[ARC5] Capacitor Tests

[Richard Knoppow]

    There are three properties of _any_ capacitor that should be 
measured: 1, its capacitance value (of course); 2, its 
dissipation factor, also stated as equivalent series resistance 
(ESR); its leakage, which is its parallel resistance. 
Electrolytic capacitors tend to have high leakage even when new. 
This is what dissipates the charge in a relatively short time.
    Capacitance and ESR can be measured with several instruments 
but a capacitance bridge is the ideal one. Leakage is measured 
with a form of ohm meter which can put a relatively high voltage 
on the cap. It should be tested at something approaching its 
working voltage. Actually, a standard ohm meter can be used for 
low voltage caps.
    A non-electrolytic cap, i.e. a paper (in good condition), 
plastic, mica, ceramic cap will have very low leakage, plastic is 
probably the lowest. They will hold a charge almost infinitely. 
Electrolytics will not and, within limits, that is normal.
    Another characteristic of capacitors is dielectric absorption 
due to interfacial polarization of the dielectric. This is what 
causes some capacitors to recharge after being shorted out. It is 
of relatively little practical effect in conventional electronics 
but becomes important at very low frequencies and in timing 
circuits. Paper has relatively low dielectric absorption while 
mica is quite high. Most plastic dielectrics are also low. This 
is one reason high voltage transmitting caps are stored with the 
terminals shorted. It makes sure that you don't get a surprise 
when touching the terminals of a supposedly discharged cap. Some 
idea of the amount of dielectric absorption can be gotten by 
simply charging up a capacitor and discharging it by shorting the 
terminals (through a resistor if you don't want a bang) and after 
the terminals are shorted for a time measuring the voltage across 
them.

On 1/8/2019 2:55 PM, Robert  Eleazer wrote:
> Today I obtained yet another discarded stereo amp, with a nice 
> big power transformer.  I also took two big caps out of it, rated 
> 8200 uf at 71V.  Unfortunately there is not much else in there 
> that is of much practical use, other than the screws and three 78 
> series regulators.
> I have assumed in the past that if you charge up a cap to a 
> appreciable portion of its rated voltage and it is still 
> holding a large percentage of that charge days or weeks later it 
> is a good cap.  However I also have a nice new 
> unused electrolytic cap that will not hold a charge of 40VDC that 
> well.
> 
> Is this a good way to measure if electrolytic capacitors are healthy?
> Thanks,
> 
> Wayne
> WB5WSV


-- 
Richard Knoppow
1oldlens1 at ix.netcom.com
WB6KBL