Why H bridge will not work?

[sheraz83s]

I have designed an H bridge and generating pwm from avr controller with dead time.But my h bridge is not working.The high volatge is shorting and several ir2113 blown up.I am attachign my schematic also.

 

[chuckey]

re: Help with H bridge not working?

IC2 is shown with short circuits on its LO and HO outputs. Different IC to IC1?
Frank

 

[sheraz83s]

re: Help with H bridge not working?

No the outputs are not short ckt.

 

[FvM]

re: Help with H bridge not working?

I don't see obvious circuit errors. Using slow 1N4007 as bootstrap diode might cause inconvenient current peaks, I would use a fast recovery diode.

But more likely the problem is brought up by details that can't be seen in the schematic, e.g. control waveforms, circuit layout, DC bus bypassing.

 

[alexan_e]

re: Help with H bridge not working?

I'm not sure it will help but I have seen recommendations for a 1K resistor between gate/source for the output mosfets, I assume the same applies for IGBT between gate/emitter

 

[FvM]

re: Help with H bridge not working?

I have seen recommendations for a 1K resistor between gate/source for the output mosfets

It's an arbitrary suggestion that can't be found in driver IC data sheets.

As a possible motivation, people expect that the driver IC output voltage might be undefined in a VCC range between zero and the threshold voltage of the internal MOSFETs, e.g. 2 or 3V. Above this threshold, the output will safely hold in low state by the undervoltage lockout function. So if the gate-source resistor suggestion might have a reasoning, it's for MOSFETs with low threshold voltage, but not for standard MOSFET or IGBT. As a drawback, the resistors will discharge the bootstrap capacitor in high-side on state and possibly demand for larger capacitors.

 

[sheraz83s]

re: Help with H bridge not working?

I am adding pwm signals for the gates of igbt.currently i am using 30v for dc bus for checking purpose only becoz my load is short ckted.

I don't see obvious circuit errors. Using slow 1N4007 as bootstrap diode might cause inconvenient current peaks, I would use a fast recovery diode.

But more likely the problem is brought up by details that can't be seen in the schematic, e.g. control waveforms, circuit layout, DC bus bypassing.

 

[FvM]

re: Help with H bridge not working?

becoz my load is short ckted

Means what?

 

[sheraz83s]

re: Help with H bridge not working?

my supply gets short circuited due to some problem in the circuit.

 

[FvM]

re: Help with H bridge not working?

But you should be able to find the reason by increasing the bus voltage from zero. Possibly also try with a reduced switching frequency. I repeat my resevations regarding 1N4007 as bootstrap diode.

Have you connected any load in the initial test?

 

[sheraz83s]

re: Help with H bridge not working?

Yes I have conencted a resistive load of 25ohm.ok i will change the diode and update you further.Thanks.

But you should be able to find the reason by increasing the bus voltage from zero. Possibly also try with a reduced switching frequency. I repeat my resevations regarding 1N4007 as bootstrap diode.

Have you connected any load in the initial test?

 

[sheraz83s]

re: Help with H bridge not working?

Yes I have conencted a resistive load of 25ohm.ok i will change the diode and update you further.Thanks.

One problem i have found is that i am not getting VB voltage just around 5v thats why my bootstrap capacitor is not charging to high value only 15v appears across it.

 

[sheraz83s]

re: Help with H bridge not working?

I have change RG=21ohms and i put 1K resistor between gate and source.I also replace IGBTs withe mosfets but still my borad is short circuit after 4-5 seconds.Also my gate signal become distorted too.

 

[sheraz83s]

re: Help with H bridge not working?

I am adding my gate driver output signal and pwm.the wave form is not smooth.
https://obrazki.elektroda.pl/9069214800_1383561210.jpg

 

[FvM]

re: Help with H bridge not working?

Looks like normal operation with medium gate resistors and large MOSFET input capacitance.

 

[sheraz83s]

Re: Help with H bridge not working?

But the problem is that my circuits gets short circuit after turn on for5-6 seconds

 

[FvM]

my circuits gets short circuit after turn on for5-6 seconds

Still a vague report.

"Normal operation" in my previous post primarly means that the waveform are expectable under specific conditions, not that they are good this way. Gate voltage rise time is a design parameter, if you want faster rise time (apparently you do), you should determine what is making it slow (I guess gate resistor and MOSFET input capacitance, possibly too large bootstrap capacitor ESR) and modify the circuit.

 

[sheraz83s]

Still a vague report.

"Normal operation" in my previous post primarly means that the waveform are expectable under specific conditions, not that they are good this way. Gate voltage rise time is a design parameter, if you want faster rise time (apparently you do), you should determine what is making it slow (I guess gate resistor and MOSFET input capacitance, possibly too large bootstrap capacitor ESR) and modify the circuit.

I am facing problem with my h bridge.I am using AVR for pwm sine wave 10khz signal.when i turned on my 15v power supply ir2113 driver supply my avr pwm output goes to zero.i also used optocoupler cny17-4 for isolation but it makes it worst by delaying the gate signals and short circuit the high voltage power supply.

 

[FvM]

I can only say that a correctly wired bootstrap circuit (as e.g. shown in post #1) won't show this kind of problems. Are you sure that the IR2113 haven't been damaged before?

 

[sheraz83s]

I can only say that a correctly wired bootstrap circuit (as e.g. shown in post #1) won't show this kind of problems. Are you sure that the IR2113 haven't been damaged before?

thanks FvM for your replies and sugegstions.I have chanegd my diode to fast diode and add coupling capacitors 100nf parallel to 10uf.Then the output at open load i am getting is 60V pk-pk, 50Hz square wave.but the problem is that i am using pwm pulses for 10ms for each leg but i am getting constant high for whole 10ms(not pulses).