0
\$\begingroup\$

I found this circuit in an older post. I was thinking about using this in place of a contactor to break under load. The load would be at 500V 300A DC. There would be overcurrent protection in line with this. I do realize that this would be less efficient than a contactor and maybe more expensive. What am I missing? There seems to be little information out there about using an IGBT as a relay or contactor and I assume it's for a reason. What forums I have found people have said it would be too dificult but the explination stops there.

igbt circuit image

I'm currently using this IGBT (FZ600R12KE3) in a small setup switching 350V 30A about 20-30 time a day and have not seen any issues yet.

\$\endgroup\$
  • 1
    \$\begingroup\$ That looks bi-directional as though it was designed for AC switching. Do you not just want DC? \$\endgroup\$ – Transistor Jul 12 '18 at 20:58
  • \$\begingroup\$ How much line inductance are we talking about? \$\endgroup\$ – winny Jul 12 '18 at 22:09
  • \$\begingroup\$ @winny the line is less than 100M so I figure the inductance is in the uH range. \$\endgroup\$ – EEDAVEE Jul 13 '18 at 19:34
  • \$\begingroup\$ @Transistor It is DC but we could have current flowing in either direction. \$\endgroup\$ – EEDAVEE Jul 13 '18 at 19:40
  • \$\begingroup\$ Less than a joule to dissipate. Should be entirely doable. \$\endgroup\$ – winny Jul 13 '18 at 19:40
0
\$\begingroup\$

First of all, that bidirectional switch arrangement is inefficient. You require two completely floating powersupplies to realise the local Gate-Emitter power.

A more efficient solution would be the common emitter arrangement. A single float PSU can be used and then two isolated driver chips (hcpl-312T for instance).

enter image description here

As to is this a good idea? well yes, SSPC exist doing exactly this. Why this is considered difficult is producing a circuit that can do this twice... Doing it once is easy, especially if you do not mind the smell of burnt semiconductors. Doing it twice or more economically n-times is harder...

Switching power electronics all about loss management and adhering to its SOA. This is a forced-commutation scenario and thus there is a very high chance there will be over-voltage due to stray inductance wanting to maintain current flow. Likewise there is the SOA with regards to switching...

Usually in SSPC there are a large number in parallel to manage local die losses & die temperature. Equally there switching profile is controlled to mitigate the increased current due to interrupting inductors along with higher-voltage shunt paths

\$\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.