[R-390] Capacitor Tester

[2002tii]

Miles wrote:

>In my opinion, the most useful capacitance checker of all is a simple
>voltmeter ­ a VTVM preferably or a 20,000 Ohms-per-volt VOM otherwise.

[much very good stuff snipped]

I concur with Miles, and will add some of what I 
see as the appropriate analysis for this 
conclusion.  I also highly recommend a variac with an ammeter as a test tool.

The first thing to consider is what you need to 
know about the capacitors:  (1) Do they 
leak?  (2) What is their capacitance?  (3) What 
is their dissipation factor (loss)?  (4) How much 
dielectric absorption do they have?  (5) What is 
their ESL (equivalent series indictance)?  (6) 
What is their ESR (equivalent series 
resistance)?  (7) What is their voltage 
coefficient of capacitance?  (8) What is their 
temperature coefficient of capacitance?  (9) Are 
they microphonic?  And (10) other subtle 
characteristics.  I have listed these in 
approximate descending order of importance to someone working on boatanchors.

As it turns out, only (1) -- whether they leak -- 
is of much importance in the boatanchor 
context.  There are some capacitors in 
boatanchors where the value needs to be in (or 
close to) tolerance (e.g., non-trimmable tuned 
circuits), but very, very few capacitors in 
boatanchors are at all critical with respect to 
any of the other parameters.  Furthermore, most 
capacitor problems besides leakage (and also, 
many leakage problems) can be determined without 
any kind of capacitor-measuring equipment.

The second thing to consider is how each 
capacitor is used.  The vast majority of 
capacitors in boatanchors are bypass caps (their 
function is to remove or reduce AC voltage riding 
on power supply lines ("B+ bypass caps" or 
"supply bypass caps"), tube cathodes ("cathode 
bypass caps"), and pentode screen grids  ("screen 
bypass caps"), etc.  When they are used to 
"bypass" power-frequency (and harmonic) ripple on 
power supplies, they are called "filter 
capacitors."  A few capacitors are used for 
signal coupling and to filter the AGC line (the 
latter create the AGC attack and release time 
constants), and some are used to make resonant 
("tuned") circuits in the RF and IF stages.

The vast majority of boatanchor cap problems are 
leaky oiled-paper power supply bypass caps.  As 
others have noted, these caps primarily fail by 
going leaky (and finally, very leaky -- i.e., 
shorted).  There are a number of "usual 
suspects," certain brands and models of molded 
plastic, oiled-paper capacitors known to be very 
prone to failure at this stage of a boatanchor's 
life.  Some radios (the Hammarlund SP-600 is one) 
have such problematic paper bypass caps that it 
is normal practice to replace all of them when 
the first one goes bad.  Other radios have just a 
few of the usual suspects, or use better paper 
caps that fail much less frequently, so mass 
replacement is generally not considered necessary.

B+ Bypass Capacitors:  As noted above, these 
typically begin to leak and then get more and 
more leaky.  As they draw more and more leakage 
current, they run hotter and hotter, often 
splitting their plastic encapsulation 
open.  However, they can be quite leaky and not 
cause obvious symptoms.  They do put a greater 
strain on the power supply, by drawing more 
current than normal, so it is well to find them 
and replace them.  So look first, for obvious 
physical damage (split cases, etc.).  Another 
good diagnostic tool is your fingers -- bypass 
capacitors should NEVER run warm -- any bypass 
cap than runs warm from internal heating (as 
opposed to being heated by nearby resistors, 
chassis, etc.) should be replaced.  (Be careful 
to avoid nearby high voltage.  Reach in with only 
one hand, and keep the other hand and all other 
body parts out of contact with any part of the 
radio.)  Look also at the resistors that feed B+ 
from the main supply to each bypass 
capacitor.  If any of these resistors appears to 
have been running hot, the associated bypass cap should be checked carefully.

Beyond those obvious symptoms of failure, your 
voltmeter and ohmmeter are the diagnostic tools 
of choice.  First, check the voltage on each 
bypass cap.  Any that show low voltage are 
suspect, unless you can find some other reason 
(e.g., the tube[s] fed by that B+ line are 
drawing more current than normal, or the feed 
resistor has increased in value).  Some have 
suggested testing bypass caps for leakage under 
high voltage.  This may assist in finding caps 
that are just beginning to leak, but I have never 
found it necessary -- if a B+ bypass capacitor 
shows a leakage resistance of less than 1 megohm, replace it.

I have said before that ceramic capacitors do a 
superior job of bypassing.  They are also much 
cheaper than film-and-foil capacitors and are 
easier to fit into the confined quarters of a 
boatanchor, so I strongly recommend using 1000 V, 
0.01 uF to 0.05 uF capacitors to replace any B+ 
bypass caps you change in boatanchors.

Filter Capacitors:  These are generally large 
electrolytic capacitors (10 uF and up).  You do 
not need any capacitor measurement equipment to 
know if these are bad.  First, look to see if the 
can is bulged or the vents are bulged or 
leaking.  If so, there is no need for further 
investigation -- replace the cap.  Next, power 
the radio up using a metered variac and see if it 
draws more than normal line current.  [Hint: 
Measure the line current when your radio is 
working properly, and use that as a baseline 
later when you suspect problems.]  Third, use 
your fingers -- if there is any perceptible 
self-heating, replace the cap.  (Observe the 
precautions noted above.)  Fourth, use an 
oscilloscope or AC voltmeter to measure the 
ripple voltage on the capacitor.  The input cap 
(first capacitor after the rectifier) may have as 
much as 10 V of AC on it, preferably less, when 
the radio is operating normally.  All other 
(downstream) filter caps should have less than 1 
V of AC, usually less than 0.1 V.  If you see 
more AC than this, the caps are 
questionable.  However, if other radio circuitry 
is drawing more current than normal (because of 
leaky bypass caps downstream, or a tube drawing 
excess current), you should fix that first and 
then re-check the electrolytic filter cap.  You 
do NOT need to test the ESR of filter caps -- 
excessive ESR will readily show up as excessive ripple.

Some very old radios, or equipment that uses very 
high voltages, may have oil-filled filter 
capacitors.  These rarely fail, but if they do, 
it will almost certainly be obvious on visual inspection.

Other Capacitors:  If the radio uses paper caps 
as interstage coupling capacitors, they should be 
replaced with modern film-and-foil types (as a 
matter of course, IMO, unless they are premium 
parts like Sprague Vitamin Q).  There will likely 
not be more than 2-4 of these in any 
radio.  There may also be mica or ceramic 
coupling caps in RF or IF stages.  These, and the 
capacitors used to make tuned circuits (mica, 
ceramic, or in rare instances glass or porcelain) 
do not typically show any pattern of failure.  If 
they do, the symptom will generally be loss of 
signal.  You should be able to locate the suspect 
tuned circuit using normal signal-tracing 
methods.  Once you have done so, a capacitance 
meter or bridge can be useful.  However, (1) you 
need to remove the cap to test it, (2) you never 
know what effect heating it and prying on it will 
have had, and (3) they cost next to nothing -- so 
you may as well just replace it.  (I have a 
pretty hard and fast rule -- once I unsolder a 
component, I don't put it back in the circuit 
even if tests good.  This practice has saved me 
countless hours of frustration over the 
years.  It does focus one on doing real 
troubleshooting and analysis -- which I advocate, 
in any event -- rather than pulling and testing 
every component in sight because one is too 
impatient to think logically.  Kind of like 
Blaise Pascal, who apologized for writing a long 
letter because, he said, he didn't have the time to write a short one.)

Good hunting,

Don


Copyright (c) 2012.  Not for redistribution