This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

DRV8353: VGLS and GLA shorted

Part Number: DRV8353

Hi

I have a motor controller for a maximum voltage of 58.8V based on DRV8353S 

Some of the boards are having problems with the DRV8353 that is heating up massively and quickly as soon as the VM and VDRAIN is connected.

Inspecting one of the damaged boards I noticed that the VGLS and GLA pins were shorted with the GND pin. the VGLS capacitor is fine and the mosfet connected to GLA is fine also.

I don't know the conditions on which those boards were damaged because they were using by customers.

I just want to know what can damage the VGLS regulator and what can I do to avoid this happening to other boards and also if I need to fix the circuit for the next batch

This is the circuit of the DRV8353, VM and VDRAIN are connected with a net tie (N1)

and this is the circuit for each of the phases of the motor

  • Hello Alejandro,

    Can you provide the register settings used?

    Damage to the VGLS regulator can be caused by a few things:

    • IDRIVE setting too high causing charge to couple across MOSFET CGD into GLx and pump up VGLS
    • VGLS capacitor connection to DRV835x too long (adds inductance)
    • VGLS capacitor ground reference is incorrect or too long (adds inductance)
    • VGLS capacitor voltage rating too low

    Thanks,

    Matt

  • Hi Matt

    The config registers are the following:

    Reg 0x02 = 0x480

    Reg 0x03 = 0x3FF

    Reg 0x04 = 0x3FF

    Reg 0x05 = 0x06D

    Reg 0x06 = 0x283

    The connections with the VGLS capacitor are very short as you can see in the picture, and the rating is 16V

    Should I decrease the current drive for the lower MOSFETs?

  • Hi Alejandro,

    I agree, the component placement is very good.

    We recommend targeting 100-300ns rise time. Since your MOSFET has a QGD of 34 nC, a good starting point would be between 113 and 340mA IDIRVE source and between 227 and 680mA IDRIVE sink.

    It may be beneficial to increase the voltage rating of the VGLS capacitor to 25V. Some 16V ceramic capacitors will see excessive de-rating with an 11V bias, and some will not. Check your capacitor datasheet  to see how it performs with 11V biasing.

    Thanks,

    Matt

  • Hi Matt

    Thanks for the reply, I will try with those current parameters and for the next design I will increase the voltage rating of the capacitor.

    On other hand, I had some more boards with the same failure, and not all of them had the VGLS pin shorted to grown, but all of them had the GLA on all of the boars is shorted, and It seems that its been shorted just starting to switch or not even switching, how this can be possible? what can be another cause for this failure? can it be something with enabling and disabling the DRV?

  • Hi Alejandro,

    Since the GLx and VGLS are very closely connected on the IC, it does narrow down to a singular root cause for the damage. Can you check the connection between PGND and the DRV8353 and the PGND at the sense resistor? This is the path that current needs to flow when the low-side gate driver is turning ON or OFF the low-side MOSFET.

    • The GLx or VGLS can be forced when a different phase is turning ON, causing a rising edge at SHx.
    • As the part powers ON, VGLS is charged through an internal LDO

    Can you measure VGLS voltage when powering ON the device without switching the outputs? If possible can you measure the VGLS and GLx voltage on a board while running a motor?

    Thanks,

    Matt

  • Hello Alejandro,

    Do you need further assistance on this topic?

    Thanks,

    Matt