ESR Meter Recommendations

[drlegendre .]

"How are
folks testing the very large electrolytics in the 25,000, 80,000 or even
100,000 uF range that are encountered in the power supplies of some of
these old machines?"

The method I use is pretty close to what you mention, though I lack an
electronic load. Here are some thoughts...

For measuring capacity - on these rather large-value caps, it's quite
acceptable to time their discharge rate through a known, stable resistance
of some reasonable value based on the marked capacitance. Make several runs
and average them out. Calculating value is then simple arithmetic - C=T/R.

For leakage, I use a variable voltage-regulated PSU with current-viewing
resistor in series. At what percent of marked voltage does leakage clear
the noise? At what voltage does it become significant and at what point
does it exceed allowable? All electrolytics leak, and the larger value the
part, the more expected leakage. Check a datasheet, or just use your
experience and make educated guesses - hint: how much power (as in heat
rejection) is the cap dissipating at working voltages vs. its marked value
and physical size?

And as you suggest, watching supply ripple on the scope, with the cap under
some nominal load, is a tried & true method. It's usually my first step, if
I think a cap might be in trouble.

(I was certain that I'd have to replace the 30 yr-old 95,000uF part in my
Altair.. but the dang thing proofed out just fine. Ran it for a day at 10%
over working voltage, and leakage sat right where it was when I began the
test. )

On Wed, Sep 30, 2015 at 1:35 PM, Chris Elmquist <chrise at pobox.com> wrote:

> On Wednesday (09/30/2015 at 10:54AM -0700), John Robertson wrote:
> >
> > As for testing large capacitors, about the highest value that gives
> useful
> > readings is around 10,000ufd. Larger than that and the ESR is too close
> to
> > zero ohms unless the cap is really bad...
>
> And that seems to be the usual situation for most ESR meters.  How are
> folks testing the very large electrolytics in the 25,000, 80,000 or even
> 100,000 uF range that are encountered in the power supplies of some of
> these old machines?
>
> One approach I use has been to isolate the supply and then fire it up
> with an electronic load while looking at voltage sag and ripple on a
> scope while that load is swept from min to max capability of the supply.
>
> There is also the Sencore LC53 "Z Meter" which can test large caps up
> to 200,000 uF out of circuit but these are typically a $400+ instrument
> on the used market.
>
> http://bama.edebris.com/download/sencore/lc53/LC53%20OPS.pdf
>
> Chris
> --
> Chris Elmquist NØJCF
>
>