Re: MMC to be?

To: tesla-at-pupman-dot-com
Subject: Re: MMC to be?
From: "Reinhard Walter Buchner" <rw.buchner-at-verbund-dot-net> (by way of Terry Fritz <twf-at-verinet-dot-com>)
Date: Tue, 06 Apr 1999 11:47:50 -0600
Approved: twf-at-verinet-dot-com
Hello Adam,

>Original Poster: Adam <psycho-at-tradewind-dot-net>
>
>Ok, I'm thinking of building an MMC cap, but I'd like it if I could get
>some feedback on my plans from those who have built them before I
start.
>My NST is 9kV at 30mA, so my target capacitance is going to be around
>.009uF
>What I'm wondering is whether I can use snubber capacitors.  The ones
>I'm looking at are 3000VDC at .047uF.  This means I should only need 5
>in series in order to get .0094, which is pretty close.  The string
>would therefore be able to take 15000VDC, which is nearly 18% above the
>peak voltage of the transformer.  I'm certainly overworking them, but
>given that others have had success driving MMCs at peak AC=DC, I think
>I'll be okay.
>This means my TC cap will cost me about $20.00 total, and will be
>extremely convenient.
>Unfortunately, all of this seems a little too good to be true.
>Please feel free to ruin my dreams before they mature!  Comments would
>be greatly appreciated!


While I am the one who runs his MMC at peak AC = DC rating, I
would not advise you to start out that way. I started with 22 caps
in series and narrowed it all the way down to pAC=DC (12 in a
string). I did this (and no eq. resistors or cap safety gap) to see
if they would survive harsh treatment. Up to now, they do :o)).
However, I will add an additional three for a total of 15 in a
string for added safety. I take great care not to get a strike to
my primary coil. During different TC setups / tuning invoved, my
xformer gap has fired many a time. No ill effects. My MMC
doesn愒 seem to care. Your 15% safety margin might be a little
on the low side. Shoot for a 30-40% margin.

My other advice is forget about using a single row of caps
as a setup. My 30-odd nF cap consists of 6 rows of 12 caps.
Doing this many-string-thing helps keep the current per cap
string low. In your setup the string sees full primary current
and this can be quite high. (100愀 of amps easily). This might
overstress the caps. If you can get those caps cheap enough,
go for more in series and then parallel a few chains. It might
be cheaper to go for a lower nF cap (meaning each single cap
here). How are they built? Who is the manufacturer? How big
are they? I have killed a set of Philips X2 MKP caps, because
they were made with metalized plastic end plates. My MMC
caps use metal foil (!) as electrodes.  Also, be sure that those
snubber caps are made with PP and NOT PET. As Gary
already pointed out, the MMC is still in the "development stage".
So, you are on your own a lot more than with a rolled PE cap
design.

I would also try to keep the V/cap down low. A lot lower than
the 3kV design you are talking about. Remember, partial
discharges are real cap killers. While they might take a 3kV
steady state DC, no one can tell you if they will take the high
frequent 3kV AC for long.

Hope I didn愒 ruin too many a dream........ (;o))

Coiler greets from germany,
Reinhard