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.

DRV8301 high side switching slowly

Other Parts Discussed in Thread: DRV8301

Hey,

I'm using the DRV8301 in a mtor control application and after a pcb redesign i am experiencing problems with 2 of 3 high side switches.

While  one pase is switching fine, the other two seem to be switching very slowly (see scope screenshots attached, switching takes several µs). What could be the reason for this behaviour?

Working phase (channel 1 is U_gate of high side switch, channel 2 low side):

broken phase (channel 1 is U_gate of high side switch, channel 2 low side):

Thanks for your help in advance,

Oskar

  • Hi Oskar,

    Have you tied all GND pins to GND including the power pad?
    What is your PVDD1 voltage?
    What is different about the working phase and non-working phase from the layout point of view?
  • Hi Rick,

    all GND pins including the pad are tied to ground in the schematic and pcb design, i have no way to check wether the pad is soldered properly, though (but i am confident it is).

    PVDD 1 and two were both 7V. Since I use shunt resistors in the phases instead of low side shunts the source of the low side MOSFETs are directly tied to ground as well. The connection of that IC pin to the corresponding MOSFET pin might be of slightly higher impedance in the non-working compared to the working phase.

    I now put together a new PCB and am now experiencing a slightly different problem. It's the same phase that is not working (and the high side switch again), but this time the gate voltage looks like this (channel 1 working, although rise time still high, channel two broken phase):

    What mistakes could i have done in the design that could cause these kind probles? To me it looks like the bootstrap cap cant deliver the necessary current to charge the gate, but the cap is a 100n ceramic (x5r or x7r, i'm not shure, also tried higher capacities) and it is the same in all phases. The voltage across all bootstrap caps is also stable.

    Thank you very much,

    Oskar

  • Hi Oskar,

    It does look like a bootstrap problem. Channel 1 shows 2V DC on it when it tries to switch. There are a few items to check when the low side FET is on.
    1) Check the BST_x voltage of the phase to make sure the voltage is correct.
    2) Check the SH_x voltage when the low side FET is on.

    The voltage should be ~GVDD - a diode.

    Are you monitoring the nFAULT and nOCTW pins? If either of these assert, are you reading the register to determine the cause?
  • Bootstap voltage of broken channel (5V per div, voltage is about 8 to 9 volts):

    S_H of the broken channel (aka output voltage of the half brigde) on channel 1, channel 2 is the high side input PWM with 3.3V of amplitude (i am using 3-PWM mode by the way):

    I will check the SPI messages later today.

    Thanks for your help!

    Oskar

  • Looks like the gate driver is working less and less. Now two phases are not working and the high side gate signals are looking like this (channel 1 non-working, channel 2 working):

    I also reenabled overcurrent protection and the low side FETs gate signals look like this (channel 1 non-working, channnel 2 working):

    The SPI register 1 shows that the gate driver ist detecting overcurrents in the low side FETs in both non-working phases wich explains the short on-period of the low side FETs. All other status bits are 0.

    Since the schematic for all phases is identical i am wondering what to change in the layout to prevent this behaviour.

    Any explanation appreciated

    Oskar

  • Here is the schematic and layout, in case you want to have look at it.

    schematic_layout.zip

    Thanks a lot!