3kj 'lyic cap bank/can crusher problem / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

3kj 'lyic cap bank/can crusher problem

Move Thread

LAN_403

ramses

Fri Mar 28 2008, 10:18PM

Registered Member #1208 Joined: Thu Jan 03 2008, 05:30PM
Location: Chesterland, OH
Posts: 154

well i made a 3kj cap bank with 12 2400uf, 450v 525v surge 'lyic caps. 4 serial banks of 3 in parallel
bussed together with doubled up aluminum sheet.

i had been running them without balancing resistors, voltage sharing wasn't too good. i recently strung them all in series, charged 1 cap to 5v through the whole chain, and measured the voltage of each cap. i ended up balancing all 4 banks of 3 to have close to the same average voltage. voltage sharing was better, and i was happy.

then i put 2 banks of 3 in series, and made a few full power shots through a can crushing coil with a clamping diode.

I put them back into the 4x3 and charged one bank of 3 to 5v through the whole string. voltage sharing sucked, so i put a 1Mo resistor across the bank, and voltage sharing was okay but not great. charged the whole thing to 1290v ( half power), and discharged. i was reading voltage across 1/2 the bank, since my multimeter doesn't go up past 1kv. after the discharge, i was reading -30v across the bank. i quickly shorted the bank, but no sparks and i was still reading -30v. voltage across the whole bank was 0v, but there was still ~60 volts distributed in the bank.

what the heck is going on here!?

additionally, for 1.5kj, the can didn't shrink much. i used 4 turns. should I use more or less turns.

thanks,
ramses

Dr. Slack

Sat Mar 29 2008, 07:42AM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

but there was still ~60 volts distributed in the bank.

what the heck is going on here!?

The balance between your series banks sucks. If the capacitance of each series block is different, then even if they are all charged to the same voltage, they will drop different voltages as the same charge flows through them all. It's not clear on what timescale you say charging voltage sharing sucks. If in the short term, then the Cs are indeed different (check the tolerance on the 'lytics, I would expect +/- 20%, would not be surprised at +80% -20%). If the long term then the leakage is different, and that's not specified as a value, only as a max. If they start at different voltages, they are likely to end up at different voltages.

additionally, for 1.5kj, the can didn't shrink much. i used 4 turns. should I use more or less turns

that's life with ltics I'm afraid, they're OK for a rail gun, but suck for coin and can crushing. I had two otherwise identical capacitors, 100uF 4kV pulse cap, and a 100uF 4kV lytic MMC. The first would happily neck a can right down to a finger diameter. The second would put ripples into the surface of the can that could just be seen if you got the light right.

BTW, after a few shots, the reverse protection diodes shorted, and they were rated 2000A surge, so make sure you over-design

Kizmo

Sat Mar 29 2008, 12:40PM

Registered Member #599 Joined: Thu Mar 22 2007, 07:40PM
Location: Northern Finland, Rovaniemi
Posts: 624

I have one question too:

I got some pulse discharge capacitors (30 x 2.5kV/40ÂµF)

Should i wire them in series-parallel to get 7.5kV / 133ÂµF or all in parallel to get 2.5kV / 1200ÂµF?

Tesladownunder

Sat Mar 29 2008, 02:00PM

Registered Member #10 Joined: Thu Feb 02 2006, 09:45AM
Location: Bunbury, Australia
Posts: 1424

Sam Barros made a 3kJ can crusher with 'lytics with mild crushing. At 3kJ my pulse cap crusher tears it in half and throws the ends out 20 feet. Think 1ms vs 30uS.

The voltage reversal will kill your 'lytic caps. At 3kj into 3 turns I get 40kA. About 60% of the voltage returns ie about 5kV in 3kV back out in the reverse direction. I was catching 12kA with diodes (actually SCR's wired as diodes) but the power was only small. You can't catch it with diodes unless you have a bank of pucks and that is more expensive than the caps in the first place.

I wouldn't bother trying with lytics. See the can crushing page on my site for details.

TDU

ramses

Sat Mar 29 2008, 02:33PM

Registered Member #1208 Joined: Thu Jan 03 2008, 05:30PM
Location: Chesterland, OH
Posts: 154

actually, the caps are +50%-10%.
clamping diode is a 3 phase bridge rectifier rated for 70A. i used the outputs so i got 2 diodes in series and 3 sets of those in parallel. Sam Barrow used an SCR switch, i use a contact switch. what can I do to get the banks to share evenly?

TDU, i wish i had those caps! As to the configuration, it depends on the application. personally, i would go series-parallel, just because i like massive pulses.
edit: i was talking to Kizmo. silly me!

ramses

Ps. TDU, i wish i had your caps, too. don't feel left out of my envy!

Myke

Sat Mar 29 2008, 04:38PM

Registered Member #540 Joined: Mon Feb 19 2007, 07:49PM
Location: MIT
Posts: 969

I think the the +50%-10% is the rating tolerances of the cap. In a power supply it's desirable to have a high capacitance.
Is this right?

ramses

Sat Mar 29 2008, 06:25PM

Registered Member #1208 Joined: Thu Jan 03 2008, 05:30PM
Location: Chesterland, OH
Posts: 154

that is correct, and i have no idea about the power supply scenario. they seem to be sharing more nicely for now.

ramses

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob