[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Dead MMC



Original poster: "Terry Fritz" <twftesla-at-qwest-dot-net>

Hi Gary,

Actually, the dV/dT rating is not the one we want.  The dI/dT rating is the
key.  It takes more current to change the voltage on big caps.  You will
find it increases pleasantly as the value goes up :-))

http://www.cornell-dubilier-dot-com/film/9422000.htm

Not that these currents in the above or for 60Hz not 200kHz.  Geek caps are
good to 8 amps RMS at 350kHz not 13.5.  Better point that stuff out before
somebody goes and...

Also not that the ESR drops with big caps that helps to keep them cool but
the real ESR is higher.  At 350kHz it is 9.5mOhm not 5mOhm as in the chart.

One may also note the measly 500 volt AC rating :-))

Cheers,

	Terry


At 10:43 AM 1/27/2002 -0500, you wrote:
>Hi Winston:
>
>Thanks for sharing this.  I think we learn more by failures than by 
>successes, I'm sorry it had to be yours.
>
>In addition to the unknown nature of the film metalization noted by Terry, 
>caps also have a maximum dv/dt rating that basically determines the maximum 
>AC current that the cap can withstand.  If you look at a manufacturer's data 
>sheet, this dv/dt rating goes down drastically as the uF-rating of the cap 
>increases.  The 1.0 uF rating of the caps you used is at least 10 times what 
>is typically used in MMC's.  (I'm a little leery of the Geek's .15 uF caps 
>for this reason, but I guess you can't argue with success.)  I suspect that 
>your caps just couldn't support the high current demanded by this application.
>
>Regards, Gary Lau
>MA, USA
>
>Original poster: "Marry Krutsch by way of Terry Fritz <twftesla-at-qwest-dot-net>" 
><u236-at-earthlink-dot-net>
>
>Hi everybody! (again)
>
>	At Gary Lau's request, I will elaborate on the failures of my dying
>MMC.
>
>	My MMC is a 0.1 uF unit, using 90 630VDC 1uF caps that are supposedly
>metalized polypropolene.  I use 30 in series, then put three 30 unit
>banks in parallel for a grand total of 0.1 uf at 18KVDC.  I did most of
>what you arent supposed to do in building an MMC.  My caps are square,
>so I taped them together with fiberglass tape, then taped them to
>phenolic strips.  There is no space between caps.  It's possible that
>they flashed over, but with less than 500 volts on each cap, I doubt
>it.  I used 10 megaohm equalizing resistors on each cap, which all still
>test good, even on the failed caps.  All failures occured with my gap at
>around 3/32", or about 10kV.  I use 2 GE copier xfmrs in series.  If
>anyone has a C&H Sales catalog, they are in there.  Don't buy them! 
>Four out of 5 failed open circuit, so the lead to plate junctiones can't
>take the abuse.  These are not "big brand" caps.  They were made in
>Portugal by some company I've never heard of.
>
>	I should say that my MMC isn't really dead YET, but with 5 replacement
>caps left out of my original ten, and 5 hard minutes run time on my
>coil, I'm expecting about 3 minutes more out of my cap.  It then makes
>the magical transformation from a $100 capacitor to a $100 paperweight
>=:-0.
>
>	These might work for a higher voltage, lower current system, but I sure
>wouldn't buy any more myself.
>
>Hope everone else's coiling is going better than mine,
>	Winston
>
>