Hi,
I just discovered this Motor Driver Forum. Previously I was posting to the Tiva micro controller forum because I am using the EK-TM4C1294XL processor board.
I was doing well developing firmware for the 8305 until my motor developed a short circuit on one winding and it took down the BOOST-DRV8305EVM with it. That was probably my fault for having protections disabled. So now I have another motor and driver board on order.
I am using the 3-PWM mode and generating a sinusoidal commutation which gives me a smooth low speed motor rotation - great for positioning. I think only trapazoidal commutation is possible with the 1-PWM mode.
I see the specification recommends a switching frequency of 200 kHz but I found that the complementary outputs could not respond that fast. The duty cycle becomes very distorted at that frequency, so I slowed the PWM period down to 15 kHz and that gave me a smooth and quiet response. Is the 200 kHz recommendation correct?
The problem I have with both the 3-PWM mode and the 6-PWM mode is that the LaunchPad boards have a very limited number of PWM capable BoosterPack output pins. Even the new Delfino board brings out only 14 PWM pins. Since my business is motion control, I want to be able to control several motors from one micro-controller. So I really want to be able to control the driver entirely with the SPI interface. I can do this with the BOOST-DRV8711 board using the "force step" and dynamically changing the step mode - kind of like shifting gears. But that solution is Kludgy.
So what I would want for a 3-phase driver is an SPI to set the signed duty-cycle for each phase independently. I think the SPI interface is fast enough to control several motors in this mode of operation.
Similarly, if I want to bypass the indexer on the 8711, I would need to use four PWM outputs for 1 stepper motor. I also need to use four PWM outputs to use the driver in split mode to drive two self-commutated motors. So what I want is to have a mode that bypasses the indexer, a trivial µP function, but lets me set the signed duty-cycle for each H-bridge independently.
Thanks,
Bob Rice