DRV8353: DRV8353 continues to die. Why changing the MOSFET may be a possible solution?

Hi Jim,

The damage mechanisms for MOSFET's is usually quite simple; the VDS rating is exceeded and the MOSFET can get damaged.

For drivers, there are several other mechanisms and the most common one is damage due to transient switching spikes, especially since the driver is exposed to the parasitic inductances in the design. My suspicion would be that the damage that you are seeing is due to high switching transients on the motor phase nodes. Have you checked the phase switching voltage to see if you are exceeding the rating of the driver. Also, you could try slowing down ( lower IDRIVE) the FET turn on/off slew rates to reduce the effect.

Hi Anuj,

Thank you so much for your fast response!

The MOSFET didn't damage is a bothering fact. The DRV8353RH is the only one that blow up that puts me into this bad situation. The voltage spike at the switching node is somewhere between 77V~80V at 60V supply voltage and 10A of the resistive load current. The diagram below is the waveform we triggered the moment before the IC fials.

IDRIVE setting do reduce the problem a little bit. I've try the following 3 setting:

source/sink 450mA/900mA      mostly fine at 20V~30V, IC almost always blow up at 50V~60V

source/sink 300mA/600mA

source/sink 150mA/450mA      IC is fine at 50V~60V with little load, but as soon as we add load it fails

source/sink 100mA/200mA

source/ sink 50mA/150mA       we only try once the IC just fail when supply voltage at 45V without load 

The waveform of the switching node seems suspicious, however 77V~80V is still far behind 102V according to datasheet absolute maximum rating and also beneath Vdrain's 100V recommended operation value. Is there any possible that if the MOSFET VDS is violated and will influence DRV8353RH's driving mechanism that results in IC damaging and somewhat the MOSFET manage to recover itself?

Best regards,

Jim Hsu .

Hi Anuj,

Sorry the diagram had some minor error here is the waveform.

Best regards,

Jim Hsu .

Thanks for the waveforms. These do not look bad.

Is there any possible that if the MOSFET VDS is violated and will influence DRV8353RH's driving mechanism that results in IC damaging and somewhat the MOSFET manage to recover itself?

Ans : No, this is not possible.

At this point, I am not sure what is causing the damage but I have a couple of questions/ recommendations :

- Could you check the high side gate switching signals as well

- Are any faults being reported before the driver damage ?

- Is the motor still spinning after the driver has *** down ?

- What voltage is being used at VM ?

Thanks 

Hi Anuj,

Wish those information helps:

- Could you check the high side gate switching signals as well

This waveform is done at 60V supply voltage, motor testing with external torque applied. final moment we triggered before IC fails.

- Are any faults being reported before the driver damage ?

We use hardware setting IC, when we see nFault pull low the IC is already damaged.

- Is the motor still spinning after the driver has *** down ?

We didn't apply any brake control visually it just stop immediately, but from the waveform above it still have a little current (the green signal) after fail.  

- What voltage is being used at VM ?

We tried both single supply and dual supply and both approach damage the IC in similar way (the fault that we have been mentioning).

Dual supply : Vdrain connected to B+ (power supply varies 20V~60V), Vm connected to a fixed 12V power supply (individual power source).

Single supply : Vdrain connect directly to power supply (20V~60V), Vm have some resistor, capacitor, MOSFET and diodes in series and in parallel to make sure it is beneath 75V, we measure Vm while working at 60V supply it lands somewhere at 66V~68V at most.  

I bet a lot on Vm being the problem, however it seems not since dual supply also fails.

I will try a way to test with different VDS MOSFET, hopefully it is the main issue.

Thanks

Hi

I do the testing again and DRV IC fail again. My test condition as below :

Vdrain supply voltage 36V only, Vm supply voltage 12V motor with external torque.

The controller is fine when motor free run, however when we add the external load gradually up to about phase current 25A then the IC fails.

We check the IC and found out that the 3rd pin (CPL) have abnormal impedance about 66 ohm only, normally should be somewhere 120 Kohm or above, other pins isn't that obvious.

I have change the MOSFET to IRF7769 which VDS 100V and it still fail, so 100Vds MOSFET and Dual supply doesn't work out for me. I have been telling my team that we can solve the problem by changing MOSFET, and it didn't work.

So what might be the problem? It seems to occur at higher current no matter the voltage, so parasitic inductance doesn't make much sense, no voltage reach the limitation.

Best regards,

Jim Hsu 

Hi

I am not very familiar with the website's interface, I will just reply it here:

1. The IDRIVE setting for the IRF77689 FET is 150mA/300mA (75kohm to GND)

2. For most of the testing the FETs are fine, only DRV fail.

I kill like about 30pcs of DRV, like only two times the FET die, and they didn't blow up just abnormal.

I think when I solve this DRV problem it wil then be the time I deal with those FETs.

Best regards,

Jim Hsu

Conversation is continuing over email.