Protection Circuit for DRV8825

Hi Arhez

DRV8825 do have internal protection on almost all possible outputs short condition. But to perform good protections, the device must be used under proper PCB layout and external connections.

When disconnect the motor online, the free ends may cause high peak voltage or electric arc caused by the self induction. That is one possible reason to blow out the device.

Could you give more info on the schematic, VM voltage and motor's parameters for us the analysis or reproduce the issue?

Best regards,

Hey Wilson

I just want to make clear that none of the ICs have been damaged, I asked you this question to be on the safe side and plan any precautionary measures I need to take before finalizing my design.  I am using Oriental's PK243-3A stepper motor in bipolar series configuration, and as I mentioned earlier in my posts we are designing an educational kit for students to understand the  basic functionality and features of a stepper motor therefore we need to be cautious as students can disconnect the motor while its running.

I will be happy to share more details like the schematic, can I get your email address? It will be more feasible to attach schematics and further info over there.

Regards

Arhez

Hi Arhez

Had given you the email address via adding friends.
So, you mean no device was really damaged when you disconnect the motor online, right? Is the nFAULT asserted in this case?

Best regards,

No device was damaged, this post is in reference to a possibility that might happen if someone breaks connection while the chip is powered. Yes the nFault pin is being used, pulled up by 5V. I am sending you the schematic in message, please suggest a protection circuitry that would help save the chip in this case.

Hi Arhez

By bench test on EVM, the hot unplugging of the outputs will cause a set of spikes on the outputx pins, VMx and ISENx. The spike on VM should be keep below the recommended VM range. It will be depent on the motor's inductor, PCB layout and the bulk caps decoupling of the VMx pin. In EVM, we can see up to 10v spike when hot unplugging the outputs. So we need to keep the V_supply + V_spike < VM_max.

The hot unplugging will also cause huge noise on the ISENx pin. The noise can also reach up to +/- 10v easily which is already much higher than our IOCP level. But it may not enter into protection because of the Tblanking time. The Tblanking time is about 3.75us while the spikes are a set of ~10ns peaks. So it will not always sampled by the OCP protection circuit. That is why you may not seen nFAULT every time when doing the unplug.

So, please have check on the MAX spike on VM, that is something may kill the device. If no breaking the rule, it should be safe to go with the device's own protections in the hot unplug condition. Anyway this is not allowed in normal operation.

Best regards,

Thank you Wilson. I will check the max spike.