3
\$\begingroup\$

I am using an Allegro Microsystems A4935 3-phase MOSFET driver in a brushless speed controller. I recently got the motor spinning, but when I began increasing the voltage, my commutation started to become jerky, for reasons I suspect to be in software.

What I'm confused about is that, when this happens, both my MOSFETs and MOSFET driver heat up. At one point I released the magic smoke in the driver, and eventually the VBB trace (high voltage, supply for gate driver circuitry) going into the MOSFET driver even burned and broke! I understand that whenever the motor gets stuck, it causes large amounts of current through the MOSFETs, but why does this cause the MOSFET driver to heat up also?

The resistors I have at the gates of my IRFS7530 MOSFETs are 4.7Ω, and I'm driving the PWM at 25kHz. The driver only gets a little warm at no load driving the capacitance of the MOSFETs. My question is, what causes this stress on the driver when there is a load? Why does sudden high current flowing through the MOSFETs cause the MOSFET driver to need lots of current? Isn't the only load on the driver due to charging the gates, which is independent of the current through the MOSFETs?

Here is the schematic for the relevant part of the circuit: Schematic

\$\endgroup\$
  • 1
    \$\begingroup\$ Possibly because when the MOSFETs burn up they fail short including the gate, shorting the output of the driver? Does it have short-circuit protection? I haven't look at its datasheet... \$\endgroup\$ – Fizz Nov 18 '15 at 7:05
  • \$\begingroup\$ Actually it does have such "ESF Pin This pin (Enable Stop on Fault) determines the action taken when a short circuit or overtemperature fault is detected", but are you using it? \$\endgroup\$ – Fizz Nov 18 '15 at 7:09
  • \$\begingroup\$ The MOSFETs aren't the ones failing though. They only get warm under the high current because of their high current rating. \$\endgroup\$ – Raphael Chang Nov 18 '15 at 7:10
  • \$\begingroup\$ I see, I misread that part. \$\endgroup\$ – Fizz Nov 18 '15 at 7:12
  • \$\begingroup\$ I'm wondering if this has anything to do with transients or "inductive kickback" during switching, but I don't really understand this concept. If someone could explain that would be great! \$\endgroup\$ – Raphael Chang Nov 18 '15 at 7:13
1
\$\begingroup\$

I would suggest you to monitor the xHI and xLO signals getting inside A4935, and also the respective Gate output from the A4935, using oscilloscope. Sometimes, due to switching of large current, there is a lot of EMI noise that penetrates in these small signals. And false trigger your MOSFET. Resulting in occasional Short-circuit condition.

I would also suggest to add 10K register between every MOSFET's Gate and Source.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.