Leaking Caps

[C Moore]

Pardon me if I sound like a dolt, but.....
I know some of you guys do A LOT of repairs.
I have never actually measured a (for example) Mallory 150 that was bad from a DC Leakage standpoint.
Any of you guys have a coupling cap that you know is bad, and want to shove it in an envelope, and send it to me.? Just so I can compare it to a good/similar cap.
I understand (or at least I think I do) the procedure of lifting the low voltage side and tying it to a series 1M resistor to ground.
Just, kind of, wanting to see this for myself.
Would be way more than happy to send somebody a free cup of Starbucks (or whatever) for their efforts.
Thank You

[Firestorm]

Mallory 150s don't really go bad that way because of their construction: metallized film. If they pinhole, the metal side oxidizes and they stop conducting. Some old Fender caps were notorious (Astrons especially), but those were film and foil. Unfortunately, when I pronounce a cap dead and replace it, I always give it back, so I don't have any of the worst examples. Let me see if I have something I dragged out of one of mine.

[tubeswell]

Sorry M8 I chuck away all my bad caps. If I remember next time I find one I'll set it aside just for you.

[C Moore]

Mallory 150s don't really go bad that way because of their construction: metallized film. If they pinhole, the metal side oxidizes and they stop conducting. Some old Fender caps were notorious (Astrons especially), but those were film and foil. Unfortunately, when I pronounce a cap dead and replace it, I always give it back, so I don't have any of the worst examples. Let me see if I have something I dragged out of one of mine.

Yeah, the 150 was just an example. But ANY coupling cap will do.
Brown Blob
Astron
etc etc
Thanks

[Firestorm]

Mallory 150s don't really go bad that way because of their construction: metallized film. If they pinhole, the metal side oxidizes and they stop conducting. Some old Fender caps were notorious (Astrons especially), but those were film and foil. Unfortunately, when I pronounce a cap dead and replace it, I always give it back, so I don't have any of the worst examples. Let me see if I have something I dragged out of one of mine.

Yeah, the 150 was just an example. But ANY coupling cap will do.
Brown Blob
Astron
etc etc
Thanks

I'll see if I have something. Caps are hard to test in circuit (even if you lift one end) because they charge and look like they're conducting DC for awhile. (Well, that's how caps work). I'm sure I yanked some out over the years before I figured that out.

[jelle]

I have yet to encounter a polyester cap that conducts DC. But I have seen many silver Mica caps. the worst one measured 11K with my digital handheld meter.

An I have seen some NOS 6PS caps that did not pass signal.

[jjman]

I found 7 old Astrons out of 8 in a 1959 Ampeg that leaked various amounts of DC. Lift the lower-DC-side, attached + side of voltmeter to the lifted lead, (-) meter lead to ground, watch voltage during warm-up and charge up. Should be less than 1vdv, perhaps much less, after charge up. Don't lift any other leads, especially those grid leak resistor/pot leads that will often be in the same eyelet/terminal. One should really know the circuit for this testing.

[jelle]

More often than not, the board is conductive. I take the caps out and test them on a Sencore LC53 at the voltage the cap would see in the amp.

Very few polyester caps go bad. The Astron paper in oil ones...yes they eventually start to leak DC.

[C Moore]

I found 7 old Astrons out of 8 in a 1959 Ampeg that leaked various amounts of DC. Lift the lower-DC-side, attached + side of voltmeter to the lifted lead, (-) meter lead to ground, watch voltage during warm-up and charge up. Should be less than 1vdv, perhaps much less, after charge up. Don't lift any other leads, especially those grid leak resistor/pot leads that will often be in the same eyelet/terminal. One should really know the circuit for this testing.

Yeah this is kind of what I would like to see.
Though, I have always read it is "better" to put a 1M resistor in series with the lifted lead of the cap. Then read, from ground, across the resistor.
Would just like a chance to see it for real...
thanks again

[Prairie Dawg]

I use an insulation tester, a/k/a megger, but you don't need the old crank type units unless you need some exercise. My present one is an IRT-2 that contains a small power supply that can deliver 1 kv at very low current and read up to 1000 megohms resistance. I also have a Shenzen Victor that can read up to 2000 megohms resistance. If I get any significant reading that is less than infinity at 500v the cap gets replaced.

Old Astrons and other paper and foil caps usually are quite leaky aside from being pretty old by now. I once bought a cabinet full of Mallorys-brown bakelite tubes as I recall-they were all unused and all of them ended up in the trash because they all leaked. On the other hand, military grade caps like Aerovox in ceramic tubes are quite good, even if they are paper and foil.

I have also built a direct reading capacitor leakage meter to do the same thing in a different way, as the Sencore LC series devices, nice as they are, are a little spendy for me.

[greekie]

I bought a whole lot of items once. Got amongst other things a package of Hunts mouldseal caps. Every single one was leaky and/or cracked in its moulded case, and got tossed.

[rdjones]

If you measure any leakage on a low voltage hand held ohm meter (even a Fluke) then the cap is bad leaky.
The "megger' and Sencore type LCRs are using a high voltage supply to test the cap at or even above the cap's rated voltage which is where the insulation is going to break down.

Note that cathode bypass and other low voltage locations are much more tolerant to leaky caps due to both low voltage and the relatively low parallel resistance.

RedDog Steve