DRV8303: OCP triggered

dominique houzet

Part Number: DRV8303

Hello, I use drv8303 with bldc motor and 3 shunts with 50A mosfet. OCP is triggered with 45A threshold. I see high currents during lowside mosfet commutation, mainly on one phase. WHen I modify motor phases connections, THe high currents occure on the New corresponding phase, as if it dépends on the motor inductance. I see high currents on the shunts and they propagate on GND and éven on Gate Mosfets . How can I reduce those short circuits during mosfet commutations?

Sincerely yours

Dominique Houzet

over 6 years ago

0 James Lockridge19 over 6 years ago

TI__Genius 14485 points

Hi Dominique,

When you change the order of the rotor phases, are you seeing the large current stay in the same H-bridge output, or does it stay with the same motor phase?

If you see the issue in the same motor phase regardless of connection, can you check your motor phases with an Ohmmeter to ensure there is not a short somewhere?

0 dominique houzet over 6 years ago in reply to James Lockridge19

Prodigy 95 points

it stay with the same motor phase! I've measured the resistance, it's ok, I've tried an other motor (same model), the behaviour is the same with only the amplitude of the large current that is a bit lower with this motor.

The high current occurs only exactly at the low-side mosfet commutation instant (very short time of 10s of ns). It seems to be linked to the way the driver manage the commutation???

Dominique

0 James Lockridge19 over 6 years ago in reply to dominique houzet

TI__Genius 14485 points

Dominique,

Are you using a TI EVM board, or did you build your own board?

What kind of commutation method are you using? Do you have hall sensors in your system?

0 dominique houzet over 6 years ago in reply to James Lockridge19

Prodigy 95 points

I ve designed my own board following datasheet PCB layout recommandations. I use FOC régulation with 1us deadtime. When high side goes low from Vbat+10 to Vbat during deadtime then lowside goes high (at this instant there is a high current and high side goes down to zéro)

Also it s sensorless.

Dominique

0 James Lockridge19 over 6 years ago in reply to dominique houzet

TI__Genius 14485 points

This sounds like you may have a shoot-though issue on the FET H-bridge from not enough deadtime.

Does the motor commutate at all, or does the OCP trigger before the motor starts to spin?

Can you please share your schematic with me?

Also, if you can take scope shots of the high side gate drive signal, low side gate drive signal, and the switching node of the phase with the issue, that would be helpful. If you have a current probe, can you put it on the motor phase that has the issue? If no current goes into the motor phase during the short condition, this may indicate FET shoot-through.

0 James Lockridge19 over 6 years ago in reply to James Lockridge19

TI__Genius 14485 points

Hi Dominique,

Thank you for sharing your schematics and scope shots with me. If the issue is shoot-through on your H-bridge, there may be a couple ways to fix this.

Increase the switching dead-time by increasing the resistor values on DTC.
Increase the resistors between the drivers and the gates. In your scope shot, there is a lot of ringing on the high-side gate and the switching node. It is possible this turns on the low-side FET briefly.

0 dominique houzet over 6 years ago in reply to James Lockridge19

Prodigy 95 points

Hello,

I've tried Max value for DTC: no effect at all. I ve tried 30ohm on Gates, this delete ringing but lower rising speed. The main shoot-through remains! With 100ohm on Gates the high side is not working...

Dominique

0 James Lockridge19 over 6 years ago in reply to dominique houzet

TI__Genius 14485 points

Dominique,

Can you take a scope shot of the high-side and low-side gate voltages, the switching node voltage, and the voltage on the shunt for the failing half bridge? You may want to look at those signals at various time scales to see where the problem might be. If possible, you should test the DRV8303EVM to see what differences might exist between your system and the EVM.

0 Hector Hernandez Luque over 6 years ago in reply to James Lockridge19

TI__Expert 8980 points

Hi Dominique,

Did James' suggestion resolve your question?

If so, please close this post by accepting the post is resolved in James' highlighted in yellow answer above me.

Thanks!

0 dominique houzet over 6 years ago in reply to James Lockridge19

Prodigy 95 points

Hi, I've bought the DRV8303EVM. I've put 150 Ohm on low side and 10 Ohm on high side. The High currents are lower now (30A) thus it works. The mosfet are still very hot.
thanks for your help.

Motor drivers

Motor drivers forum

DRV8303: OCP triggered