PWD13F60 failing on dev board, shorted FETs

[KConn.2]

Hello,

We are creating a design using the PWD13F60 and have done some testing with the EVALPWD13F60. We are running the gate driver portion at 12VDC, and the motor voltage up to 200VDC, but starting with lower voltages like 24, then 48, etc. We are running the PWM at 50KHz.

Everything works fine for a bit then the driver gets very hot and fails. We measure the 4 internal FET's and one is shorted. In one case was a low side FET, in another it was a high side FET.

The motors are unloaded, so the current should not be very high. One of the motors was a very low power motor, just rated to 40 watts, the other was higher power rated to 500 watts.

Our setup is very simple with this dev board, has anyone else had issues with this dev board not running reliably for very long?

Thank you,

-Kyler

[lauria.michele]

Hello Kyler.

The board has been fully evaluated and is available since many years, so we do not expect issues in the design.

The power dissipation of the internal MOSFETs depends on several contributions, please make sure to check following aspects in your evaluation:

• It is recommended to measure the actual current flowing in the motor windings. The inductance and parasitic capacitance of the windings my draw a current exceeding the device specification even if the motor is unloaded, depending on the motor rating and design
• Switching losses are also important, 50 kHz might be fine for some motors but might be too high if the sum of all losses contribution (conduction and switching losses mainly) exceeds the power budget of the device.
• Please check to be using an adequate dead-time at the logic inputs. A too short dead-time might result in temporary cross-conduction during the switching, which in turn increase power dissipation and therefore the temperature of the MOSFETs
• Make sure that the switching frequency and duty-cycle you are using allow for an adequate supply voltage of the High-side (i.e. the charge of the bootstrap capacitor). Driving the high-side with low BOOT-OUT voltage increases lossess. In case the Cboot voltage should not be sufficient, consider adding an external bootstrap diode between VCC and BOOT, with adequate voltage rating

Thanks and regards

Michele

[KConn.2]

Hi Michelle,

Thank you for your response. Our EE team is working on a solution to better protect this part and they seem to believe they understand what to do.

One more question I have. I noticed when I drive this chip it can only PWM at a maximum duty cycle of 98.8%, which at 50KHz means a low time of 240nsec. If I PWM above 98.8% the output to the motor appears to shut down. I verified the PWM signals form the MCU are as expected at these levels. Also, we leave the low side on all the time, and just PWM the high side in this case.

Is there some requirement for this part to have a minimum down time every n seconds or something like that? I looked through the datasheet and it doesn't really mention if this is required and how frequent. Is this like a watchdog feature?

Thank you,

-Kyler