That's why I like my old solid state Sencore high voltage capacitor analyzer and my even older tube type Semcor (not Sencore) high voltage capacitor tester's. I have seen this many times in the past. Someone sent something in for repair and said they tested all the large bulk filter capacitors and they tested well within spec. I test them at their rated voltage and they fail. I wouldn't say it's a common failure but it does happen. What I have found over the years is that if a aluminum electrolytic capacitor test over +20% or higher capacitance than what is printed on it, it likely has high leakage current. You don't even need a capacitor analyzer / tester to test high voltage caps. As long as you have a high voltage power supply the only other thing you need is a series current limiting resistor. Just hook up the cap in series with the resistor and monitor the current. Good rule of thumb for maximum leakage current is .01 x C x V = maximum leakage current in uA. That's .01 times capacitance times voltage. In the case of this capacitor that would be .01 x 710 x 200 = 1,420uA or 1.42mA maximum leakage current.

The capacity tester is really spectacular. It reminds me of the Saul Steinberg cartoon where he explains which of the cockpit instruments is for coca cola. If you do not have such fancy equipment the high voltage test could be made with a voltage source, like a variac, diode, resistor and voltmeter. If you have it you have to make use of it, it really looks nice.

Can’t wait to steal snippets from this supply. I love seeing what circuits test equipment use

Whoa! Your vintage capacitor tester is an absolute thing of beauty and joy for ever, even more so than the HP supply.

That's a really unusual fault! I have test hundreds of this type of capacitor and, as you said, very few are bad. I always test for value, ESR and leakage at rated working voltage. Regards, David

Just turned mine on the other day after having not used it for a couple of years. A little smoke came out, so I opened it up. Tuns out R11 is just dissipating 3 or so watts all the time and had a piece of lint on it. Cleaned it off and it's been running fine.

Note that when that large cap is installed, it covers the markings on both the cap and the board. I'm sure that guy marked it before removing it.

Amazing that you even thought to check for a high voltage problem with the capacitor...

in hinsight the board and capacitors are marked, but mounted both will not be visible. so marking the cap was a good idea I think.

As a technician who has specialized in the repair of commercial photographic flash units, which typically have main capacitors in the 330-500Vdc range, the fault is quite common. High voltage capacitors can even fail like this one did and "self heal" to continue on with a long life afterwards. Self healing is a design feature of certain capacitor types and brands.

I always test capacity, ESR, and leakage of capacitors before using. Just like your ESR chart, I have a similar one for acceptable leakage current. Unfortunately the leakage test I do must be manually performed with an external resistor and voltmeter, because I don’t have enough work to justify a dedicated test unit.

if you get electrocuted with it, you'll be like Marv in Home Alone 2

Had a few of those old Sprague capacitors that had long term failure. Very variable, the one rotted away, but the one next to it from the same batch was absolutely perfect, almost meeting spec even before any reforming. Reformed at operating voltage via a resistor for a day, and then left it open circuit for a month, where it had dropped less than 1% of the voltage due to leakage. OK it was a 15000 uF 6.3V rated capacitor, charged to 6V, but impressive after close to 4 decades of non conditioned storage. Still have one or twp of the 100uF 6V rated ones, that are 1.5in long, and 0.5in diameter, which I thought was a typo in the data entry, but they were those dimensions. All failed, capacitance all over the place, and ESR in the dozens of ohms ranges, but so big, even for the 1960's era. Kept because I might get a tube amp, and hide a modern one in the case to use as cathode bypass capacitors.

Tradition has it that Hewlett Packard capacitors never fail. Sprague naturally.

Incredably cool old school tester man

I bought some cheap and dangerous high voltage generators from china precisely because of stuff that malfunction only at "high voltages", whether they are capacitors, resistors diodes, transistors, switches, etc. It has been really helpful

I used to work for a company that built fault detectors and sensors that were designed for use on medium voltage lines - which is anything from2400 to almost 70,000 VAC. We tested a lot of capacitors from vendors who assured us they were good at those voltages only to find they failed (sometimes spectacularly) when in that environment.

This HP supply reminds me of a series of linear DC supplies from a company called “Power Designs”. Probably about the same vintage, maybe a little older (they feature similarly hand drawn PCBs, but were new enough to be using 2N3055 output transistors and DIP op-amps to set the voltage/current levels). You can still find a lot of them used, they’re commonly a pretty generous voltage range with more limited current. I picked up a dual model that has symmetrical 50V/500mA, I think I paid $80 shipped. They’re seriously well built and very quiet. My unit had an asset tag from a nuclear research lab in California, I believe Power Designs was originally a California company.

Interesting, wonder if you could reform the cap by slowly raising the voltage. If you can't find a replacement cap , cut open the old cap, replace the guts with several radial Nichicons and glue it back together.

Encountered the same a few times now 🙃

I've always wondered about these new capacitor testers using low voltage to test leakage and ESR. I have a couple and enjoy them but always wonder. I do have an old Eico capacitor tester that can test leakage at higher voltages and will pull it out occasionally, but never had any positive results like you did. I'll keep the old tester!

Have a theory on that shorting cap. The previous owner tried to re-form the cap and brought the voltage up too fast and shorted it hence the shoring at voltage 'X' I have done this myself to an NOS computer grade cap blew it a 100v. It reads perfect ESR, capacitance but shorts when voltage goes near 100 d'ho.

With the capacitor installed, he couldn't see the + on it...

So no carbon track detected since it would be in parallel with some part of the capacitor and not in series? Interesting.

I have seen those caps get leaky at higher voltage but never arcing. Must be a defect in the insulator separating the plates or something like that. Maybe it got over voltaged at some point and that blew through the insulation?

I always admired the old meter's capabilities and design. You should open up the old cap and look for the arc path. What is the vintage of that PS?

Hi there, I in your movie the discharge "thing" and love it. Can you please tell me more about it? Tnx in advance

Thank, interesting. What is the checker condensator please?

Take the Cap apart. Let's see what the actual fault looks like.

Cool capacitor tester, what is the make and model?

What make and model is that fabulous capacitor tester of yours?

What is the make and model of that beautiful, old-school capacitor tester?

Looks so good and IT IS so Bad.

This type capacitor are pretty expensive, i repaire big switching power supply with through hole capacitor because capacitor with screw cost me 80$, few through hole capacitor in paralelle 10$ with better ESR and life!

As @TheDefpom would say "It's always the capacitor", but I don't know if he's seen this type of fault before!

I wonder if Mr. Carlson's low voltage leakage detector would have sensed that. He claims it's as good as testing them at working voltage, so maybe.

What model is the vintage tester you used?

never say never,....

Sounds like it may have moisture in it, you could bake it at 90C or so for an hour or two. (*) ... ... (*) Advice from random strangers on the internet may be detrimental to your health. Proceed at your own risk. Don't forget to video it in case it explodes.

I’ve thought about this before, why couldn’t you use a smaller more modern cap of the same value and voltage rating? What is so special with these “computer grade” caps?

Why assume it was a he who took out the cap😂😂😂😂😂

Correct me if I'm wrong. I'm under the impression that ESR is very important for switching power supplies. Especially at rated switching frequency. ESR has nothing to do with a leaky cap. A linear PS could care less about ESR, as long as the cap wasn't leaky. If there was a good example of a leaky cap, this would be a good one. Just as an FYI Antique Wireless has a cap tester that will provide voltages from about 10v to 600v.

Talk en trash.