MMC for a 32kvac load

[Rowdy]

Ok I have 32kvac HV supply rated at 180mA output. I know I need an MMC for 12-15nF. But it needs to be portable enough to travel with. Was looking at these but no specs on them other than what you see, not sure I
could go by Cornell specs it its compared to.

https://cdn1.bigcommerce.com/server800/64f33/content/Graphics/1.5u-2kv-mmc-2b.gif

Any ideas guys?
Also the HV transformer is a gapped transformer and has OVLD protection, if shorting the output it trips the OVLD in a couple of seconds. 100K 250W resistor works ok but still trips OVLD. I need a capacitive load.

[Weston]

A MMC for that rating is going to be pretty expensive / large, you are looking at ~ 15J of energy while the standard CDE 942C series caps people like to use in MMC stores ~ 0.2j without any derating. The high voltage does not make anything easier.

A MMC might be more compact than a single commercial capacitor because you can push the ratings on the capacitors a bit more.

Is this for a tesla coil? You mention a gapped transformer which is pretty unusual and you also mention some sort of resistive balast.

[Rowdy]

Hey Weston, yeah that is what I thought but didn't hurt to ask for backup. It is a Hi pot used to test bucket trucks. We have a cap at work we use to calibrate the current meter with, but I travel around alot and taking a huge cap just isn't viable. Resistors work ok for lower ranges but the problem is the 180mA range and the OVLD
trips, you have be very quick to get the readings. Just trying to find an easier way, and less stressful on the variac that feeds the HV transformer.

[Mads Barnkob]

Ok I have 32kvac HV supply rated at 180mA output. I know I need an MMC for 12-15nF. But it needs to be portable enough to travel with. Was looking at these but no specs on them other than what you see, not sure I
could go by Cornell specs it its compared to.

https://cdn1.bigcommerce.com/server800/64f33/content/Graphics/1.5u-2kv-mmc-2b.gif

Any ideas guys?
Also the HV transformer is a gapped transformer and has OVLD protection, if shorting the output it trips the OVLD in a couple of seconds. 100K 250W resistor works ok but still trips OVLD. I need a capacitive load.

Getting 15 nF from a string of 1.5uF/2kV capacitors, you need to put 100 in series for 15nF/200kV. You properly only need a voltage rating of 100kV to still have some good overhead (32kVAC ~64kVDC+33%)

So if you could find a 0.75uF/2kV capacitor, that would make you able to build the same with just 50 capacitors in a string, along with 10M balance resistors on each cap, that it properly a viable solution, if there is no big requirements on the current capability.

You can easily play around with MMC configurations in my MMC calculator: http://kaizerpowerelectronics.dk/calculators/mmc-calculator/

[davekni]

How about a string of 66 of PHE845VW7100MR06L2 1uF 760VAC (1.5kVDC)?  They are rated for X1 (across-line) use, but not for X2 (line-to-ground) use.  As long as failure doesn't put human life at risk, these should be fine.  You'd get 50kVAC rating for the string.  Cost is ~$5 each, so ~$330, plus the cost of 10meg balancing resistors and whatever you use to mount and interconnect them.

[Rowdy]

Ok this is probably my ignorance about film caps, but lets do this....

https://www.mouser.com/datasheet/2/88/942C-22274.pdf
942C20P15K-F
Now this cap is rated for 2000Vdc and 500Vac....and depending on how it is made it could be 1000Vdc and
400Vac. Now most Tesla coils i have seen the specs for , are based on the DC rating on a primary side of
the tank circuit. This is AC not DC.   
I also realize at this point, my own problem that started this thread is gonna be expensive and probably not so
portable to travel with. That's ok, the company I work for needs to work this out and I will continue to use resistors as the other service techs.
But the problem above still remains, DC rating. Why not the AC rating of the cap? Maybe this needs to be moved to another thread?

[Rowdy]

Also thanks guys for the replies !!

[davekni]

"But the problem above still remains, DC rating. Why not the AC rating of the cap? Maybe this needs to be moved to another thread?"

That is an excellent question without a simple answer.

Cap AC rating is generally RMS.  Taking into consideration the duty cycle of DRSSTC use, RMS voltage is generally well below cap's rating when the peak voltage hits the cap's DC voltage rating.  However, cap vendors generally test AC with sine waves, usually 50 or 60Hz.  With DRSSTC waveforms, caps can still fail even when all parameters are within spec limits.  That's especially true for caps with large differences between DC and AC voltages, such as the WIMA FKP1 22nF 6kVDC 700VAC caps I'd used initially.  Internal electric field stress points along foil edges will slightly break down the PP dielectric at the edge until enough charge has moved to make the field more uniform.  For DC with limited AC on top, that charge stays put, as the field doesn't change too much.  With even low-duty-cycle AC, voltage reversal forces repeated slight dielectric breakdown at foil edges.  Eventually the dielectric fails.

After the FKP1 caps failed, I built a new MMC using 0.33uF 630VAC 1200VDC Chinese induction cooker caps.  First I ran stress tests on two parts, running DRSSTC waveforms at progressively higher voltages.  Had to reach +-2kV to get degradation.  This was at low ~1% duty cycle, for several days solid.  By +-2kV the caps were getting rather warm even at 1% duty cycle, which may have contributed to their failure.  (No active cooling.)

That's the other factor for DRSSTC use:  A few days of operation is all a coil is likely to get in it's lifetime (or it's owner's lifetime).  Cap ratings are designed for years of continuous use.  So the reliability decrease of going a ways over AC rating is tolerable for something being ran for a few minutes at a time on a few days each year.