DRV8353R: DRV8353SRTAR IC failure Debugging

Hey Akash,

I will try to help with this issue. 

1) Are these the first 4 units you have tested?

2) How have you confirmed the damage? Impedance measurements with a good unit?

3) Any specific test/ more info on what was being performed when the damage happened?

Best,

Akshay

Hi Akshay.

Thanks for your response.
I want to explain my problems a bit more appropriately.
1.) Yes, these are the 4 units we have tested.

2.)IC's 1&2 were damaged due to a design issue by our team, we pulled up DVDD regulator to 3 pins which was consuming 40mA of current but according to datasheet it can only give 10mA, due to which I feel the DVDD regulator went into CC mode and shorted DVDD pin & GND Pin, We were not able to produce PWM due to this
3.)IC 3 went bad due to some voltage spikes present at the MCU enable pin, this shorted the Internal TVS in MCU Enable to Ground as it showed a drop of 0.001V in DIode Mode in multimeter and a resistance of 15 ohms in resistor mode, No SPI was occurring and no Gate pulses as well.(This can be a Layout issue as well; we are still debugging this)

4.) 4th IC SPI communication is still going on, but there's no PWM, as there's always an OTS fault. The fault bit is latched onto this fault, and we are not able to clear this bit as well, and thus, there's no PWM.

Please try to help us debug IC 3 & IC 4 issues, as IC 1 & 2 feel like a design error done by us.

Hey Akash,

Thanks for the clarity. 

3) How did you confirm the DRV IC was damaged in the third unit? If MCU was damaged, then does the device get proper signals if the MCU gets replaced? If driver was not enabled and active then it makes sense that it would not toggle the gates.

4) Is this the same was the issue described in the following E2E?

https://e2e.ti.com/support/motor-drivers-group/motor-drivers/f/motor-drivers-forum/1485254/drv8353r-drv8353r-went-into-ots-and-isnt-working-after-i-let-it-cool-over-night

If so we are debugging this on that thread.

Best,

Akshay

3.) Yes we replaced the IC and Communication was established making sure that the enable IC was damaged in the previous chip
4.) Yes, it's the same issue I'm facing. I applied all the instructions you gave there, but that didn't fix the issues.
A Short Update regarding the same after a week
So we replaced IC 3 with IC 5 and tested our Controller with both Resistive and Motor Load.
Resistive Load showed promising Outputs with Desired Waveforms on the oscilloscope, but the moment we started testing the board with a PMSM motor and started changing the PWM Duty cycle to spin the motor, as described previously for IC4, IC5 went into Fault mode, showing under-voltage and Over-temperature shutdown faults. Both ICs went beyond 150° C the moment we turned them on.
We did some measurements on IC 4&5 and found out GB_LS has shorted to Ground and this might have damaged the IC, We still need help in understanding why the Gate pin got shorted to ground, No such issues are there in IC 5 though.

Hi Akash,

Thanks for the update! If GB_LS is shorted to ground on both IC4 and 5, I strongly suspect an abs max or abs min violation on GLx to GND or SLx to GND. Inductive spiking becomes more severe the more current that is flowing in the motor, and could eventually lead to damaging the driver. This is also heavily dependent on how fast the MOSFET slew rate is. What is the IDRIVE configurations you are using? What is the MOSFET Qgd?

Regards,

Anthony Lodi

Hi Anthony,
GB_LS is only shorted to IC4 not 5, We are still trying to debug why IC5 failed.
I understand inductive spiking can be a concern for Motor Loads but we were moving the motor at increments of 30°(Trapezoidal control) and not completely spinning the motor.(PWM with 50% duty of 10 kHz, 12.6 us deadtime)
Also at no Load, our Motor consumes a max. of 3A and we also had a 6A fuse in our Power cable so high current shouldn't be a factor.
Regarding Idrive these are the parameters-
300 mA charge current and 700 mA discharge current of high-side drive
150 mA charge and 700 mA discharge current of low side drive
We arrived at these values based on clearing the fault of VDS which was set at 0.7V(This value is relatively high as the Rdson of Mosfet is 3 m-ohms but below 0.7, there was always a Vds fault)
The MOSFET used is CRST045N10N where Qgd is 19 nC and total Qg is 90 nC, We are using 3 of them in parallel per Leg.

Hi Ghosh,

Thanks for the followup! I will look over this more this week and get back to you.

Regards,

Anthony Lodi

Hi Ghosh, 

Understood. You mention that at no load the motor consumes a max of 3A - could you help me better understand the reason for paralleling 3 120A FETs? How much current do you expect with a load? 

The gate current that you are using doesn't seem terribly high since you are driving 3 parallel MOSFETs with a sum total of Qgd = 57nC, however this will ultimately be determined by PCB trace inductance and the amount of transients that can be induced as a result. Especially with 3 parallel MOSFETs, PCB routing can become difficult leading to higher inductance. What gate resistor values are you using? Could you provide a schematic?

What do you have set for the TDRIVE configuration? this is bits 9-8 of register 0x04.

For IC 5, could you do a pin to GND resistance measurement for all pins and compare to a good unit to see if any pins are seeming to be damaged? 

Regards,

Anthony Lodi

Hi, our full load current is 150A, and 2 MOSFETs, along with heat sinks, are required to drive the Motor. We used 3 MOSFETs to make sure they don't heat up, as it's our first prototype.
Gate Drive LS Register (address = 0x04h) was set as 0X2135(for Tdrive Configuration)
I have sent you the Schematic, Layout, and Oscilloscope images in DM Please check that.
As we have damaged all our ICs, we have ordered new ICs for replacement. Kindly allow me some time to do the Impedence measurement comparison.
Let me know if anything else is required from my side to debug this.

Hey Akash,

Thanks for the update. I will aim to review and provide feedback next week.

Best,

Akshay

Hi Akshay,
We received a fresh batch of ICs a week ago and are facing some new issues now.
The gate pin of Phase B low side always turns high(7.8 V) when we turn on the Enable Signal; no Input signals are given. We have faced this issue with 4 ICs till now.
We tried connecting a Gate-Source pull-down resistor, and that didn't solve the problem
We also removed the MOSFET, and the Gate signal is still always high.
Can you please help us debug why the driver is behaving this way?

Hey Akash,

Can you show me a waveform of INLB, GHB-SHB (differentially measured)? Did you do a impedence measurment with a good unit to see if there is any shorts?

Also any update on IC 5 that failed?

Best.,
Akshay