Best way to switch capacitor for coilgun

[smashedProton]

Hello all.  From the title it's pretty much self explanitory.  What is the most proper modern way to switch a giant capacitor into a big inductor.  I know to have a free wheeling diode to protect against reverse emf if I'm using a transistor.  Is an scr the best way to do it?  SCRs are awefully expensive...  What about a capacitor supressed relay?  Would the contacts still flash over and weld together?  That wouldnt be good...  Thanks in advance

[c4757p]

SCRs are awefully expensive...

Must be those solid gold ones...

[Zero999]

An SCR with a high enough current rating might be expensive - I haven't checked though.

[nctnico]

Did you look into a spark-gap?

edit: I looked a bit further. If the voltage is high enough then a triggerable spark gap is definitely the way to go. It is much more robust than a thryristor and costs very little.

[gotvolts]

What voltage and how much capacitance?

[fcb]

An SCR would be the ideal way to do it.  I would suggest an IGBT or two, but it wouldn't be any cheaper.

What is your peak capacitor, capacitor capacity, and estimated peak current?

[smashedProton]

It's going to be multi-stage, that is why I need the electronic switching.  Each capacitor is going to be at 350 volts and 3700 uF.  227 joules per capacitor. Very very fatal if I get shocked by it...  I understand the safety, and have played with it before.

[SeanB]

Look for old forklift SCR packs and use the hockey puck ones as the SCR's. With a low repetition rate you will not need the 20kG of heatsink they typically have bolted to them, just a pair of polished steel plates clamped across the ends with a thick copper wire bolted to each end. Those should handle the current with little problem provided you have enough gate current to switch it on very fast.

[staze]

IGBT's are what modern strobe camera flashes use... they probably could handle the current, and might be easier to control (not positive on that). Though, they're probably more expensive.

[G7PSK]

If you have really high currents and highish voltage ignitrons are very robust which is why they were the choice for spot welders before solid state came in. I used to have a Sciaky spot welder which had two in tha could switch several hundred amps at 415 volts over a fraction of a cycle that was on the input side the conductors were about 40mm by 15 mm solid copper bar. The output was 2 copper plates 20 mm thick with a hole in the middle and a slot.

[Psi]

Each capacitor is going to be at 350 volts and 3700 uF. 

*Psi takes one step back 

Exactly how many of those are you going to have in series?

[smashedProton]

Two in parallel for the first stage, then one for each subsequent stages.  A total of 6 capacitors.  The coilgun would be mounted on a tripod, so the capacitors wouldn't be anywhere near my face

[Marco]

Each capacitor is going to be at 350 volts and 3700 uF.

At those kind of voltages the currents for a discharge in the 10 msec range are actually not that high, less than a kA, which is pretty low for SCRs. You don't even need hockey puck or even stud devices since the repetition rate is low. A SCR in the 5 buck range can handle that.

Since you are talking about a 100 buck capacitor I don't see why you are calling the SCRs expensive ...

[smashedProton]

I got the caps for free.  I was also wrong about the price of the caps.

[Marco]

Still, do you consider say the  VS-70TPS12PBF expensive?