[SOLVED] IGBT Heating in DC-to-AC Full Bridge Application

Hi, so basically I'm building a full bridge DC to AC inverter using SPWM passing through an LCL filter. I've already produced the necessary signals, and able to produce a sine wave at the filter output however I am not able to increase my bus to 400V due to the following reasons:

1. The signal gets disrupted at 250V input. (I assume it's due to the bus voltage noise at higher levels)
2. Q2 and Q3 gets heated up alot more (around 20 degress Celsius higher) than Q1 and Q4. The IGBT I am using already has built in body diodes, and I also attached external SD diodes to Q3 and Q4.

Can anyone explain why #2 is happening and what can I possible do to fix it? I need to reach 400V input, however at 225V, the temperature of the diodes mentioned at #2 are already at 90degC.

Thanks!

Here's my main bridge schematic and at the output is an LCL filter.

what is switching frequency, maybe it is too high for igbt?
Is your layout good?
If not then noise will happen, and yes at higher voltage it would be probably more.
But i dont know what you say, because do you mean your inverter doesnt work when your bus is 400vdc, or do you mean you cannot produce the 400vdc?
What is you gate drive like? You have negative turn off voltage like you should?

- - - Updated - - -

You have nice tight current loops?...you get the high di/dt current loops low in area?

Hi treez!
I think that the noise really is due to my layout but right now that's the least of my problems. :/
Sorry for the late response by the way.

- - - Updated - - -

treez said:

what is switching frequency, maybe it is too high for igbt?
Is your layout good?
If not then noise will happen, and yes at higher voltage it would be probably more.
But i dont know what you say, because do you mean your inverter doesnt work when your bus is 400vdc, or do you mean you cannot produce the 400vdc?
What is you gate drive like? You have negative turn off voltage like you should?

- - - Updated - - -

You have nice tight current loops?...you get the high di/dt current loops low in area?

Click to expand...

Q1/Q3 switches at 24kHz and duty varies from 0 to 90% duty cycle while Q2/Q4 switches at ~60Hz (grid synchronized application)
My gate drive is from 0-12V. Maybe this is why my circuit is having problems at higher bus voltages? Please explain the negative turn off voltage? However, I'm using a gate driver IC (IRS21867) which is said to be used also to drive IGBTs.
Any bus voltage higher than 250V, I get spikes at my output waveform.
Also, I'm using STGP10NC60KD which should be okay for my operation since I'm only operating at 100W max power at 230Vrms.

Okay, I was able to solve the problem and apparently Q2 and Q3 were heating up cause Q3 was being falsely triggered due to the noise at the gate driver, providing a return path for the positive cycle of operation (operating time of Q1 and Q4). I was able to remove the noise at the driver and was able to get rid of the overheating and I'm now able to increase my bus voltage to 400V (which is the required bus for my specs.)