Tube DIY Asylum

Do It Yourself (DIY) paradise for tube and SET project builders.

Return to Tube DIY Asylum

Message Sort: Post Order or Asylum Reverse Threaded

caps...rated at 660VAC, will they withstand 950VDC?!

Posted on October 31, 2001 at 20:09:06

Posts: 130
Joined: March 27, 2001
need to find some 12uF oil caps, and located some on ebay. they are rated at 660VAC, and i was curious if they would be stable under about 950VDC?!

any help would be appreciated!


Hide full thread outline!
Re: caps...rated at 660VAC, will they withstand 950VDC?!, posted on October 31, 2001 at 21:11:24

AC 660V is equvalent to DC 1.4*(2*660) V


NOT NECESSARILY!, posted on November 1, 2001 at 04:50:50

I've seen many caps where the dc voltage is spec'd is far lower than what you might think from the peak ac.

Best check with the manufacturer... i still don't know what to do!, posted on November 1, 2001 at 06:15:44

Posts: 130
Joined: March 27, 2001
basically...i was in line with what kjt said. i was figuring that they would be fine for operation, but now...vry has put doubt in my mind. the problem is they are NOS, and i don't know if i could hunt down the manufacturer (or if they even exist still). anyone else have an opinion? i'll play this like "ask the audience" on who wants to be a millionaire and go with the majority - and hope these things don't blow up and soak me with oil ;-)


Re: caps...rated at 660VAC, will they withstand 950VDC?!, posted on November 1, 2001 at 07:00:29

Why the extra 2 factor ?

Isn't Vmax = sqrt(2)*Vrms ?
Then we would have Vmax = 1.4*Vrms so for these caps Vmax = 1.4*660 = 924V hence, no, they can't be used for 950V.

I've got the 'same' problem, I've got loads of caps rated at 300V dc. I thought they were rated for AC when I bought them and now I'm not sure what to do with them..

Just like to be cautious with these things - I wouldn't like them to explode when there's people around.

There's two solutions: either sell them or use them at a much lower voltage.



problem finding cheap oil caps rated that high, posted on November 1, 2001 at 08:16:15

Posts: 130
Joined: March 27, 2001
i actually haven't purchased them yet (so that is good). they are on ebay. my problem is that i am trying to come up with a power supply for some 211s, and it is tricky to find the parts i need. i have a good design done up in duncan's power supply program...but now i'm trying to source the caps!


well...i guess i'll keep looking!


Re: problem finding cheap oil caps rated that high, posted on November 1, 2001 at 09:19:02

Check out the Axon Novas at Micheal Percy 1200v 20uf.


Re: caps...rated at 660VAC, will they withstand 950VDC?!, posted on November 1, 2001 at 09:26:03
dave slagle

Posts: 5211
Location: NYC
Joined: April 27, 2001
the "2 * 1.4 * ac" comes from the total max difference P-P of the rms AC waveform... so a cap with 660VAC RMS will actually have a max potential P-P difference of 1848V you have to realize that that potential only exists for a fraction of time, and not continuosly like DC so some caution is needed...

i have run some 370VAC caps at 1KV and they do tend to short, but thats right at the edge of that 2.8 number... i suspect the 660VAC caps should be OK for the voltages you want... make sure the whole thing is fused properly, and make sure the cans are attached to ground... i think the 1.4 number is a safe one.. it gives you a factor of 2 for "derating" and of course if possible do as vry says and check with the manufacturer....



Some additional cap rating thoughts..., posted on November 1, 2001 at 10:27:27

Only you can decide whether the risk is worth it, but...

1. I've exchanged emails with Eric Barbour on the topic of NOS PIO caps and he went out of his way to stress that the "oldies" are filled with PCB's (extremely carcinogenic) and that if one explodes/leaks you have essentially a toxic waste disposal problem in your listening room. Prolonged breathing of the fumes *will* cause health problems, so you have to rip out carpeting, destroy upholstered furniture, drapes, etc. He recommended never exceeding 80% of the rated voltage, allowing for power spikes/shorts. This leads to...

2. Allowing for power fluctuations - assuming you're building a 1200VDC supply, and use 120VAC wall voltage, a 10VAC spike may produce a 100VDC spike (simplifying for the purposes of discussion). Thus you need at least 1300VDC capability, plus 20% for safety now has you running caps rated at 1560VDC.

3. Another way to get hi voltage capability is to stack caps in series with resistors across them to equalize the voltage. The downside is this reduces the total capacitance. You must also take into account that most resistors have a limit for working voltage of 400V across them so you may want to use resistors in series too. By stacking the caps you can safely handle higher voltages.

4. I've had more than one amp repair tech recommend sticking with the conservative 1.4xVAC = VDC rating formula to keep you out of trouble. This consensus opinion comes from years of repairing blown amps so it may be a bit pessimistic. But then again...

Good luck and may the force be with you,



I don't think that's right., posted on November 1, 2001 at 11:13:09
Charlie G

Posts: 327
Joined: May 2, 2001

If you look at the P-P of an AC waveform, you're seeing the potential of one line as referenced to itself at a different point in time.

But a capacitors only experiences the voltage across the two terminals at the same instant.

For 660VAC RMS signal (with no DC bias) the potential to the reference at the waveform peaks is only 1.4 x 660.

Now, how that translates exactly to the AC ratings of caps, I am not sure, but I don't think an AC rated cap is meant to handle 2.8 x V, certainly not DC.

Charlie G


Let me see if I can confuse you some more...., posted on November 1, 2001 at 14:19:44
Pete Millett

Posts: 706
Location: Dallas, TX
Joined: April 6, 2000
I believe that the answer you seek is even more obscure.

I looked through datasheets for a number of "dual-rated" capacitors to see if there's a correlation bvetween the AC rating and the DC rating.

They're all over the map, from a low of VDC = 1.4x VAC, up to VDC = 3.0 x VAC Most seemed to be either close to 1.4 (especially smaller film caps) or closer to 2.2 (especially big motor-run caps).

I know I have some old GE paper/oil caps that are rated 660VAC/1500VDC.

I think part of the discrepency is that AC motor-run caps may be rated for the AC line voltage, but in operation the cap can get more than the AC line voltage imposed across them (due to the inductive nature of the motor) - one thing I read said up to 1.5x the line voltage. That might explain why a 660V motor run cap might actually be good for 990VAC, or 1500VDC. Maybe....

Also, one set of datasheets I was looking at showed a DC rating of 1.7x the AC rating, but also listed a DC "transient" rating of 2.3x the AC rating.

In general, I suspect that caps designed for motor-run applications are very rugged and over-rated. They have to withstand line surges and wicked inductive spikes when a motor is turned on and off. I'd guess that the 660V caps will be OK at 950V.

As someone already mentioned, I would be a little worried about old-stock caps, possibly containing PCB's and likely without a safety mechanism (newer caps are all marked "protected", which I think means that they fail in a less-than-explosive manner (open circuit)!).

I bought some 12uF/660VAC motor run caps from All Electronics (www. for $3.50 each - seems like a great price. These are modern polypropylene-in-oil caps, rated 660VAC, protected. I haven't run them over 750VDC yet (I'm planning on using them for an 845 amp someday at 800V-1kV). At least if they fail they won't hurt you.

Hope this helps....



Re: problem finding cheap oil caps rated that high, posted on November 1, 2001 at 20:29:03
Try these two places. They have some high voltage stuff, some new, some old. I imagine there are others like them.

Fair Radio
Surplus Sales of Nebraska



Page processed in 0.033 seconds.