[Prev][Next][Index][Thread]
Re: UK MMC bulk order
>Does anyone have any definite findings either way for the need for
>equalising resistors in an MMC? I seem to remember seeing a comment
>that they were unlikely to help.
>
>
YES.
I made a small EMMC with 11 of the .056uf 1600v Panasonic caps. I just
soldered the leads together and took it over to the bench and hooked it
across a NST and a spark gap to see what would happen.
blasted it for about half a second. got interrupted.
Came back two days later to get something else off the bench. Brushed
against the middle of the string with the back of my hand (not on purpose)
and got a nasty surprise. Even though the end to end voltage was 0, some of
the caps seemed to have a full charge on them. I then broke out the meter
and found that 4 caps had over a 1000 volts left. others had varying
amounts. An interesting thing was that the polarity was random. That
convinced me. I got some 10M 1/2W carbon film resistors and put a resistor
across each cap. Then I charged the thing to 20kv and checked it again.
Nothing. The bleeder resistors cost $0.06 ea. 22 x .06 = $1.32, a very small
price to pay for safety.
It may not make any difference to performance, but for such a small cost and
piddling amount of trouble, it just does not make any sense to leave them
out. Also the random polarity of the residual charge left on the individual
caps of a "discharged string" might well cause trouble after time.
BTW my mmc consists of two strings of 11 for a total of 10nf. I put the
strings side by side inside a 1 1/2" O.D. polypropylene tube 12" long.
perfect fit. Potted with epoxy resin. A 1/4" X 1 1/2" brass bolt serves as a
stud on each end. Much smaller than the equivalent rolled poly in oil
version.
just My $.02
deano