2
\$\begingroup\$

I'm working on a coilgun and I need to discharge a capacitor bank through a coil to create a strong magnetic field. My coil has a resistance of 266 milliohms and when fully charged, the capacitors are at 200V with a capacitance of 4700uF total. This will result in a peak current of 750A (ignoring wire resistance and ESR of the capacitors). Up until now I have just touched the wires together to close the circuit but I need to time the switching in a controlled way. My problem is that I have no clue what component to use to switch ~700A 200V. Of course the 700A peak is only for a very short amount of time and the whole pulse should last around a millisecond.

I have tried using 5x irfp260n MOSFETS in parallel, each MOSFET should be able to handle 200A peak so it would total 1000A when connected in parallel. Unfortunately one of the MOSFETS always dies after a couple shots (I did take care of the reverse voltage generated by the coil). I also am aware that connecting MOSFETS in parallel is generally not a good practice.

I have looked into thyristors but the inability to switch them off through the gate makes them useless for me. "Gate turn-off thyristors" looked interesting but I could only find really expensive ones (in order of hunderds of euros). IGBTs are known for being able to handle more current but I failed to find any that were rated for this kind of current.

Does anyone know a solution to switch 700A 200VDC for 1 millisecond?

\$\endgroup\$
8
  • \$\begingroup\$ ”ignoring wire resistance and ESR of the capacitors” ESR and ESL will be dominating your circuit. Don’t ignore them. Show Vgs, Vds and Id waveforms for your MOSFETs. \$\endgroup\$ – winny May 13 at 20:22
  • \$\begingroup\$ IGBT like this one? (about $20 one-off) IGBTs have advantage over MOSFETs at higher voltages. You might want to read up on driving IGBTs. \$\endgroup\$ – Spehro Pefhany May 13 at 20:35
  • \$\begingroup\$ @winny I don't think ESR is dominating my circuit, I don't exactly know the ESR of my capacitors but since 10 of them are connected in parallel the resistance decreases by a factor 10, also the wires carrying the current to the coil are thicker and much shorter than the coil itself. Lower voltage (60V) measurements with my scope support this hypothesis. \$\endgroup\$ – Gustav May 13 at 20:38
  • \$\begingroup\$ @SpehroPefhany Thank you! This one indeed seems really promising, I'll definitly look into that one. \$\endgroup\$ – Gustav May 13 at 20:50
  • \$\begingroup\$ Then you have good capacitors. What about ESL? Any progress on the oscillograms? \$\endgroup\$ – winny May 13 at 21:16
1
\$\begingroup\$

You are working with 200V power source, and irfp260n has absolute maximum drain-source voltage 200V. So this mosfet is not suitable for the task, you have to have some margin. And you are not even considering transients and ringing, which is inevitable. Use mosfets with much higher voltage rating. Like 500V. And consider ringing, because you are talking about coil.

\$\endgroup\$
0
\$\begingroup\$

Very fast turn-on is essential here so you need a very powerful gate drive. You don't care so much about ringing etc. So go ahead and pump 10A into each of the gates.

If this is still too slow to protect your FETs, maybe check out some GaN parts which have much lower gate charge.

Another step you can take to slightly delay the current surge until after full turn on is to add a tiny series inductance with the caps. Around 100nH

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.