[TCML] Another MMC question.

[Finn Hammer]

Barton B. Anderson skrev:
snip
> It's funny, I argued the case of voltage on MMC's for a long time, but 
> it seemed that 90% of the TCML argued against it. Now I'm hearing just 
> the opposite. I feel vindicated in some respect, but also stupid for 
> throwing the TCML's old MMC chart up for reference.
snip

Bart, all
I don`t think the *right* voltage rating has been determined yet.
It is still a question of economics, need for durability or the 
individual coilers boldness, if you like.

The voltage rating may be determined on the basis of the VDC of VAC 
rating from the spec sheet.

If you adhere strictly to the VDC rating, the cap will end up as swiss 
cheese, if you use the VAC rating, the cap will end up over designed, 
and unnecessarily expensive.

Fact is, the caps are still conservatively rated from the manufacturer, 
as they should be.

Some of us will still remember how the caps shrunk in size, just after 
the MMC *craze* set in, and I feel strongly that the cap manufacturers 
were aware of the work done on this list at the time.

I have constructed a cap, which is in daily museum duty.

The buss voltage is 560Volts, so with DC resonant charging the bang 
voltage is 1120V. With 90% reversal, the ACpp voltage is  2128V.

The caps that I used: CD 940C12W1P5K-F Have an AC rating of 500V and a 
DC rating of 1200V, so

Vac= 1500
Vdc= 3600V
http://www.cde.com/catalogs/940C.pdf

This cap has been checked by removing random individual caps. These caps 
have then had the end caps cut off, has been unrolled, the layers of 
dielectric been separated and inspected for damage.

The quality of all layers was virgin, ie, no pinholes, no self healing 
telltales, no milky discoloring.

I therefore stand by this recommendation:

run your caps between Vac and Vdc, and you will be fine as long as 
current peak and rms is not exceeded.

I`ve been there and done that, so I realize that to the newcomer, it is 
hard to spec a cap. There is nothing to say about that than: study.

I recall an experiment I did back then:

whipped a coil together, and ran it with 2 caps in parallel:
They lasted 20 sec.
Then 8 caps in series/parallel. 4 strings of 2
They lasted about 15 min.
after that, 27 caps, 9 strings of 3
They never blew up

Once in a while it pays to get down and dirty and get hands on 
experience with things.

Hope this helps

Cheers, Finn Hammer