The Semikron IGBT thread
Steve Conner, Sun Jun 15 2014, 06:33PMAfter blowing up a couple of Semikron IGBT bricks in my Odin project, I am trying to figure out the reasons for their demise.
I'm worried that they may simply not be as overbuilt as the older Powerex H and F series. The newer devices like the 8 and T4 series that I destroyed have smaller IGBT chips in them running at a higher current density, and the Semikron applications manual actually states that currents over Icrm will damage the chip metallisation and cause premature death. I've always suspected this would be the case, but never seen it stated in any Powerex application note.So: Who all else has used Semikrons? What part numbers? How far have you pushed them over the Icrm rating and how many hours of runtime have you racked up at this current?
Re: The Semikron IGBT thread
Hydron, Sun Jun 15 2014, 09:27PM
I have used a couple of SKM200GB125Ds at up to ~750A (2.5x Icrm) at ~80kHz. Did not see any issues, and I guess this is a pass/bang test.
Not sure about the run time, would not have been particularly long (10s of minutes max). There is a possibility (can't remember which bridge I used for the test) it did a 15 min run induction heating a pan at 1-2kW, would need to see if I have a pic of it to find out which bridge was used. Coil is now running 2x CM300s, so it's not easy to do further tests unfortunately.
Edit: forgot to note, these IGBTs were being driven by UD driver + GDT at +/- 24V. Have attached a scope shot of them running at ~600A (2x Icrm):
Hydron, Sun Jun 15 2014, 09:27PM
I have used a couple of SKM200GB125Ds at up to ~750A (2.5x Icrm) at ~80kHz. Did not see any issues, and I guess this is a pass/bang test.
Not sure about the run time, would not have been particularly long (10s of minutes max). There is a possibility (can't remember which bridge I used for the test) it did a 15 min run induction heating a pan at 1-2kW, would need to see if I have a pic of it to find out which bridge was used. Coil is now running 2x CM300s, so it's not easy to do further tests unfortunately.
Edit: forgot to note, these IGBTs were being driven by UD driver + GDT at +/- 24V. Have attached a scope shot of them running at ~600A (2x Icrm):
Re: The Semikron IGBT thread
Goodchild, Mon Jun 16 2014, 02:35PM
I have pushed the SKM200GB125Ds up to 1200A without issue on a couple coils I built for a museum early last year, I have yet to cook a semikron. I also used +-24V GDT gate drive from a phase lead UD style driver.
Goodchild, Mon Jun 16 2014, 02:35PM
I have pushed the SKM200GB125Ds up to 1200A without issue on a couple coils I built for a museum early last year, I have yet to cook a semikron. I also used +-24V GDT gate drive from a phase lead UD style driver.
Re: The Semikron IGBT thread
Steve Conner, Mon Jun 16 2014, 03:33PM
Thanks for the info guys! Seems to be all 125s so far, has anyone tried the 123, 128 or 12T4 series?
Steve Conner, Mon Jun 16 2014, 03:33PM
Thanks for the info guys! Seems to be all 125s so far, has anyone tried the 123, 128 or 12T4 series?
Re: The Semikron IGBT thread
Goodchild, Mon Jun 16 2014, 07:23PM
Forgot to mention, I have also take the 128s up around 1kA as well.
Goodchild, Mon Jun 16 2014, 07:23PM
Forgot to mention, I have also take the 128s up around 1kA as well.
Re: The Semikron IGBT thread
Reaching, Wed Jul 02 2014, 12:31PM
Hi Steve :)
i still use semikron igbts for all of my coils. two coils i build 4 years ago are still running fine without a single igbt failure.. this coils are running at least 2 or more hours per week. they should have a few hundred hours runtime already without a single failure in the entire system since then.
for smaller coils i use the skm200gb126d series and for the larger one skm400gb126d.. the ocd protection is set 20-50A below the max rated current of the datasheet. so even the big coil runs at only ~350A peak current. thats enough to produce some nice sparks and make people happy :)
the semikron modules i destroyed failed due to gate overvoltage ,not overcurrent.
Reaching, Wed Jul 02 2014, 12:31PM
Hi Steve :)
i still use semikron igbts for all of my coils. two coils i build 4 years ago are still running fine without a single igbt failure.. this coils are running at least 2 or more hours per week. they should have a few hundred hours runtime already without a single failure in the entire system since then.
for smaller coils i use the skm200gb126d series and for the larger one skm400gb126d.. the ocd protection is set 20-50A below the max rated current of the datasheet. so even the big coil runs at only ~350A peak current. thats enough to produce some nice sparks and make people happy :)
the semikron modules i destroyed failed due to gate overvoltage ,not overcurrent.
Print this page