DRV8332: Blows possibly because driving in non complimentary mode

Hi Dennis,

The device is designed to be operated in various modes. Although complementary is recommended, non-complementary can be used.

Can you provide more details regarding the device failure?
1) Are you monitoring the OTW/ and FAULT/ pins to determine if a problem is being ignored?
2) What is your PWM frequency, and min/max duty cycle?
3) What is your bootstrap capacitor value?
4) Does your 12-V GVDD/VDD supply meet the requirements of section 9.2.1.2.2 of the datasheet?
5) Can you provide a screenshot of the DRV8332 with the circuitry around the device?
6) Do you have a heat sink attached?

Hi Rick,

1) We can monitor the OTW and FAULT however to date we did not formally do so as our initial experience with the device was so good and our motors were running well. It was only when we started stressing the motors things started going wrong. And when they did we switched off immediately and lost any history... SO yes we can try to make some sort of log and dump it to a PC in future. But we will power down the boards if we see anything a miss.

2) We started at 100 Khz with a 255 level PWM ...  so we used settings of between say 20/255 to 200/255 duty cycle during our early tests. Which went well. When we stalled the motors the torque was low so we increased to 250/255 ...  thats when we got our first blow up.  We read that you should not run at 100% to allow capacitor to charge up. We also worried about the PWM maybe being too high. So we pushed the PWM frequency down to 16KHz

3) 100nF caps between BST A B C to OUT A B C

4) My data sheet is labelled SLES274B –AUGUST 2013–REVISED JANUARY 2014,  there is no such section as far as I can search in the document. However We drive the 12V from a bench power supply with 30A capability and good overall regulation.

Note that we copied much of the design from your ALTIMUM evaluation board design. Which says:

For all applications, it is recommended to add 26-Ω
GVDD power supply and
 GVDD_X pins to limit the inrush current on the
signal by reducing the load current to prevent further internal bootstrap diodes.

We did not fit these resistors.

5) As we copied the reference design explicitly you will see there is a lot of room around the DRV8332 exactly as in the reference.

You will see that the main blow out occurs around the pins 19 to 23. This blow out occured twice.

 

6) We fit a 45mm x 45mm x 10mm heat sink onto the chip using thermal double face adhesive and we screw the heat sink against te chip for good measure. the heatsink becomes grounded by screws contact.

I added fast recovery diodes to the phases in case the internal flyback diodes were not doing their job, but a 3rd and 4th blow still occured.

We noticed the blow was occuring when the PWM got to 250 so we restrict it to 200 for now but that is low for our needs.

We aso noticed that when we did hit 250 on the PWM, some strange stuff happened and it managed to somehow disrupt our CAN BUS, I guess the VDD was getting messed with somehow and the CAN bus shares the regulted down to 5V from the VDD. Could the DRV8332 not like the 250 level and pull on the VDD?

One last comment/query is regarding the order one updates the PWM and RST signals. Do you need to be careful to set all RST false, and then only tru after the PWM is set on the A B C, is there a best sequence?

Many thanks

Dennis

Hi Dennis,

I should have asked earlier, but thank you for pointing to the datasheet version. SLES274B is the DRV8332-HT (High Temperature) while my questions are for the DRV8332 (Industrial with datasheet SLES256E). The devices are similar but have a few differences.

Your post will now be routed to the group supporting the DRV8332-HT. I hope the questions I asked will be helpful to them.

Hi Rick,

All I know is that we are using the DRV8332DKD chip, I dont know if that is HT or industrial.

BEst

Dennis

I checked and I believe we are using the industrial version.

the data sheets seem very similar. I didnt even notice a difference.

Best

Dennis

Hi Dennis,

Thanks for confirming you are using the industrial version.

"5) As we copied the reference design explicitly you will see there is a lot of room around the DRV8332 exactly as in the reference.

You will see that the main blow out occurs around the pins 19 to 23. This blow out occured twice."

Was an image attached? If so, would you please resend. It did not appear.

"6) We fit a 45mm x 45mm x 10mm heat sink onto the chip using thermal double face adhesive and we screw the heat sink against te chip for good measure. the heatsink becomes grounded by screws contact."

Being on the corner of the package, are you perhaps overstressing the package with the heatsink, or not making good enough contact in the corner? Please see post e2e.ti.com/.../229936 for heatsink recommendations (spacers and max pressure).

"One last comment/query is regarding the order one updates the PWM and RST signals. Do you need to be careful to set all RST false, and then only tru after the PWM is set on the A B C, is there a best sequence?"

The PWM and RST signals can be changed at any time and in any order.