I have a 3-phase bridge rectifier which uses IGBT's shown in the image below. The circuit works perfectly and the voltage on the output (unfiltered) is exactly as expected. However the gates are referenced to the collector. Now in order to turn on an IGBT the gate voltage must be greater than gate-emitter voltage.

When I reference them to the emitter the circuit does not work but does work when referenced to collector.

Can anyone clarify why this is??


enter image description here

  • \$\begingroup\$ Do you really have P-channel IGBTs? \$\endgroup\$ – winny May 3 '18 at 14:40
  • \$\begingroup\$ No the NMOS PNP combo was just constructed in LTSpice to implement a generic IGBT as this is what they consist of. \$\endgroup\$ – MXG123 May 3 '18 at 15:13
  • \$\begingroup\$ Oh! But think about Vgs for the MOS. \$\endgroup\$ – winny May 3 '18 at 15:23
  • \$\begingroup\$ Same issue. Vgs is required to turn on an nmos but in this scenario it would be Vgd thats connected instead of Vgs. It should be the other way round. Is there a better way to model an IGBT in spice so that turn on is referenced to gate-emitter and not gate-collector? \$\endgroup\$ – MXG123 May 3 '18 at 15:36
  • 1
    \$\begingroup\$ Exactly, that’s why your gate drive reference is wrong. \$\endgroup\$ – winny May 3 '18 at 15:53

As it turns out the pulses were out of sync, they were shifted over. This effectively created a short between gate and collector. It turns out that setup automatically triggers each device correctly and at the right time as the emitter falls below gate/collector by Vthreshold for that section. This maybe useful as an autonomous rectifier without any control.

I wanted control over the gate voltage so with the right pulse timings I can reference it to the emitter and works as expected.


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.