Hi,

I build a full H bridge inverter at 40Khz starting from a same project at 25Khz.
The output stage use IGBT IXGH20N60BD1 with fast reverse diode.
Logical PWM signal comming from a CPLD controller.
IGBT driver is a double IR2110.
Main bridge voltage is 350 Vdc (from 220Vac line)
Out power is 1.5 Kw

The problem is on hi-side driver on IR2110 that create a cross-switch problem. It's really a strange problem and I really don't understand.

With CPLD controller I have a large cross conduction delay (5 uS) to be sure about t-on and t-off delay of IGBT (isn't so fast devices..) but when I check the output signal on the gates and I see:

- HI-SIDE=1 / LO-SIDE=0 -> at the end of the ON time of HI-SIDE, the gate signal goes low but sometimes it turns on itself (it's not seem an oscllation but an high fixed signal). When the LO-SIDE become ON, the HI-SIDE gate goes low but cross-conduction increase.

- HI-SIDE=0 / LO-SIDE=1 -> at the end of the ON time of LO-SIDE, sometimes the gate signal of the HI-SIDE goes ON itself, before the real on signal from the controller. This "false" on state can be as long as cross-delay (so the on pulse will be add to the real ON period) or shorter and I see as pulse on the output stage of the inverter. This add noise, cross.conduction and so on.

To debug the system, I try with a low main voltage ( 50/60 Vdc) and a fix duty cycle to be sure about bootstrap capacitor charge but the effect is absolutely the same !

--- Have you same ideas about ?

Thanks to all in advance

Francesco

 

A schematic might be helpful.

 

Thanks to your attention.

Follow schematic ..

Thanks again, waiting for your kindly replay

Francesco

 

What voltage are you using to pins, Hin/Lin? The same 3.3V you have on pin SD?
The way I understand the ir2110 you need at least 5V to turn those three pin on.

 

Logical supply voltage can be down to 3 V (minimum value from data sheet).
This value can be used, I'm sure. This circuit, in this configuration, is working good at 25Khz.
I think that at this higher frequency (not so high !!), IGBT start to go in auto-conduction but the strange effect is that I see the signal just on pin of IR (serial gate resistor is low - 10 ohm - but, looking at the level, seems to that the signal is comming from IR and not from the gate). This happen even if I decrease the main bridge voltage at 40 Vcc.

If you need more information, don't esitate.
Thanks for your attention

Francesco

 

In IGBT use the is something called "tail current" that effects turn off of the device.

In this Tech Note; http://www.irf.com/technical-info/whitepaper/choosewisely.pdf
on the third page it says that for motor control, the switch frequency should be under 20kHz for IGBT's. maybe there is something in that link that will help.

 

Hi Francesco,

I know it has been long time (almost 4 years, but I have a problem which is identical to yours at here. It is a very critical issue on my thesis, I have to fix it as soon as possible. Do you remember what was the root cause of that problem and how did you fix it?

I am looking forward to your relpy.

Thank you
Mehmet

 

Francesco made 3 posts in 2010. I doubt he is still reading this thread, unfortunately. You might try starting a new thread with a detailed description of your project and what issues you have. Be sure to include drawings, links or anything else that would be helpful.