2012-03-07 05:35 AM
Hi,
we have problems in using the STEVAL-IHM031V1 power stage. As this is a quite fuzzy problem, we hope someone has some hints for us. Our test setup is a STM3210E-EVAL Board, the named power stage and an approp. BLDC. We implemented a block commutation by hand not using the STM32 FOC motor control library. The implementation is quite similar to that one: http://www.mikrocontroller.net/articles/STM32_BLDC_Control_with_HALL_Sensor but not the same. We changed the dead-time, tried running with and without active freewheeling but couldn't detect differences in behaviour between these settings. With all settings, the motor drove. At some time, sporadically, something on the power stage broke, and it got a short-circuit (voltage broke down, current went in saturation on the laboratory power supply). This scenario we had with two boards now. Just before the defect, we didn't change the software, i.e. with the same SW state, it already worked. We assume, that it breaks because of wrong/bad switching timings but can't be sure, as the problem is not reproducible. We don't use over-current protection in SW at the moment. In the manual of the power stage, I can read something about over-current and short-circuit protection. Does someone have experience with those defects, e.g. can say, which parts on the board would break first under those circumstances? There's nothing ''charred'' to see under a magnifying glass. Except over-current protection by software, what else can I do to prevent those defects? Thanks a lot in advance! Best regards, Sebastian2012-05-10 07:58 AM
Hello Sebastian,
from your description I guess you should be running your motor by mean of Hall sensors. In this case, may be possible some noise on the sensors outputs cause the MCU to detect a wrong state leading to an undefined state of the PWM outputs? BTW, I agree you should be enabling over-current protection soon :) Regards, Billino