High voltage capacitor needed & questions

Jon, you made me think actually. I have a hobart cybertig with arc start, and fooled for a short time with HF start, and did not see any HF style arcs. I will try to spend more time to see if my HF system is broken.

See if you can find a place that services electronic air cleaners -- the kind that go into the furnace. Those have 6KV to 10KV caps in the .001 to .002 uF range. Two mfrs are Honeywell and Trion. I used to know a lot of guys at HON but that was a while ago now.

TV sets used to have 500 pF (.0005 uF) 20KV "doorknob" caps.

Here is a surplus place that has transmitting micas, 2000 pF (.002 uF) 6KV for $99. (Ow!)

Cornell-Dubilier (CDE) makes a mica paper cap, partnumber: KVX10S222K0T001 that is an .0022 uF 10KV cap. Here is a list of CDE distributors:

That's really backwards reasoning, and quite wrong too.

The freq is typically somewhere between 1 and 8 MHz. You can figure the RMS current from the voltage level and freq. That rating would be very conservative because it's basically like a Marconi spark xmtr rather than a CW-excited tank at the operative freq.

Wrong eh? I guess you didn't read the federal law section I included that mandates the marking of all non-PCB ballasts and capacitors manufactured domestically or abroad after June 30th 1978? Here, I'll include it again:

From:

Subpart C--Marking of PCBs and PCB Items

Sec. 761.40 Marking requirements.

(g) Each large low voltage capacitor, each small capacitor normally used in alternating current circuits, and each fluorescent light ballast manufactured (``manufactured'', for purposes of this sentence, means built) between July 1, 1978 and July 1, 1998 that do not contain PCBs shall be marked by the manufacturer at the time of manufacture with the statement, ``No PCBs''. The mark shall be of similar durability and readability as other marking that indicate electrical information, part numbers, or the manufacturer's name. For purposes of this paragraph marking requirement only is applicable to items built domestically or abroad after June 30, 1978.

So, once again, if you have an oil filled AC capacitor that is not marked "No PCBs" per federal law, it likely contains PCBs.

It would appear the the OP should take a look at the EPA info regarding decontamination for his shop.

Pete C.

As I read what is quoted below, it would appear that any manufactured *after* July 1 1998 (e.g. for about the past seven years) new capacitors and ballasts are no longer required to be so marked. (I presume that they are thinking that all are out of service by now -- or will be so by the time these recent capacitors look old enough to be questioned.

But beside that -- capacitors manufactured *before* June 30th

Unless it was manufactured after July 1, 1998, or possibly if it was manufactured before July 1 1978 and did not happen to use PCBs.

He should at least take steps to see whether what it leaking

When was the welder in which this capacitor was installed manufactured? That may have been covered earlier, but I do not remember it, and with my current news server, backtracking the threads often fails.

Enjoy, DoN.

Quite possible it expired and also quite possible manufacturers have continued to mark "No PCBs".

The marking law corresponds with the marking law. There is a strong probability that older non-marked capacitors do contain some level of PCBs. The gist of the law overall is that non marked capacitors are to be considered guilty until proven innocent as more likely than not they will be found guilty.

Not sure what percentage of capacitors manufactured prior to the ban were non-PCB. PCBs were pretty pervasive back then, likely only a year or so before the ban that any significant quantities of non-PCB capacitors were produced, also likely that those were marked No-PCBs as well since the law appears to have been in development from sometime in

Certainly a good idea to investigate further. Presumably the cap can be bagged and sent to a lab.

A specific date was not given, only a 70s vintage which would put it in the right timeframe for PCBs.

Pete C.

Yes, I stand corrected. The original post was from a man looking for a new lighting ballast for his shoplight, and not a capacitor from a welder.

I just dug up two old Cornell Dublier mica transmitting caps. Each is rated 0.004uF at 12kV and 22A @ 3MHz. Used in series they would provide the correct capacitance of 0.002uF and should be quite suitable for 6kV use. They would fit into an approximately 6" cube box each.

Remove the MUNGE from my email address to contact me directly for more info.

Just wanted to say something, possibly unrelated. To the person looking for mica caps, first check if they are indeed bad. They should show infinite resistance when unplugged. You can also try charging them with low voltage, and see if you get a spark in a screwdriver test.

The reason why I am saying this is that I also was dismayed to find that HF was not working in my welder. It turned out, simply, that it was wired for a remote arc starter. I rewired it for the local starter and it all works.

You're not going to get any sparks off a 0.002uF cap charged to low voltages.