DRV8303: Over-current mode didn't get corrent current

Li,

Could you please also try setting 0.822V for Vds level and check where the OC trips?

Have you tried the same experiment with only one FET installed per phase? I see the schematic has two FETs per phase.

Regards,

-Adam

Adam,

Thanks for your reply.

Yes, I have two boards. One board only install one FET. The other I install two FETs each leg. Two FETs board  I got 20Amp when offset voltage is 0.403V.

I also tried different OC adjust offset. When setting at the smallest offset, I only can get  6Amps. when I set the largest voltage offset, I got 20Amps.

Bests,

Li

Li,

Could you please explain the values of your two current sense resistors?

What sense amplifier Gain setting are you writing to register 0x03? Please let me know the full data being sent to register 0x03.

Also, we may have to adjust your current sense resistor values.

Have you tried adjusting the current sense per section 8.2.2.3 in the datasheet.

Hi Adam,

Do you mean shunt resistor is related on the current limit mode?

My current sense is 1mohm each. That means the bus current sensor value is 0.5 mohm.

I tried to adjust the sense amplifier to 80V/V. It's 0x180c. 

Today I also tried set it back to default value. But I didn't see any big difference.

I also set the OC_ADJ_SET to 22, which is the 0.822V.  I get the maximum current is 13V.

When chip goes to OC , it's very easy to damage the chip. till now , I already burned out 10 pieces.

Thanks,

Li

BTW, The Vref is 3.3V. I connect to the Power directly. I didn't add TL431.

Hi Li,
Could you help us understand if the issue is only seen on the design with 1 FET per phase or 2 FET per phase ?
If you are seeing this on the design with 2 FET's only, then a possible reason is that the two FET's are not thermally coupled to each other. Due to this , one of the FET's will take most of the current and this could cause a trip.

Regards
Anuj

Hi Anuj,
It happens in both situation.
The detail issue is the current limit mode will works when the current reach up to round 10A. I detect the OTW signal. When I continued to increase the torque, the chip will burn out.
I very appreciate you could help me to find the issue.

Thanks,
Li

Hi Anuj,
Today I did lots of tests. I found phase voltage will increase noise when increasing torque. When the current reached up to around 10A, the phase peak-to-peak voltage will over 65V. That caused OC happens. So do you have any better idea to reduce the noise? what should I improve when I re-design the board?
I checked the other board which doesn't use DRV8303. The other board also has that kind of noise. But their gate driver can handle those noise.

Thanks,
Li

Hi Li,
It appears that the issue is caused by the high peak-peak spikes. Please refer to the link below for some of the best practices and recommendations when you design your next board.
e2e.ti.com/.../746299

DRV8301 has a fixed amount of response time to detect an OC event and this may simply be too short for the spike you see. The other device you are looking at likely has a longer delay time to recognize the OC. Both are not ideal , since if the delay time is too long you are stressing your FET's and could damage them.

Do you have an opportunity to replace the driver with our new driver series www.ti.com/.../DRV8323 ?
These devices have configurable VDS filter times
Anuj

Hi Li,
I hop ewe have answered your questions. May we close this thread ?