testing capacitors

[xcortes]

When building my first gainclone I blew (literally) a couple of psu's. The diodes are easy to replace but the caps are harder for me to get living in Mexico.

I'd like to know if they still work so that I can use them for other amps (I have the other parts, including transformers, waiting to buid a pair of active woofers to go with a pair of 3 litre Jordans).

How can I know whether the caps are damaged or not?

BTW: my first gainclone works perfectly now. I didn't kill myself or set the house in fire in the process of building it!

[jackinnj]

I have over 40 years of DIY'ing -- in my youth in the physics lab (even though I was a chem major) , then mostly ham radio equipment, instrumentation, audiophile stuff and programming -- embarrassingly I blew up the caps on several GC's I built by soldering them incorrectly ! Welcome to the club, it proves that we are mortal !

FWIW I use 1000u/63VDC electrolytics bypassed with 10N/100V mylars on the supply rails. All the other bypass caps are 100V. I have used a number of different input caps including 100V EPCOS silver caps, mylar caps from Mouser (B.C. Components) and some exotic General Electric and Corning caps.

[falcott]

Originally posted by xcortes
How can I know whether the caps are damaged or not?

I use a multimeter that measures capacitance. Mine measures from 40nF to 100uF, so I cannot check things like 1000uF decoupling caps...

[Mr Evil]

You can put the 1000uF cap in series with a < 100uF one and measure the difference from the < 100uF alone, since the total will be less than 100 and thus within you meter's range. The value of the 1000uF cap can be calculated from that.

[johnnyx]

The caps are probably ok.

It's not always useful to measure capacitance to test for failure. I use three capacitance meters; one on my Fluke DMM, and two stand-alone capacitance meters.
When a capacitor fails it is usually due to leaking electrolyte or evaporation due to age. Measurements of capacitance show a small change, but within tolerance, for the DMM and one of the cap meters. The other meter shows a greater reduction of capacitance for faulty caps. So only one of the three shows a faulty cap as being faulty. The others would show it as being ok.
It depends how they measure it, a leaky cap has a higher impedence (esr etc), which doesn't always show-up in the capacitance reading.
Cyril Bateman designed a meter in Electronics World that would show faulty caps, but it isn't simple, partly because losses are frequency dependant.

[thomas997]

Build an ESR meter then.

Here is one example:

http://www.qsl.net/iz7ath/web/02_brew/15_lab/06_esr/

[falcott]

And another:

http://www.siliconchip.com.au/cms/A_100895/article.html

[thomas997]

Originally posted by falcott
And another:

http://www.siliconchip.com.au/cms/A_100895/article.html

Thanks!


Oh and dont read this:
[SIZE=0.6]If you dont happen to have 9$ to spend, use a certain search engines cache feature to reveal the full article.[/SIZE]

[falcott]

Originally posted by thomas997
Oh and dont read this:
[SIZE=0.6]If you dont happen to have 9$ to spend, use a certain search engines cache feature to reveal the full article.[/SIZE]

Oops... I found this by doing a search with a certain search engine, and didn't realise that they charged for viewing more than the first paragraph or 2. Glad you found a way to get the whole deal.

(If anyone asks, I never read a thing....)

[falcott]

By the way, for those interested, part 2 of the above article is:

http://www.siliconchip.com.au/cms/A_101187/article.html

with a couple of interesting links to the author's website at the bottom of the article.