DRV8350R: Some problem about the TVS diodes between G-gate and S-gate in the lowside

Hey Shengzhong

I will consult the team and aim to provide feedback by Wednesday of next week.

Best Regards,

Akshay

Hey Shengzhong,

Thank you for your patience.

The TVS diodes being used at the moment is not protecting from negative transients since it is 15V rated. The GLx cannot be lower than -5V transient so this diode will not protect from negative transient. The bidirectional TVS diode being used at the moment will only protect against positive transients.

You could add a gate resistor to further lower your gate drive current, as lowering gate drive current will help mitigate negative transients.

In the meantime, I can consult my team for more information on this issue.

Best Regards,

Akshay

Hello  Akshay：

   Thanks for your reply.I am not sure weather there is a negative trainsient.I only use this for preventive effect .So it makes me puzzled that how this  TVS destroy the mosfet?Make it switch faster or slower?

  You say we need add a gate resistor. Do you see my schematic diagram? I have add a 510Ω and our rise time is 250 ns.Do you means if I add there is no negative transients ,the TVS is not necessary for my pcb?  I add the waveform before I add the TVS.

 Waiting for your reply.

Best wishes

shengzhong

Hey Shengzhong,

I will consult my team and provide feedback by early next week.

Thank you for your patience.

Best Regards,

Akshay

Hello Akshay : 

          OK. But can you tell me the reason why the TVS will lead the damage? I am afraid that if I add the unidirectional TVS diode between GLx and SPx,the mosfet will be broken again .I only made ten pcb.

thanks

shengzhong

Hello Shengzhong,

I would like to know whether the different grounds you have (DGND, SPGND etc.) are connected with each other using a net tie or a 0 Ohm resistor. If they are not, I recommend connecting them and seeing if the issue persists. This would help diagnose the issue.

Also, I would like to make a correction to the previous reply. If you decide to add a unidirectional TVS diode, then it has to be between GLx and ground.

Best Regards,

Akshay

OKH

Hi Akhay:

      I send the picture about the DGND and SPGND(PGND).There is a common mode choke between  PGND and DGND. But reference ground of these drive chip like drv  ,mosfet is SPGND.And some sample signal ,we use DGND.So you can see  the SCH I send before,I add

the two-way TVS between GLX and SPGND(PGND).

thanks

shengzhong

Hey Shengzhong,

I will aim to provide feedback before the end of the week.

Best Regards,

Akshay

Hey Shengzhong,

I have discussed the issue with my team and we noted that the bidirectional TVS diode will not help with suppressing negative transients. You want the diode to conduct if there is a negative transient. GLx will have to dip to -15V transient for the diode (an unlikely event) to conduct with a bidirectional TVS diode.

We recommend implementing a diode in the following manner:

We are not entirely sure why the TVS diode is causing the damage, so we could suggest some testing to be done to help diagnose the issue. However, you do run the risk of potentially damaging the board.

Please let me know if you are interested in this approach and I can suggest the testing approach.

Best Regards,

Akshay

Hello Akshay，

       Thank you for your reply. Yes,after my testing our system with this board.I can use one pcb to test adding  unidirectional TVS. I have got a modle of TVS  SMF18A. I will tell your the result after my test.

thanks

shengzhong

Hey Shengzhong,

For the unidirectional diode we would recommend a diode rated between 15-18V.

If you wish to test the bidirectional diode on potential reasons for failure, we would like to obtain waveforms of SHx, V_GS, voltage across bidirectional TVS diode, nFault and INLx.

Best Regards,

Akshay

Hey Akshay :

       I am sorry I didn't get the waveforms of those. When I added  these TVS (SMF18A) and sent the run order.The motor just run for a while .And the power get into the protection state . I found that the mos is shorted. I decide to remove the TVS although I don't found the reason.Maybe the tvs make the switch time of the mos too long  so that  the upper and lower Bridges are conducting at the same time？But I didn't measure this .When I need know this , I'll take risk to do this again but at present the problem makes me puzzled a lot.Anyway , I have known that it's not a question of TVS slection,right?

thanks

shengzhong

Hey Shengzhong,

The waveforms might help explain if the issue we are seeing is due to a selection problem of the TVS diode, as we would be able to see how the TVS diode is operating and its effect on the outputs.

Best Regards,

Akshay