IGBT gate driver

Thread Starter

dhakal50

Joined Jun 28, 2011
5
I am using a IGBT gate driver IR2117 to drive a IGBT which drives a 4- quadrant PMDC motor of 75 V dc. I have a problem with driver. The input to the driver is PWM technique. The problem is the IGBT driver chip always blows out when I switch on the dc supply. Actually if the duty cycle is 50% then it is fine but when the duty cycle is increased to 95% it always blows out. Sometimes IGBT too.

Any help will be appreciated from my heart.
 
The problem is the IGBT driver chip always blows out when I switch on the dc supply. Actually if the duty cycle is 50% then it is fine but when the duty cycle is increased to 95% it always blows out. Sometimes IGBT too.

Any help will be appreciated from my heart.
Could be bus pumping forcing up the power supply volts ?
Watch the power supply when you power up and see if it pumps up way above normal voltage on 95% duty cycle.
 

praondevou

Joined Jul 9, 2011
2,942
Do you have a schematic?
Turning ON the DC Power supply, you mean you turn on the bus voltage and the controller is already working, or you first apply the bus voltage and then turn on the controller voltage? (Even if you turn them on at the same time there will be a delay, what's that delay?)

To avoid burning and do some tests you could lower your bus voltage until you find out what's going on.

What IGBT are you using?
 

Thread Starter

dhakal50

Joined Jun 28, 2011
5
Ya I have schematics. At first I turn on the control circuit and only I turn On the DC Bus. There is sufficient time delay between these two turn on. By lowering the DC bus voltage sometimes the IGBT driver doesn't blow out. The motor I am using is a 75 V dc and it draws 5.8 amps at full load. The IGBT I am using is G4PC40FD.
 

Attachments

praondevou

Joined Jul 9, 2011
2,942
Q1 and Q4 don't have an emitter connection to their drivers? Or is it just not drawn?
Did you measure your deadtime?
Are you sure that the outputs of the opamps and therefore the output of the 2117s are OFF at the moment (and shortly after) you turn on DC?
 

Thread Starter

dhakal50

Joined Jun 28, 2011
5
The emitter of Q1 and Q4 is not drawn. Sorry about that. My dead time is about 600 ns. I usually turn on the drive circuit then only turn on the DC bus. I don't know whether I have to turn on DC bus first then turn on the driver circuit or do it in opposite way. Usually my Chip IR2117 blows out not the IGBT.
 

praondevou

Joined Jul 9, 2011
2,942
No , the way u do it is better, first control side, then power side.
As the failure seem to originate from the driver and not the IGBT, can you put less bus voltage on your circuit and see if you have a negative voltage spike at your driver output at the moment the output signal goes from High to Low? If yes, how big is it?

600ns is bigger than the turn-off time of the IGBT but can you increase it a bit? Just to make sure.

What's the distance from driver to IGBT?

For testing purposes you can also take the load out of your circuit and put a light bulb in series with your bus voltage, if cross-conduction occurs at least the IGBT won't blow.
 

prepka

Joined Oct 5, 2020
13
I looked at your circuit and you are not creating the needed dead time for your drivers and getting shoot through that blows the IGBT and blows the driver. You need a fast diode like
1N4148 across resistors R17-20 to fast discharge the 510pF capacitors that are delaying the turn on. Without it,the turn off is delayed also and you blow the IGBT when the opposite side is still on at 95% causing the shorts.
 
Top