Hi Everyone,

I have designed a H-bridge circuit and I made a prototype board for which I have also attached the schematic of the circuit.

The board is designed for 300 V DC operation. But i started to test it with 48 V DC(safer side). I tested it as two half bridge circuits seperately.

The resulating 48 V square wave has peaks as the frequency is increased.
At 5 KHz , I don t have any peaks but at 30 KHz I have these voltage peaks.I have also attached the oscilloscope readings.

Could someone please help me out.
I am so exhausted.

I tried increasing the value of MOSFET gate resistance ,
but it ain t seem to work

Please help me out.


Thank you guys.

 

Inductive ringing?

 

Inductive ringing?

Hi,

Could you please elaborate ?
I am not sure.

For testing , I just have a resistive load. i do understand that it is not a pure resistive load. It has some inductance.

 

Well I think your gates are ringing. But I'm no pro at this.

Post your schematic and someone will come along that knows how to help you better than I can.

 

Well I think your gates are ringing. But I'm no pro at this.

Post your schematic and someone will come along that knows how to help you better than I can.

Hi,

Please check the attachments.
I have posted the schematic and waveforms.

Thank you anyways.

 

Oops, sorry, I see you did indeed include the schematic. Here's a portion of your drawing. (Some folks won't chase down attached files.)

 

Do you have any load on bridge then testing?

 

Do you have any load on bridge then testing?

Ya.. I do have a resistive load for testing . 45 ohm - 100 watt..variable resistor

 

See if this Google serach may help you https://www.google.com/search?q=h-brige+snubber+network

 

You might try a fast diode for the clamp diodes. The internal FET diodes are pretty slow.

 

You might try a fast diode for the clamp diodes. The internal FET diodes are pretty slow.

Hi,

I have MUR860 diodes parallel to my MOSFETS.
is that not good enough ??

Thanks.

 

See if this Google serach may help you https://www.google.com/search?q=h-brige+snubber+network

I will check that... thanks..

 

Hi,

I have MUR860 diodes parallel to my MOSFETS.
is that not good enough ??

Thanks.

That should be a good one. One other possibility would be some line inductance, so you might try a few microfarads from the top to the bottom of the bridge.
After that some more scope pictures of the FET pins and maybe the current (looking for shoot thru)

 

That should be a good one. One other possibility would be some line inductance, so you might try a few microfarads from the top to the bottom of the bridge.
After that some more scope pictures of the FET pins and maybe the current (looking for shoot thru)

Can you please make a rough sketch about the capacitance to be placed from the top to bottom of the bridge.I didn t quite get that.

And tomorrow I could post some more important scope pictures from the mOSFETS...

Thanks!

 

Just from the positive supply to the H and ground to the H. Close to the FETs themselves. I don't know that this is it, but if the wires are fairly long from the power supply to the H is would match the simulation.

 

I just tried searching around for an IR2770 and I can't seem to find it. Do you have a datasheet for that? When you say you're testing the half bridge, does that mean you've grounded one side of the load? Also, I'd assume from the scope images that you're probing the load, but I don't see a note to confirm that. If you could clarify these things and get a scope image of both gates and the load simultaneously, it might shed some light.

It might also be helpful to send a picture of your test setup. The fact that you're getting 70V spikes with a 48V DC power supply indicates some inductance in your circuit, your measurement equipment, or both. I've had prototype setups for switching circuits that didn't work at all due to some inductive wire loops, but worked fine once the loops were shrunk down to a PCB with smaller components.

 

How close are your gate drivers to the mosfets? There is a International Rectifier application note that states something like, keeping the driver to gate distance ~1inch or less. When using a longer wire or trace it will give problems.

 

You might try a fast diode for the clamp diodes. The internal FET diodes are pretty slow.

you can try to use snubber circuit
and if the gate signal is ringing you can lower the gate resistors or use beads

 

Just from the positive supply to the H and ground to the H. Close to the FETs themselves. I don't know that this is it, but if the wires are fairly long from the power supply to the H is would match the simulation.

Hi,

I tried placing bypass capacitors. But the improvement in my waveform is not significant. I have attached the latest waveforms. Could you please check the waveform and tell me what could i do next?

NOTE:This time i used a 24 V DC power supply instead of a 48 V DC .
I have a variable resistor as a load.
The resistance was found to be 48 ohms and the inductance was 90uH.(Measured using a LCR meter).



Thank you!!

 

I just tried searching around for an IR2770 and I can't seem to find it. Do you have a datasheet for that? When you say you're testing the half bridge, does that mean you've grounded one side of the load? Also, I'd assume from the scope images that you're probing the load, but I don't see a note to confirm that. If you could clarify these things and get a scope image of both gates and the load simultaneously, it might shed some light.

It might also be helpful to send a picture of your test setup. The fact that you're getting 70V spikes with a 48V DC power supply indicates some inductance in your circuit, your measurement equipment, or both. I've had prototype setups for switching circuits that didn't work at all due to some inductive wire loops, but worked fine once the loops were shrunk down to a PCB with smaller components.

The IC is IR2110.
Both the half bridge voltages are measured w.r.t ground.
I have also attached the waveform.(Power supply was 24 V DC)
But this measurement was done before adding the bypass capacitor...

Even I think the problem is because of the prototype board. Now I am planning to make a PCB design and check the real cause.