table 97. 

The decreasing: capacitance characteristic with decreasing 
temperature and increasing frequency is a maximum of 12.65 
percent from -25°C to +85 C and 120 to 1200 cps as shown in Table 
B7. However, capacitance change variations of 59.6 percent and 
29.8 percent maximum for sintered and foil capacitors, respectively, 
over a temperature range for -55°C to +85°C and 120 to 1200 cps were 
measured. Similarly, the dissipation factor increases with decreasing 
temperature and frequency, within the frequency range of the tests 
as shown in Figure 10. These data and physical data comparisons shown 
in section 2.1 of this report are primarily for reference purposes. 

3.2 Commutating Capacitor Test and Results 

A method of correlating capacitor losses, while operating in 
commutating circuits, to capacitor losses measured with more 
conventional test equipment and sinusoidal voltage was sought In 
this study. 

Test 

A 2.5 microfarad, 500 VDC metallized polycarbonate' capacitor was 
mounted in a calorimeter and it's dissipation factor determined 
utilizing sinusoidal voltages over a frequency range from 4-00 cycles/ 
second to 50 kilocycles in a. 25°0 ambient. 

While mounted in the calorimeter^ th<? capacitor was tnen connected 
to a static inverter and operated as a commutation capacitpr. The 
capacitor heat loss was calculated from the calorimeter heat measure- 
meants. Calibrated oscilloscope pictures of the capacitor voltage 
and current waveforms vara obtained . during this test for use in analysis 
of the- cvpacitor losses. 



^v^i^7.v.y^*^v^^^-;a5m > 



*^W ^TsS^^.r:.- 



100.0 



10.O 



\5^. 



O 

CO 
C 

o 

•H 

to 

09 

•H 
P 



1.0 



.1 















Av«r»pe Dissipation Factor for T*nt*lum Capacitors 














k : 


;■'. 


.:•:!■■■• 
1 ' 














::!. 












itfi •-: 


. r ;i 




tH: 


:.i ;■; 


Hit- 








:•; 


HH 


:t— 


r f~ 










































,. t . ._;; ;- 




rt- 
















ii:i 
























ti- 




















' : r; : : 


















: ; 




r H.i 




































































-: ■!- 


! . 
















' '■ 
























::■•:-'■:"• 


; J ■: or:- 


: - 1' 


H=1 


^ 


p?*' 


<M 


ft-. 


: -1?'>^ 


ew-j 


!ii£ 






































;;;: 


1 .k: 


i ': :^ : n 


r .! i-frr- 


'.'• . 












•:::|!H: 






-!-- 




























. . : : 












• ' l.: Ir 


rff'f:- 


-.;: 


;::* 


»_& 


^Ti 


Wi 


' t r " 










:■; : 1 






















ki 




















M.:' :: 


: ■ -:iri 






iV : : 


•Hi 


■:k 












'-.'■' 


k: 














1 ■ 
























.": '. a -Hi 




.r* r !". 






4;.,.: 


i : i" 












Tin 










: :-i 


~ 


1 1 ■ ■ ■- 






:.-:i ' 




















i::: 














k- 






:H 


::•: 












^f 








$ 


*■£' 


W* 






: : .'H' '1 


if.ii.T_: 


i-H- 


k: 




j '.".' 


H-i 






.' i i 


, !T' 
























i ■ • 










;::■ 


'?:■:■ 


! '.' i 


kill 


i-to 


-H 


Hi: 




ri: 








;:.:; 




Hi: 


if 


: H ; - 














! . 
i 






.. 1.. 

.-.■I:. 
! ■ 














: " I 


1 ; h~ n 


+* TT -j 








:'- : : 


k- 










' ; -i 1 




V- 










t.i.T 
- "1 

H | 


■ - k 




k: 








- 






: :^ 


Lk; 


;k 


'-■ S B 
■'.iS it 


k'kt 


rk 




-:;: 


■if 


kt 




;-H : 


ii ff 


~ 


if 


4 k : - 


H H: 




1 ; 







'Jjj 


VT' ,! 


: -4 


. . . 

■■■- 


.; ■ 


ri' 


; :; : 






ft 

fit 


S 

!-i+- 


* 


fef 




' : ::r 


--rtti 

kkit 

- ni !] 

IS" ': 




! ! _l "! 


Vl" 


4H-' ; 


T; ■ ' 


r-; 1 


■fi 


-. j . 


k 


I 


iif 

TT*T 


;. f— l - - 


'44 i- 










■■"- 




.... 






-it 


'==: 






^ 


S# 


* 


^ 




ttn. 


f*#i4- 


itWF 


T^ 




-— _ 


Mtl- 


it;* 


i; i i 


l.k-i] 


.-,;- 


p* 


i«:ii 


















■. 


..- 










-"v 


kH 


fH± 




i i 1] 


n-i-1- 




9Sii«j. 




-(- ;+j: 


IS 






f'--' 


^ 


irL; 


44,1-. 


ki 

--;4- 


f ft 


H R -,- 

: ;i ■ 
lit;, 






"a 


:.; 


t- 1 ^ 


4— 


■-.— 


. ■ i . . -. 










.-:: 


--';; 


"'! ■ t 


tkk- 


! .-_ 




.:_:.. 


1 i-'i VI 


? V& : 


iiii:_ 






—■; 




■ : ' i :-f' 


i ; ii 


:.U_. 




i.r-ff 


T'":t 








V ; 




















' ?. : 










r|S u 


: Si!. 




":l-_ 


;i;-: 




iiii 


iiii 


1 :i. 


i'- :i 


trs 


-H""r- rr 


1 LHj 




!-■:■ 




T. : i 


.i-.j 


— :-.- 










1-; 






:■:: 










■.". : ; -; 


-f -'t; 








i:H 


ii ii 


:-^.: 




.!-: 


it-k 


itk_ 


i i ih 


if: :- 








i.i 1 !' 






• ..■' 










■ T 


; -■ ; 




:'i:f 






« : iili 




;.':. _ 1 




_-;. 


i':'^. 




i'H: 


4 :-;i 


r-L-r 


"i-^V 


4 lit 


fifi 


;-';: 




li,i : 












«k- 


:i ' -. 




.1,.: 




"k 


1T< ; 


, : . : 


i ; r-. ; 


t.-i-r 


nir 




n i".r 


f:f- 


:.:„-. 


: "j ; '" 


i-ik 


-Hij- 


ikii 


■ : f 


fh| i. 


ii ; l- 


+ t:r 


,33 
■j-fei- 




li !: 


;:.n: 


fil - 












;.:* 




iii_* 


-i ■' it 


i I f ' 




^ 


& 


ft 


it*: 


kk 


Ikl 


jif 


;"!1"j" 


lift 


sp- 


tfft 


3itT 


ii!-! 


il.i± 




k ; i| 


iRl 


i# 


ifi-f 


•1. !" 


T**t 
ItH- 










' : 






:!;:. 


nn 












^t;| : 


k ! " : '- 


-trri- ij 


T-r. : r; 


::- 




TT!T 


'*:Vt 


Hii- 


Hf 


TTTT 
?Tit 


irfl' 


!iii 


4T|f 


ihl" 
vi-ii 


itl 


k 1 


iTft 


ik-: 


tit 












•:'■; 




k : 






k:r 


^!t 




4 >-■ ■ 


i n Hi 


it ihi' 


:1t! 


.. ^ ^ .. 


liiTj'^j 


ti_t 
-iij-i. 


t ;_!_;_ 


+|T] 


irm 


l-Pf 


-\ i :.7- 

-n : ': 

i 1- ; T " 


fflj 

•f s *- 


(Hi 


f':" 


-V-i 


:«T 


tjH 


ft 














1 i i 


: <:i 




IT 


Jilt 


f 1 


/-;■] 




!>il 


...:. : J 


1 -':- 


-4 r . 


!'1i 


pf- 


-jii:r. 


fk 


lu t: 


•i 1 ^ 


4-I.1l 


i I if 


!f!f 

■r!i 


]k* 


H '■' 

"H "t": 




-' 1 

Hi 






;;..: 








Mi 




s£ 






/ill 


^4 


$ 


fe ;; 




8 


!"■; r i 


iitilr 




. , i : 


iili 


; 1 i_! 

1TT1 


: hi j- 


f f:- 


Ii' : 


I-t'4 

;|iji 


■ iif 
-r,i1 


fi- 

1 ' M 


Ik 

j ! 'i 


H-H 






■ r ■ 


■ f ■' 












Ik* 




ki: 


^ 


^ 


# 


i 


j] , 


Hi'. 


**-i 


.j: ;.;■'- 








JjJJ. 


f: 


th 


ri i : 


if 




in-;. 


ik: 


if; 


■ j k 


.k 




-■ k "r 

* ; i 


k rtt 

;-• rp 












H-; 




: 1 ~. 








' i : 


■si; 


kjl 


m' i: 


f if 


l 


%i 


i_H i 
T™ 




-i i : -j 

i-! i-. 


"*i 


i;t: u 
k'"k 


i ' j ! ' 

-i -ii V 

f+ ' !■!-- 

T TTTT 


i t-M- 


-1 | i 
— 


— 


.:l. 




:ii |- 
fjH 


iili 


if 

■ : i , 


: : f 
iili. 


ilk 


: if: 

if 

iif 


44.-"-" 

;iH 
n 1 ' 


■Ir:"l 


li! 


i - : 












i kl 






"r*s 


UK! 




i-i-ii 


:k 




iTr* 


■S4J T": 




•ri 1 




if 


:iii 


Hiv 


r'i *- 


: ; j.. 


44 it. 


j Hi 


iiii 














; ■ 






ik 


j"k 




k .: _i 


',"i 


!&> 




^§ 


jTki- 


J ■.; I: .|f 




•W 




^ft 




jirr 




i.r-l ;v 


!n-f 


!il- : - 


■Si-t 




ktr 


.iff 


iiii" 




H :J 












' ' : i 




- : ; ; 








if 


::rr 


4 l.i"'~ 


k-i 


ki! 


i] Tii ; 


■ifii 


j ' '.:'. 


■TTTH 

Tj il 


Lii-: 




#++ 




ft," 


; : '1 


If 




1"'i! 


i'" 






1-H-t- 


■ ; ; | 










i". : : 






. : k 




'" f 




1 : 1 ! 


1 : »' 


i; "Ii. 


'ii Ml" 


_■'■ u |J 




H-! - 




!i-i; 


r-'-'i. 


ii ; l 


!'::! : 




i ;-i-, 


n : = 


' 4 •■ ■ ■ 


T ; k 








T-'i 














,.,. . 










-!■; 


ti-.f 


■i'S 


:-f --- 


>-ii' ' '■ 


-j-ii* 


.-.!_- 




i !4- : 


i.t'i 






i;l : - 


; '.; t 


I] k 


i;:r 


: ■ ! • 










; ; :.-: 


4^- 


: ^- 




— 


_i.:.j i 


yji 


W 


in 


raa 


wS 


-IS 


$d 


ftl 


^ 


»-: 


&A 




:f 


M 


w? 


] ;-. J i 






Hli. 


r~jr 


iiii 


! i !-T 


4: it 


" r. . , 


Hii 


i:: : 


. . 1 


" j k 










.' i-i — 




! :r T 




ffifi 


m ' r 


S iiiv 




f : TT 




LlV;- 


\ -i : 




:U;i 


if; , L 


4 :-V\ 


r ' 1 r 


-rHi 


-H." : 


.!:.i 






1 H; 1 


■^-t-— 






T**» 


-l-Hf 


f"t 


tjk 
l-i.'i 


:T.!n 


^ 


^ 


^ 


P 


J^ 


^l 


i^i-S 


IJ't'l ^ 


IliS-t 


+W 


S 


#tt 


jaii 


*-k 


Sfl 


f& 


■*4r^ 


t*BS 


kT 


if I f 


fr ; 


jj ; 


4j4+ 


Trrt 


h=t *"T-t- 


' ' ' 










v. 




im 


frrf 


-inn 


% 


tMIi 


'• w 


1 


F^ 




'^ 


^ 


■prr 


] i_|j 


;.1.!1 

.Hj* 


■H-rt 


kfe 


k : 


■Mi- 


"14X 


5-1T 


Jk- 


*p"i 


nf 




U44. 


Ii- 


























h 1 " 


i:.4 


'kt 


kTrt 

; iif" l r 


kki; 

Sir 1 






ii -' 


iiL 


r;::i 


fr : 






■ i- , 


j':.l± 


H'h 


k;-" 


j -- 












};.!: 


f-l 


• f-i- 




f k 


fl 


H$ 


fcj 




rkT 


ft 




k-Tm- 


;;■[;■ 






ki- 
ll]! 


7 -1j 


"■f* ii 


k / 


■ ■ i- : 


:::.!. 




|i-H- 




i ! ' '• 


















r f 




- -1 ..[ 


■k; 


HI"! 


-j - 


11m 


ijlt 


-1-: !| 


Z'1 T 


k+4 

i-fi;- 4+ 

hrf| 


j-ll-t 

J ' . 

Ill'' 


lif. 


j.;i|- 


ff'* 4 
H-l-'-- 


fi 


LUi. 


if 
i in 


1 : T 




if: 






l.f 




f 4 " 




f :: 


f 






in 


■ i- ! 


", tt 
T. r .": 


i^i.i 








t-'i-i 




J.j:'.,: 


i-i-j- 1 

'Hi 


■1 '-'-■ 


Tfili. ff 


4- .: '— 

^T - 1 ■-■ 
_f -j_kki- 


f | h 




T+H 


fj? 


kil" 


SB 


a 


In 


:r. 


jkix 


Tt : ■ 


kf 




H :.:: 




- ! 4i 


t',; 
fkL 




..... i 


"'"1" + 


Tit" 


4ll|, 

jjjjU 


m.r 
Liu, 


.. .1 k 




k: 
"] T ' : 


liir 
kj-k- 


trir 
tk 

m 


T-l-rf 

r; fl- 


: -i 


1 r 'l 






rrh 

hitl 


ik 

!J4iJ 
FT]- 

-i_ 


-'■)■. 

ii.a 
1m 


kkj 

mr 

Hk 
-Uii 

1LU. 


knl 

jv: : 


if 


- "' f 


! ; f 


.: :l1 


-1-.J4 

iuii. 


ini 
fil 

kf 


■3+r 

fif 

XLu. 


Ibii. 


Mi 


f r 


frfr 
ILOi 




ftrttn. 

rH-i-f 
<-■-■ . 1 1 




4i(:r 


Jk rr 


-r n- 1 



.1 .2 # 3 .i* .5 .6 .if .7 .B .9 1*0 1,1 1.2 1.3 

Fr»jqt»aBy (ic) 






<>r,Wf»*£;-.y *-:-«*- T-.- _«%^j.'-irs. i s lfii( r,>^ 



Report No. R63SCW-31-5 
Page 31 



Results 



The dissipation factor of the commutating capacitor versus 
frequency characteristic is Shown in Figure 11. It should be 
noted that the dissipation factor for this capacitor is much 
greater than for the polycarbonate capacitors shown in Figure 9 
but the results of the test are valid. 

Actual edaauta.'fciig capacitor losses while operating in the inverter 
circuit were 0.91 watts. Capacitor voltage and current waveforms, 
during this test, are shown in Figure 12. The encircled portions 
of the waveforms in Figure 12 were subjected to analysis to 
determine the RMS voltage and currents. A detailed wave analysis 
was considered necessary because the waveforms contained a 
significant amount of ripple. 

Evaluating the rms volt-amperes of the encircled portions of the 
commutating capacitor waveform was accomplished as follows: 

1) Voltage and current values from expanded portions of the wave- 
forms as shown in Figures 13 and 14. were scaled at 5 micro- 
second ordinates as illustrated in Figure 13* These scaled 
values were squared and replotted as squared voltage and 
current curves. 

2) A polar planimeter was then used to measure the areas under 
these new curves (i.e. volt -seconds or ampere-seconds) . 

3) The areas under the squared voltage and current waves were 
divided by the 4-00 cycle/second time base and the square root 
of the resultant was taken to obtain RMS values of the voltage 
and current. 

4.) An assumption was made that the commutating current pulse 
conformed to a quarter sine wave function during an interval 
of 47.5 microseconds and had a peak value of 10.2 amperes. 
It was also assumed that the voltage function during this time 
interval conformed to a cosine function. From these assumptions, 
RMS voltage and current values were calculated for smooth wave- 
forms as shown in Figure .15. 

5) The RMS values obtained in 4-) above were subtracted from those 
obtained in 3) above. The difference between these RMS 
voltage and current values was attributed to the ripple frequency. 



-*-' v--?-i ^*\*?± *±?:,cp r - 



' V ~ V '" i »V~"':- , * i >e*.'' :■, ^-■ r T-:: '.■■,,.,",; .»*:;. -, . ■ 



E 

O 

o 



„ ! 






q 

•v ' 

♦H 
C 
CO 



C 



H 

ev 



o 

o 



c 
o 



CO 




% -J-Q 



tf> «0 N» 10 ' i/y . <* K> €Sj — o 



"-. ?w* m '?r^,T<i^ ; <^-^ : w;;*vf^^v^ii^w'. ■* i : -'^s- 



■ , >v^;<r TWK ?t:S5*«KTJE?-?.' -MWBBH 




~i*w o'.j^qpnUjjRi 




^"f-Hh+ 



in 

CM 



lUJ 

p 
b 



t 



^^-■.^"•-W^-r^/^rfW^^e,^.^,^.. 



Report No. R63SCW-31-5 
Page 36 



RW CTEHEHfT AID VOLTAGE FOR CCWTOTATIOH PULSE 





RMS CUX8LNT CMCULKTToN RtAS VOLTA<5-ff CALCULATION 

-t,«47.5>U5ffc j T*a56yUsecj f = 400CPS ; •f.m^^iSJCc 



flrrr\S = 



r 



p 






>L * Ip sin tc^t 



-L A«ns s. -Xi 



7T LX 4uu J. 






-r-2 

s Xp ^ 



" J » 



4uU 
I/uw\s - o,/38Zf 

Xp - 'O.2. 



Cruras -s. 0.093 £p 



FIGURE 15 



Report No. R63SCW-31-5 
Page 37 

From Figure 13, the commutation current pulse duration is 47. 5 
microseconds. The shape of this pulse approximates a quarter 
sine wave with a frequency of 

f = 1_= 1 =5.28 kilocycles 

4t 4(47.5)xl0-bsec 

The shape of the voltage waveform during the commutation current 
pulse interval is cosinusoidal. The RMS volt-ampere for these 
wave functions during the commutation interval is derived in 
Figure 15 and numerically is 16.6 volt-amperes. 

The sums of the squared volt-second and ampere-second values from 
the waveform, a portion of which is shown in Figure 13, are: 

(1) 4. 505.217 Volt -usec = 1820 volt 2 

2500 usee 

(2) 6646 amp* -usec = 2.655 amp 

2500 usee 

Vrms = 1820 = 42.6 volts 

Irms = 2.655 = 1.63 amps 

The capacitor loss from the commutation pulse with an equivalent 
5.28 kilocycle frequency is calculated using a dissipation factor 
of 1.85 percent for 5«28 kilocycles in Figure 11, 1.41 amperes rms 
and 11.8 volts rms in Figure 15 J 

VI (D.F.) = (11.8) (l.a) (.0185) = 0.307 watts 

Subtraction of the rms volt-amperes of the commutation pulse from 
the rms voltage and current totals yield the rms volt amperes of 
the ripple: 

42.6 = 11.8 = 30.8 Volts rms 
1.63 = 1.41 = 0.22 amps rms 

The product of the RMS voltage and current of the ripple and the 
capacitor dissipation factor determined with sinusoidal voltages 
at the ripple frequency yields: 

(1) Dissipation Factor from Figure 11 for 80 
kilocycles S 8.9$ 

(2) VI (D.F.) = (30.8) (0.22) (.089) = .603 watts 



Report No. R63SCW-31-5 
Page 38 

The sum of the watts attributed to the ripple and commutating 
pulse is: 

0.307 watts (commutation pulse) 
0.603 watts (ripple) 
0.91 watts 

The measured commutation capacitor loss in the calorimeter was 0.91 
watts and .a favorable agreement was achieved with the analysis above. 

3.3 Life Test and Results 

Description 

A total of fifty (50) capacitors were subjected to life test of 1000 
hours in 85°C ambient while energized with 4-20 cps voltage. This test 
was conducted to determine if the capacitors could withstand AC peak 
voltages equivalent to the DC voltage ratings with the exception of 
capacitors having DC voltage ratings above corona starting voltage 
region of 325-34-0 volts. Corona starting voltage is the voltage level 
at which ionizing of air or gases entrapped between layers of the dielectric 
roll takes place. These ionized gas pockets create hot spots that can 
progressively deteriorate the dielectric. Fifty (50) capacitors (33 
metallized polycarbonate, 12 polycarbonate/foil and 5 metallized paper 
capacitors) were mounted on the inner walls of a sealed aluminum box, 
shown in Figure 16. The surface of the aluminum box was maintained at 
a temperature of 85 +3°C for the duration of the test. Figure 17 illus- 
trates instrumentation and indicating lights in parallel with fuses that 
were in each capacitor circuit as illustrated in the elementary diagram 
in Figure 18. 

Ratings of the fuses in series with the capacitors were selected to be 
approximately three (3) times the rated capacitor current. Detection 
of a shorted capacitor during the test was by observation of the indica- 
tor lights that were on, indicating the fuses that had open circuited. 

Capacitors that may have failed in an open circuit mode only would have 
been detected when capacitance and dissipation factor data were taken 
at the conclusion of the test. 

Capacitors with 300 and 4.00 VDC ratings were energized with 325 volts 
peak. Capacitors with 200 VDC ratings were energized with 212 volts 
peak. The 135 and 157 volt DC capacitor ratings were energized with 
162 volts peak. 



Report No. 63SCW-31-5 

Page 39 













MOUNTING OF LIFE TEST CAPACITORS 



Report No. R63SCWw3t-5 

Page 40 







LIFE TEST MONITORING EQUIPMENT 

■■■■■£■ ' ■■■■■■■■i 




u 

IS 







mmmmmm l 

nnmmm*m* j 









Pi 

2^i^^,«] w lis 

ill 



a 

Is 



t 

i 



sill 



Report No. R63SCW-31-5 
Page 42 

The 325 volt peak limitation was maintained to prevent damage to 
the capacitors from corona. 

A strip chart temperature recorder shown in Figure 17 was used to 
continuously monitor the temperature of the aluminum box contain- 
ing the capacitors. The 420 cps voltages were recorded twice 
during each normal working day during the test. An elapsed time 
indicator was energized from the 420 cps voltage source to indicate 
total test time. 

Capacitance and dissipation factor data was measured before and 
after the 1000 hour 85°C ambient test to determine load-life 
characteristics. Survivors of the 1000 hour life test in 85°C 
ambient were placed in a 125°C ambient and the life test was 
continued for 370 hours. 

Results 

There were thirteen (13) capacitors out of fifty (50) that failed 
by short circuiting during the 85°C ambient test. 

Capacitors 3,4>5»&H * n Figure 18 developed short circuits upon 
step application of the 230 volt RMS, 420 cps power. This step 
application of power resulted in two (2) 350 V RMS transients and 
three (3) 280 V RMS transients from resonant effects between 
the capacitors and the transformer inductance. 

Capacitors 20, 21, 23 and 30 in Figure 18 developed short circuits 
upon being energized. 

The following capacitors developed short circuit during the life 
test as tabulated: 

Capacitor No. Hours to Failure 

13 
32 

12, 33 
31 

The capacitors that failed during this test were from one (l) 
capacitor manufacturer. The fifty ($0) capacitors subjected to 
this life test were obtained from six (6) capacitor manufacturers. 



234.7 


362.7 


474.8 


666.7 



Report No. R63SCW-31-5 
Page 43 

Analysis of the capacitors, that failed, by the capacitor 
manufacturer revealed that the causes of failure were attributed 
to construction techniques that have been improved upon since 
these capacitors were manufactured. 

Five (5) capacitors developed short circuits during the 370 
hours test at 125°C ambient. Four (4) of these five (5) capacitors 
numbers 2, 19, 30 and 49 in Figure 17 failed within the first 
six (6) hours of the test. The fifth capacitor, number 1, failed 
between 80 and 14-0 hours of operation in 125°C ambient. 

The majority of the capacitors surviving the 1000 hours 85°C 
ambient test exhibited a small increase in capacitance and a small 
decrease in dissipation factor is illustrated by the test data in 
Table B8. Capacitance and dissipation factor data were not obtained 
after the 370 hour, 125°C ambient test. 

4.0 Conclusions 

Appreciable size and weight reductions of commutating and load 
filter capacitors in static inverters may be realised with the use 
of polycarbonate capacitors instead of paper dielectric capacitors. 
These potential reductions are attributed to the smaller dissipation 
factor (i.e. power factor) of polycarbonate capacitors. 

Polycarbonate capacitor power losses are quite low between excitation 
frequencies of 0.4- to 5.0 kilocycles but increase significantly 
between ^.0 and 30. kilocycles. In commutation circuits, capacitor 
losses may be kept low by reducing resonant frequencies and magn:' tudes 
b.tween Lnductive-capacitive components (L~C ringing.) 

Survival of polycarbonate cpacitors, from fiv^, manufacturers, in the 
1000 hoar 8^°C ambient and 370 hour, 125' C ambient test is encouraging. 
This t,ypi'; performance suggests that a smaller voltac:^ derating than 
the ';> to 1 from D.C. to A. C. applications for paper capacitors may 
be utilize."!. This voltage derating for AC applications is done to 
prevent excessive capacitor heating caused by increased capacitor dissi- 
pation losses vorrj-it; frequency. 

5 . Re c oinme n da t i o n s 

Life testing of polycarbonate capacitors should be expanded in 
quantities of capacitors, voltage levels and temperature levels and 
renewed in an effort to define the relationship of reliability versus 



Report No. R63SCW-31-5 
Page UU 

Additional developmental effort^ by capacitor manufacturers, to 
reduce the dissipation factors in the extended frequency range 
from 10 to 50 kilocycles should be encouraged so that greater 
static inverter efficiencies may be realized. 

Radiation effects testing of polycarbonate capacitors should be 
accomplished to provide designers of aerospace static inverters 
capacitor degradation characteristics from radiation encountered 
in the space environment. 



Report No. R63SCW-31-5 
Page 45 



National Aeronautics & Space Administration 

Levis Research Center 

21000 Brookpark Road 

Cleveland* Ohio (U135) 

Attn: B. Lubarsky MS 500-201 (1) 

K. L. Cuaeings MS 500-201 (1) 

H. T. Musial HS 77-1 (l) 

J* B, Dilley MS 500-309 (1) 

George Mandel MS 5-5 (1) 

V.'JU l*lli MS 500-203 (1) 

J, rYSttitter MS 500-109 (1) 

C, S. ^orcoran, Jr. MS 500-201 (l) 

E. AittmtnikMS 500-201 (l) 

A. C. terr MS 77-1 (1) 

7. Oouraah MS 500-201 (3) 

R. D. Shattuok MS 21-5 (1) 

Dorothy Morris, Librarian MS 3-7 (2) 

Alice Dill, Report Control Office MS 5-5 (1) 

National Aeronautics & Spaee Administration 
Goddard Space Flight Center 
Greenbelt, Maryland 
Attn: F. C. Tagerhofer (l) 
H. Carleton (l) 

National Aeronautics & Space Administration 
Marshall Space Flight Center 
Huntsvllle, Alabama 
Attn: James C. Taylor (M-ASTR-R) (l) 
Richard Boehme (M-ASTR-EC) (l) 



National Aeronautics & Space Administration 

Manned Spacecraft Center 

Houston, Texas 

Attn: A. B. Eickmeir (SEDD) (l) 

National Aeronautics & Space Administration 

4th and Maryland Avenue, S.W. 

Washington 25, D.C. 

Attn: James R. Miles, Sr. (SL) (l) 

P. T. Maxwell (RPP) (l) 

A. M. Greg Andrus (FC) (l) 



Report No. R63SCW-31-5 
Page 1th 



Naval Research Laboratory 

Washington 25, D.C. 

Attn* B. J. Wilson (Code 5230) (l) 

Buraaii *f Navel W fi fifW 
I*partaent of the Na^ 
Washington 25, D.C. 

Attn* VvT. Beataon (Code RAEE-52) (1) 
Milton Knight (Code RAEE-511) (l) 

Jet Propulsion Laboratory 
4800 Oak Grove" Drive 
Pasadena, California 
Attn: G. E. Sweetnaa (1) 

Diamond Ordnance Fuze Laboratories 
Connecticut Avenue & Van Ness Street, U.W. 
Washington, D.C. 
Attn: R. B. Goodrich (Branch 940) (1) 

U.S. Amy Research & Development Laboratory 

Energy Conversion Branch 

Fort Monmouth, New Jersey 

Attn: H. J. Byrnes (SIGRA/SI-PSP) (1) 

Engineers Research & Development Laboratory 
Electrical Power Branch 
Fori Belvoir, Virginia 
Attn: Ralph E. Hopkins (l) 

Aeronautical Systems Division 
Wright-Patterson Air Force Base 
Dayton, Ohio 
Attn: Capt. W. E. Dudley - ASRMFP-3 (l) 

University of Pennsylvania 
Power Information Center 
Moore School Building 
200 South 33rd Street 
Philadelphia U, Pennsylvania (l) 

Duke University 

College of Engineering 

Department of Electrical Engineering 

Durham, North Carolina 

Attn: T. G. Wilson (l) 



Report No. R63SCW-31-5 
Page 47 



National Aeronautics & Space Administration 

Scientific and Technical Information Facility 

Box 5700 

Bethesda 14, Maryland 

Attn: NASA Representative (6 copies + 2 repro.) 

AiResearch Division 
Garrett Corporation 
Cleveland Office 
20545 Center Ridge Road 
Cleveland 16, Ohio 
Attn: W. K. Thorson 

Westinghouse Electric Corporation 
Aerospace Electrical Division 
Lima, Ohio 
Attn: Andress Kernick (l) 

G. M. Defense Research Lab. 
General Motors Corporation 
Santa Barbara, California 
Attn: T. M. Corry (l) 

The Martin Company 

P.O. Box 988 

Baltimore ,Maryland 21203 

Attn: Mike Monaco MS 717 (l) 

General Electric Company 
Specialty Control Department 
Waynesboro, Virginia 
Attn: Mr. Lloyd Saunders (l) 

Lear-Siegler, Incorporated 

Power Equipment Division 

P. 0. Box 6719 

Cleveland 1, Ohio 

Attn: Mr. Robert Saslaw (2) 

The Bendix Corporation 
Bendix Systems Division 
Ann Arbor, Michigan 
Attn: K. A. More (l) 



Rap&rt NO.R63SCW-31-5 
Page AS 



Tha Bendix Corporation 
Bed Bank Division 
1900 Huloan Building 
Dayton, Ohio 
Attn: R. N. Borngtiaw (1) 

VARO, Incorporated 
220U. Walnut Str^t . 
Garland, Texas 
Attn: J. H. Jordan (1) 

Aeroapaca Corporation 

P. 0; Box 95065 ' 

Los Angeles 45, California 

Attn: Library TachnJ^ft -Documents Group (1) 

Engineered Magnetics Division 
Gulton Industries , Incorporated 
13041 Cerise Avenue 
Hawthorne, California 
Attn: Burton J. MoComb (l) 

AiReseareh Mfg. Co. 

Div. of the Garrett Co. 

P. 0. 5217 

Phoenix, Arizona 8501O 

Attn: Mrs. J. F. Mackenzie, Librarian (l) 

General Dynamics Astronautics 

Dept. 963-2 

5001 Kearney Villa Rd. 

San Diego, California 

Attn: R. Schaelchlin (l) 

Prestolite 

Toledo, Ohio 

Attn: J. F. Carey (l) 

Director of Research Lab for the Engr. Science 
Thorton Hall 
University of Virginia 
Charlottesville, Virginia 
Attn: A. R. Kuhlthau (l) 

Radiation Effects Information Center 
Battelle Memorial Institute 
505 King Ave. 
Columbus, Ohio 43201 
Attn: E. N. Wyler (l) 



Report No. R63SCW-31-5 
Page 4-9 



Ihsajpeon->Ramo Vooldridge, Incorporated 
72$|; Piatt Avenue 
CleiNtLand, Ohio 443B4 
Attn* J. R. Murray (l) 

Lear Siegler, , Incorporated 
Power and Controls Division 
632 Tinton Avenue 
New York 55, New York 
Attn: J. Rambusek (l) 



*"»*■. 



Report No . R63SCW-31-5 
Page 50 -.-",'■■ 



APPENDIX A 
Comaatatlng Capacitor Analyses of Operation and Specification 
Analysis of Operation 

» Half Bridge Circuit 




C - Commutating Capacitor 
L - Commutating Reactor 
SCR - Silicon Controlled 
Rectifier 



St*2 



Figure A 

Analysis of operation for this circuit is based on these assumptions: 

l) Inductive load current does not change during the commutation 
interval,. 

2) , JFull load turn-off time is one half that at no load. 

3) When capacitor voltage is zero, turn-off interval is over. 

The equivalent circuit shown in Figure B is for the condition when SCR2 has 
just been triggered into conduction and SCR1 conduction blocked at beginning 
of commutating interval. gi 



-. :* 



^ 



Ik- 



-X, 



loM> 



«* 



Figure B 

Writing the voltage equation *f or_the loop 

E + Eq * 1 J !]_ (t)dt + L dl 2 (t) where t is turn-off 



C Jb 



dt 



interval. 



E Q is the charge on capacitor 
at start of commutation 



Report No. R63SCW-31-5 
Page 51 

Equations describing the equivalent circuit, shown in Figure B, in Laplace 
notation are i • "' 



2s 2s sC 

I 2 (s) = IjCb) - I 

.-';■ ■ ' s~ ' . 



sC 



I^s) = E + 2L I c 



1 + sL 
sC 



Substituting equation U) into equation (3) gives 



E„(s) = E - E . -2L_I 



c 



S_ 



2s s(l+LCs 2 ) (l+LCs z ) 



(1) 



(2) 



E c (s) = £ - I^s) 

".,'• ~scT' .;."" ' ,("3) 

Combining equation (l) and (2) gives 

E + 2LI =1,(8) (1+ sL) 



U) 



»-l 
oC E c^ s ) = E c^) = S - E(l- cos tat) - 2L I o> sin u)t (5) 

, 2 

where OJ = \ I 1 

y lc 

when E c = o, turn-off interval is over. Solving for t 

E = E cos Ut - 2 L I W sinWt (6) 

2 



Report No. R63SCW-31-1 
Page 52 

At I Q = o (no load or current that goes thorough zero at end of half cycle) 

£ cos wt = S ; cos wt = .5 ; i*>t = tt/3 
2 

t = tt/2w 

Using the second assumption that turn-off (t off) at full load is one 
half of t, t off = tt/6w = ^ ^p-—- 

6 

Using t off in equation (6) gives 

E = E cos n/6 - 2 L I Q U) sin tt/6 
2 

Substituting 1/ fLC for U> 

TCT 2 H— 

0.366 E = I V L/C (7) 

To determine the capacitance (C), divide equation (7) by t off, which 
gives 

0.366E = 6 I Q 
t off ttC 



C = 5.22 I t. nff 

o ■ 



(8) 



Inductance value (L) may be determined by multiplying equation (7) by 
t off, which gives 

L = 0.7 E t off (9) 

I. 



E in equations (8) and (9) are minimum steady state source voltage and 



;,. Report No, R63SGW-31-5 
Page 53 ; 

Capacitance values for circuit configuration*! shown in Figure A are 
determined by the minimum steady state voltage levels, I Q and to if 
of silicon controlled rectifiers listed in Table I. 

TABLE I 



Silicon Controlled Rectifiers 


Minimum Source Voltage Le 


Type 


I t off 


25V 50V 90V 




(amps) (usees) 


C(ufds) 


G.E. C-ll 


6 12 


15 7.5 4.2 


G.E. C-40 


16 12 


40 20 11.1 


G.E. C-55 


60 20 


550 125 69.0 


G.E. C-80 


120 25 


*■ 625 313 174.0 



Similar analyses for circuit configurations shown in Figures C and D will 
yield capacitance values equivalent to one half the values shown in Table I , 



<»—— \LQAb \ —~* 



^J. 



»*CR| 



SCRZ 




Figure C 



Figure D 



Capacitance values, in parallel type inverter configurations, shown in Figure 
E, are equivalent to one fourth the values of C tabulated in Table I. 



sou] I )|_ ' Jt Z ' 



Wt 



— E 



Figure E 



Report No. R63SCW-31-5 
Page 54 



Values of commutating capacitances selected for the capacitor survey- 
are 5, 15 and 50 microfarads. Specifications for these capacitors are 
contained in this appendix. Selection of these values is based on the 
use of multiple series and parallel capacitor networks for half bridge, 
full bridge and parallel types of inverters and converters. 

Peak recurring capacitor currents are determined as follows. 

From energy storage equations 

LI P 2 _ E 2 C 



I = E yC/L, where E is maximum steady state source voltage 

Substituting C and L from equations (8) and (9) 

I = 2.74. I Q E max. for values of C (10a) 

p E min. 

I = 1.37 I Q E max. for values of C/2 (10b) 

p E min. 

Peak recurring capacitor currents for the selected capacitance values 
are determined from silicon controlled rectifier types, I Q and C values 
from Table I and are shown in Table II. 





TABLE 


II 








Rectifier 
Type 


Voltage 
Level 


Capacitance 
(ufd) 


Capacitor Circuit 
Configuration 


Ipk 
(amps 


C-11 
C-4.0 
C-80 


25-35 
50-65 
90-105 




15 
10 
50 


C 

C/2 

C 


23.1 
28.5 
109 



^*„ 



Report No. R63SGW-3I-5 

Page 5$; ■*•;,■ 



E8QIKBBRING SPECIFICATION 



1. Scope — This specification is for an industry survey of capacitors for 
application in 115/200 volt, 3 phase, 400 cps output static inverters 
and converters in space environments . The criteria to be used in 
this survey are: 

A. Yoluae to capaoitanee ratio 

B. Weight to Capacitance ratio 

C. Power factor to thermal resistance ratio 

D. Volume and weight to energy-storage ratio 

E. Cost to energy-storage ratio 

2 * Range of Capacitor Ratings 

A. 5 microfarads +20-10$ over temperature range 

Capacitor No. 12 3 

Peak Voltage 35 65 105 

Peak Voltage (lOQasec D.C. Voltage 52 97 157 

Transient) 
Peak Current Amperes at D.C. Working 4.6 8.6 11 

Voltage 
Peak Current and Voltage Waveforms Fig. Al Fig. Al Fig. Al 
Peak Current Amperes (100 msec. D.C. 

Voltage Transient) 6.85 12.8 15.1 

B. 15 microfarads +20-10$ over temperature range 

Capacitor No. 

Peak Voltage 

Peak Voltage (100 msec. D.C. Voltage 

Transient) 
Peak Current Amperes at D.C. Working 

Voltage 
Peak Current and Voltage Waveforms 
Peak Current Amperes (100 msec. D.C. 

Voltage Transients) 



1 


2 




3 


35 
52 


65 
97 




105 
157 


13.8 


26 




33 


Fig. Al 


Fig. 


Al 


Fig. Al 


20.5 


38.^ 




49.3 



Report No. R63SCW-31-5 
. . .-- Page 56 

C . 50 microfarads +20-.!0Sjt over temperature range 

Capacitor Bo. 1 2 3 

Peak Voltage- 35 65 105 

Peak Voltage (100 msec. D.C. Voltage 52 97 157 

Transient) 
Peak Current Ampere at D.C. Working 46 26 110 

Voltage . . 

Peak. Current and Voltage Waveforms Fig. Al Fig. Al Fig. Al 
Peak Current Amperes (100 msec. D.C. 68 129 165.0 
- Voltage Transient) 

3. Physical Size and Weight— Vendor to recommend optimum size and weight 
proposed capacitance value and voltage rating. 

4. Ambient Temperature — -55 to +85°C heat sink ambient with a maximum 
of 125°C capacitor hot spot temperature. 

5. Method of Capacitor Power Loss Taisfer — Conduction through mounting 
surface to heat sink ambient. 

6. Construction — Hermetically sealed. Capacitor to be subjected to 0.25 
atmosphere within equipment enclosure during optional life. 

7. Shock — Capacitor is to withstand three 35 g shocks in each direction 
along the major axes. The applied shocks are half sine waves of 0.008 
second duration. 

8. Vibration — While energized, capacitor is to withstand the following 
sinusoidal vibration requirements: 

Frequency Force or Displacement 
5-20 0.3 inches D.A. 

20-100 5 g*s 

100-500 lOg's 

500-2000 15 g' s 

Duration of the applied vibrational forces is four 15-minute logrithmic 
sweeps from 5-2000-5 cps at the specified levels and 10-minute dwells 
at each resonant frequency found during the sweeps. 



Report Ho. R63SCV-31-5 
Page 57 

9- Radi*foOft« ~€apacitor shall have a tolerance to the following integrated 
radiation dosage without ifcalfunetion: 

A. 5 x lO^UTT Fait lew^rons/on 2 

B. 5 x 10 7 ,RADS (Carbon) Gamma Rays 

10. "lrTrrtl?**" 1 I^ f* ObieotiTB — >Energized for 3 years continuously while 
exposed to the radiation listed in 9 above and 85°C heat sink tempera- 
ture. Capacitor to remain within oapaoitanoe tolerance. 




Report Ho. R63SCW-31-5 
Page 58 



FIGURE A-l 



PEAK CURRENT AND VOLTAGE WAVEFORMS 



FIGURE A-l Tl = 74.5 microseconds 
T2 = 1256 microseconds 



Report No. R63SCW-31-5 
APPENDIX B Page 59 

Capacitance and Dissipation Factor Test Data 

Data contained in Tables Bl through B3 were obtained by bridge 
measurements and calorimeter data in 25°C ambient with sinusoidal 
voltage waveforms. 

Data contained in Tables B4 through B6 were obtained by bridge 
measurements over the temperature range of — 55°C to +85°C and 
frequency range from 0.4 to 10.0 kilocycles. Electrolytic capacitor 
test data contained in Table B7 were obtained by bridge measurements 
over the temperature range from -55°C to +85 d C and 120 to 1200 cps 
frequency range. 

Life test capacitor data contained in Table B8 were obtained by 
bridge measurements before and after the 1000 hour test in 85°C ambient. 
These data were taken in 25°C ambient with a test frequency of 1 
kilocycle . 

An example of the dissipation factor corrections applied to the 
bridge data is given below: 

From Table Al, capacitor No. 1A, the dissipation factor as 
determined from the bridge data at 10 kilocycles is l.llOSt. 
As measured with the calorimeter, the dissipation factor at 
10 kilocycles is 0.975/6. 

By ratio of the bridge data of the dissipation factor for 
capacitor number IB and 1A, the corrected dissipation factor 
for capacitor IB at 10 kilocycles is: 

tjlo* x -975 = -840* 






V. o 



«o. 




u. 
o 






CD 

UJ 
_l 
CD 
< 






o v 
a. oc 

K & 




<\1 

o 



u 

hi 
-L 



CD 
UJ 

CD 

< 



o o 



*(S j^lu <c<ou^^ <c"OoA^u <r"QJ^^ 



u 

(0 




C\J 

CD 

LU 
_J 
CD 

<t 



t/1 




CD 

bJ 

<t 
i- 



* 





v> 




8 


v» 


*•. 


1 


* 


* 


5 


"V 




iu 


mi 


<1 


V- 


s 


* 


& 


$ 




M 




1 


V 


^ 


Jf 




c\ 


?■" 




K 


SS 




1/ 


! K 


*v. 


«•> 


•J 


ft 




V 

£ 






<* 


•S 

S 


£ 


O 






(J 






A «4 VJ _, m 

m vi o 55 - 



<r \u A '-u u ^ 

-^ W *1 Ul v^ r * 



<> 



i^ ^ Uj,A Uj - , 
" ^ <\J vi lo >J *X ' 






z z 



^ 



"x 



V 


^ 


•<. 


K 


\* 


v. 




X 




\ 


*""» 


.< 


V 


•J 


>i.' 


.* 



1> 



v 



*' <!■ 




CL- 



•JD 



is 

(J 

"O -O 
OS -Ol 



<u Q, 



'u 






* 




^ 








-1 




*' 


X 




•<l 






fv 






S 


* 

^ 




V 


V/ 




*■'* 






5* 


V 




* 


^ 
*■ 




<• 


K, 




i 


-.J 
hi 




i » 


-* 




•\ 


W 




rx 






> 








'* 




V. 


\ 




\ 1 













* 



■-a 









V) 

X 








! 


• 

+ 


i 




1 


i 


.0 

1 




* 


s\\^ 


X^X 


£ 


X 


:v 






^ 


\\ 


A 


>o 




\s 


N 


5 


sX 


\ s 


^ 


X 


^ 


s 


s\ 


\ 


\ 


X 


s^ 


5 


X 


\ N 


^ 


X 








• 








i 












1 
























^ 
















T 








i 












i 












I 












| 














^ 








4 

0> 








<5 




4 




^ 


a 







a 

X 




^ 






a 

^1" 





s 




^ 
















f4 


ri 


M 


rO 






A 


A 


X 


<f\ 




1 


A 


ti* 


lC\ 




^ 


A 


A 


A 


cO 




| 






































^ 












$ 






X 






^ 












\XX 


x\\ 


5 


N 


^ 


^ 


N 


s 


5 


^ 


N^ 


<; 


^ 


^ 


5 


X 


\ s 


NN 


X 


^ 


s 


\^ 


X 


X 


<> 


N 


5 


^ 


X 


\^ 


S H 






« 

* 

X 


i 








i 


■S 


is 


^ 

"< 


< 




1 


< 




3 


3 




| 


"4 
< 


fe 

^ 


5 


3 

US 




^ 












v» 






•-6 


o 
m 








1 


41 


A 


oi 




1 




A 


i 






| 


A 


H 

5 




R 




| 












ro 




In 






OO 




? 


■* 
% 




i— 
? 






1 






a 







| 


00 


si 


ri 






^ 












\ 






0» 








i 


00 




i 


to 




i 




•ft 


A 







| 




m 


00 


00 
m 






^ 












^. 




in 






5. 




^ 


rfi 




A 
a 






s 












$ 


00 


CI 


ft 







^ 












N\\"i 


XXX 


5 


X 


^ 






s^ 














y 


nN 


^ 






X 








>s\ 


N^ 


^ 


5S 


\" 


^ 


X 


^ 




Q 

oo 
rvi 






1 




^ 


5 
> 




OO 

OO 


C-r; 




b 




-8 


1^ 


5 

u 

vi 




| 


A 

Vn 










^ 












V"t 






1 

r4 


1 






^ 




Q 

pi 




« 

^ 




$ 








> 




^ 














| 












'»"\ 




^1 
oO 
If' 

(3 


DO 


o 
a 

OO 


<5 
Q 




! 








* 








Q 

o 

rtl 


OO 

oo 






$ 




4 


c5 


5 

in 




$ 












- 


i 


S 


-A 






IN 




^ 












^. 




w. 


O 

V 




ON 
'i 




1 











1 


--a 




3 






^ 


5! ° 




m 


Ho 




^ 














U\ \ 








L\3 


K N 


k\ 




— - 

<1 


OO 




k\\ 


\ V 1 


k\\' 


S^ 




\ \ 




vA 


s\ 


X 


X 


lx 


-> 




rr, 


OS 


Pi 








-a 

OO 






oo 
«\ 

Oo 

0- 

4 




3 


1 


? 

^ 


r: 


^1 
O 

ri 


•* 

1! 
















^ » 




J? 

Ml 




IS 






o 

IN 


4 


1 










1 












n"'i 






A 




ft 




■X, 


% 

oo 


•J 

Oo 


OO 






1 




< 


o 
a 




•- 


A 




00 

ri 




1 












- 




•> 

•** 


1 

f4 


V9 

r- 


^ 
^ 




o«> 








^ 
g 


8? 

r> 




i 








Ci> 


^3! 
ri 


1. \ 












* 


1 ^ 

i -i 


o 

-*• 


-J5 


r-. 




B 






4 








Ok 


0, 


00 




^ 




r- 


r4 


rr> 

ri 




I 














V 
o 

lo ! 


Co 






<-t- r- 00 ^s 
^ ■ — 





CD 

LlJ 

1 

CD 

<t 

h- 



a. <& 





<J) 




Ml 


a- 






p 




s 








'-' 










'■t* 


^: 




-11 






X 


v.> 




\'> 


1~ 










\ 




-> 








-J 











:i 


•.J 




fl 


*u 






J 




M 


'■-n 






■ J 






~^ 




A 






1 .. 






^ '"; 


*-~ 




■ r 


-c 




^ .' 


ii. 






CD 



UJ 



u 

10 



vO CO 
05 vol 





z 



o _ 

05 



00 



s. 



£ 






Hi 



"J 
U. 






'Hi 





s 

Ul 

u. 

e 

u. 

a 

■5 
J> 

Hi 

« 

4. 

in 
a: 

N 

Ut 
O 












1 




! 




1 






<S\ S 


s\\\ 


5 


tf 


N N 


N 


\ S 


<: 


^ 


^ 


N 


^ 


s^ 


^ 


* 


N 


N N 


N^ 


N 


<^ 


2 


,\ 


\' 


\ 


\ S 


sN 


^ 


N 


N S 


^ 


N v 








41 




u. 


U. 


u. 


1 


4e 


U. 
41 




a. 


q. 
C 


^ 


«> 

c 










| 












^ 
















SE 


o. 






(A 
<3 


2 


1 




Oo 

OO 




in 


i 


i 


«DO 




Oo 








^ 












f 














£& 


ti- -> 


3 

* 




Oo 
r4 


■*? 


s 


| 


"8 


? 




*< 


? 
•t 


^ 


a 


? 








I 












^ 














&* 


-a 


o 


« 
Q 

N 


<s' 




1 


1 


& 

V 


s 


SI 


is 




^ 


CM 


Oo 








-1 












^ 
















*4 

5* 


J" 


| 


!J 


^ 


UJ 


§ 




Co 





# 
* 


vu 


^ 


<0 


O 








| 












S 














\\\1 


\\N 


5 


k 


\' 


^ 


N 


s 


5 


^ 


S^ 


s; 


^ 


N v 


5 


N^ 


S s 


^ 


N s 


^ 


s 


\ S 


S^ 


k 


^ 


h 


5 


sN 


N 


k S 


sX 


^ 






4> 


5 


8 


^ 


4! 


^ 


u. 


^ 


1 






^ 




«: 








| 




lQ 


is 

1 




>. 


| 


A 










■J 


/ 

* 




3 


to 
in 


If 


V 
m 


*> 


§ 


a 
% 


A 


j 


J 


1 


1 


Oo 


Oo 








| 








2 
1 


t 


| 




^ 
^ 








^ 

u 
k 


H 




« 
& 


A 


U 






g 




% 


a 


< 


! 


$ 


p» 


2! 








| 




5 
If 


> 
< 


* 
« 




1 


<4J 


I 








III 


%*■ 


of 




Co 


m 
-* 












9- 


Oo 


OO 


1 


til 

on 


Oo 








1 




T 
<J 

4. 




V 

u. 

it 




| 


~T 










■4 




3« 


£ 


1 


<J 


-9 


|u 


^ 


s 








01 


1 




O 

■r- 








| 












| 












i 


\\\N 


\\\ 


5 


s; 


\ X 


^N 


^ 


sN 


2 


sN 


M 


^ 


^ 


£ 


3 


S S 


X 





N 


$ 


s 


S: 


s^ 


^ 


\^ 


X 


5 


sN 


^ 


s^ 


N' 


ft 






1 


* 


4e 


1 


1 




1 


« 
s 


t 


* 




1 










^ 


1 


^ 


5: 


4e 


I 


^ 


§ 


i 


i 


i 


1 


* 




A 




j£ 






? 


4 
;* 






rl 


it\ 


? 


X 

a 


1 








A, 
ifl 


s 


§ 


? 




1 


o* 




| 




5 
o 




^ 
% 


A 


*,' 


^5 








rt 


*, 








iO 


to 


t 




1 


■2 

in 


o 


OJ" 






| 










? 


| 


l>\ 


In 




ir\ 


^ 

?. 




A uj 


«J.Sc 




0» 

0; 


r- 

oo 




N 


1 






t^ 
* 

T; 






1 


1 


in 
% 
ri 




"rt 

o» 


9 


| 








ir> 


in 


i 






Oo 


o 




•■i 




^ 


SO 





^ 


\U 


S 






5 


5 


« 




^ 


fc 


v9 


o 


Ul 


^ 

■^ 


1 


rn 


< 
ft 

\ 







UJ 


1 


^ 
■+ 


4 


5; 


^ 


§ 




\\\ N 


K\v 


K^ 


\ 


v\ 


[ N V - 


KN 


si 


K 


Ls 


^N 


\' 


K^ 


^ 


3 


N 


k^ 


s^ 


s? 


fc 


% 


•S 


$ 


!- 


^ 


^ 


s 


^ 


sV 


\ S 


s\ 








i 

s 


£ 

*: 


* 


5: 


* 

s 


^ 


5: 


a 


1 


1 


I 


1 


J 


^ 


^ 

«: 


^ 


1 


1 


a 


I 


i 


? 


^ 
















/ 






OS 


5; 


OO 


i 


OO 


* 


•n 






^ 




i<\ 




■J 




I 


■£ 


5 


% 


jj 

>* 




1 


t 


* 


*4 




•*> 

ii 








J 


a 


in 






I 


ft 


i« 




rt 


if 

IT* 


i 




a 


0- 


oo 


3 




•* 




2 


i 


^ 


1 


F 

lO 


if> 


A 
^ 


a 

* 


A 
o 

OO 




4 F 


1 1 u- 


4- 

3". 


>> 








I 






i 






1 


i 








1 


1 




N 




oo 
lit 


a- 


^ 

^ 


in 

N 


i 

N 


ri 


4 

in 


o 






3* 


cQ 


o 


A 


UJ 


<*> 


^ 

\ 












g 


\8 




V6 




U) 

3 


S 


5 






on 


Uj 
5 


1 


1 


i 


«J 


§ 













CO 

CD 

Ld 

_J 
CD 
< 

1—