AnsweredAssumed Answered

(Huge) problems with STSPIN32F0 custom PCB

Question asked by Max Schachtschabel on Jan 26, 2018
Latest reply on Apr 3, 2018 by T J
After some time experimenting with the STEVAL-SPIN3201 Motorcontroller board, we took it a step further and build our own little PCB with the STSPIN32F0, three power half bridges and an additional Bluetooth chip for external communication (the final goal is the wireless control of two seperate Motors using one joystick).
We pretty much build our board very closely to the schematics of the evaluation board, so that the existing software will run on the custom board just as good as before; the only difference being the shunt resistor values on the three motor phases (2x 0.01 Ohm in parallel instead of 2x 0.02 Ohm in parallel). I adapted the software by altering the "RSHUNT" value in the "Power stage parameters.h" from 0.01 to 0.005.
Like stated, the rest of the circuits and components were designed to work exactly like the evaluation board. However, once we connected and tested our PCB, we arrived at a point were none of us knew what to make of the observed behavior: After the STSPIN32F0 was successfully flashed with the software, the Hall sensors and motor cables connected and a supply voltage of 11.1V was given by an external power adapter, a modest speed ramp was set in the software's main-routine and the motor started flinching a few times, before staying still and vibrating fiercly. The power adapter reported that the current limit of 5A was reached and the voltage dropped rapidly, before we disconnected the System from the power source.
Wondering what went wrong, we did a few more tries, until we noticed that after another reset, the STSPIN32F0 did not respond anymore and there were no 3.3V on the VDD pin measurable. Probably having fried one of our Motorcontrollers, we started from scratch and build a new PCB, carefully managing the voltage pins and connections and making sure that no soldering error could occur. We tested the board again without the motor attached, and measured the outputs of the three gate drivers, which behaved exactly like on the evaluation board.
Next step, same results: The motor went crazy, the voltage dropped for a few seconds to about 8V and the chip played dead after a few tries.
On a third try, we connected the SW pin to GND and supplied the VDD pin with another external power adapter, hoping that the internal DCDC buck regulator could not cause any harm to itself and the rest of the board, since the internal control should detect this and latch off (as stated in the datasheet under "External optional 3.3 V supply voltage", page 22). Sadly, we ended with the same result: no running motor and the external 3.3V power source reported a short circuit.
Having used 2 seperate boards and destroyed 3 micro chips, we are know freshly out of ideas and not eager to fry any more hardware. The following questions remained open:
- Why does the controller think he would need 5A and more to rotate a small BLDC motor without any load?
- What exactly is the cause of the chip destruction? 
- Why do we observe this behaviour when we are using the same exact hard- and software as before with the evaluation board?

We would be very glad if someone can help us figure this out soon.
Any suggestions or feedback is much appreciated!

So long,