> I would not expect very smooth operation in 'D' or step motor mode (no > encoder > feedback/PID), especially with IGBTs and their long switching times that > neccessitate a large dead zone. Do you have the dead zone time set > properly? I > think the default is 5 Usec which is (deliberately) huge. (note: too > small > a dead time will be fatal to your IRAM) Same motor, same power supply (12V only), almost the same config, just MESA 7i39 and motor runs smooth in "n" mode. If you mean dead time - I set it to 500ns, as I found something like 400ns at IRAM datasheet. I checked dead time on MESA outputs with scope and it is correct. Real switching has a delay of 800-1200ns, but switch-on/off times of IGBTs are 100-330ns.
12 volts only - maybe that's the problem? Motors are rated for 48V. Something about IGBT's? Big resistance maybe? Will try 48V. By the way, what max voltage I can go with this setup when I use phase current limiting? > I would get feedback running before worrying a lot about smoothness. Tried "qi" mode with PID (almost the same working config as with 7i39) too: bldc init passes somehow, but does crazy things due to kinky movement especially when PID turns on. > You definately do not want kinks in the top of the voltage waveform. > if anything you may want a voltage step at the sine zero crossing to > compensate for the dead zone (if the required dead zone is a significant > portion of the total PWM period). My PWM frequence is 12kHz, servo thread - 4kHz. So when and why kinky waveforms are used like this http://www.postech.ac.kr/ee/cmd/102_BLDC_pedal.htm ? ------------------------------------------------------------------------------ This SF.net email is sponsored by Windows: Build for Windows Store. http://p.sf.net/sfu/windows-dev2dev _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
