cancel
Showing results for 
Search instead for 
Did you mean: 

Unusual high current (almost short-circuit?) on startup with openloop STSPIN32G4

ALedi.1
Associate III

I am using a custom PCB based on the EVSPIN32G4 schematic and STSPIN32G4 and IRF7759 MOSFETs. Since the 10 kW motor that I try to run doesnt spin, even with 30 A current at startup, I wanted to use the board in openloop, so I changed the OPEN_LOOP_FOC in the .ioc file to true, and set startup to look like this:

0693W00000QLW2oQAH.pngWhen I start the motor, the current in phase B rises to 350 Amp before an undervoltage error stops the motor (power supply cant keep up with current). I think if I connect a battery, the MOSFETs would burn through. switching frequency is 30 kHz and with 15 kHz I have the same issue. Current waveform is as follows:

0693W00000QLW70QAH.pngand the oscilloscope measurements

CH1, yellow = Phase V highside gate,

CH2, blue = Phase W lowside gate

CH3, pink = Phase V lowside gate

CH4 green = Phase W highside gate, all measured to ground

0693W00000QLW7oQAH.png0693W00000QLW8XQAX.png0693W00000QLW8cQAH.png0693W00000QLW9LQAX.png0693W00000QLW9QQAX.pngIf someone has an explanation what goes wrong here, please give me a hint.

6 REPLIES 6
Not applicable

Which app do you use to take this screenshots? As they are quite clear

ALedi.1
Associate III

I use the high frequency plot log function from motor pilot to log the file as CSV and then use excel to plot as diagram

ALedi.1
Associate III

I already tried to increase both the speed ramp final value to 3500 RPM and also to increase the current to 20 A, both didnt change the behavior, the motor doesnt spin.

MK.1
Senior

Hello, I could be wrong, but it seems to me that there is a short circuit due to the lack of dead time on the waveforms.

ALedi.1
Associate III

thank you for the suggestion. The deadtime is set pretty high at 2000 ns. The rise of the green line is due to the lowside FET being switched off, so the potential of the highside gate rises but the HS FET is not yet switched, the switching occurs at the second impulse, this is due to the fact that I have GND as reference for all channels, and not the source of HS FET for the highside channels. 0693W00000QMbEmQAL.png

MK.1
Senior

Yes you are right.

You can make an oscillogram of signals

 on the gates of transistors of one phase of the motor, but connect the common wire of the oscilloscope to the output, and not to the ground?

If you connect the 3 and 4 probes of the oscilloscope to power and ground, you can see what happens at the output.

Just be careful, because the common wire of the oscilloscope will be under high voltage!!! It is better to galvanically decouple the oscilloscope for power supply !!!