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

Re: [TCML] MMC design

Hi Paul,

Don't go with the single string of 8. Your rms current handling would be good, but the voltage standoff is Questionable (which is only one step above Terrible). I hate to see coilers build wimpy cap banks. Voltage standoff is very important even with MMC's. Granted, some have managed to use rather low voltages with MMC's, but there have been enough coilers who have killed their caps due to a poor voltage margin.

In my opinion, if your going to build an MMC, build it to last and handle what you throw at it. Standard derating for cap voltage is 2.5 x Vp. So in your case, sqrt(2) x 9000 = 12.728kV x 2.5 = 32kVp.

The 2 strings with 16 caps per string for .013uF rated at 32kV is excellent. That is very robust and hits your LTR value right where you want it. It's only 32 caps so it's not expensive. Since the old days, tank caps have always been the big dollar item. MMC's have allowed us to cut that cost down to something reasonable.

Take care,

dunnsept wrote:
I'm going through some of the MMC calculators on the net trying to figure this stuff out.
I have a 9kv NST and LTR capacitance is 12.5nF
output of the NST is ~13kv p-p   (close enough)
so, assuming 0.1 2kv caps
1 string of 8 = 16kv and 13nf
2 strings of 16 = 32kv and 13nf

is 16kv "high enough" above the NST?
is it just a waste to go with 2 strings of 16, should I, in this scenario, stick with 1 string of 8 and call it good?
if the caps were 0.047 and 2kv, then 2 strings of 8 = 16kv and 11.75nf
or 3 strings of 11 at 22kv and 12.82nF

which is more important in designing the MMC?
is it more important to be as close as possible to the LTR value or have enough voltage "padding" above the xmfr? is there some other factor in here to use (like the discharge energy or something?)

Paul (I didn't expect the Spanish Inquisition) in Michigan...

Tesla mailing list