stepper motor vector control

Hi Gary,

Not completely sure what your are after, but have you looked at the DRV8828 and DRV8829. With these devices (2 required for stepper operation), the current is controlled by varying VREF with a DAC or setting a fixed VREF and varying the Ix inputs.

PHASE and ENBL control direction and enabling of the outputs.

App note http://www.ti.com/lit/pdf/slva641 provides more details on how to use these devices. 

Hi Rick,

I want to independently control d and q axis, with say, 8 bit resolution.  And I want to use the back emf to control position detection.  I believe that I have a means to go down to dc which would be a big improvement over current solutions.

But I either have to use something like a DRV8833, which is a small step above simple bridges or I have to build my own.  Using two 8828/9 chips has the DAC built in, but only 5 bits.  All the "fancier" chips assume that I don't want to control the field in the motor. They are for simple (micro) step and forget systems.  I want more.

I'll probably just go with the 8833, build my own current sense amps, appropriately trigger the a/d, and do the current loop in software.

Thanks for the input. 

Gary

Hi Gary,

One clarification. The DRV8829/28 is not limited by the 5 current control bits. It also has the ability to control the current by varying the VREF input directly.

For instance, if you have two 10bit DACs, each DAC output can be fed into a DRV8829 VREF input with the Ix pins set to a logic high. This will allow the change at VREF to directly control the chopping current, with 1024 settings.

The equation on page 8 of the DRV8829 datasheet shows how to determine the chopping current.

There may still be a few items to work around. The PWM chopping frequency is fixed, the accuracy may not be what you desire, and a BEMF measurement would have to be created.

Hi Rick,

Yes I understand all that.  I'd rather just generate the PWM signals than have to put DACs into the system.  I need the current measurements anyway for the BEMF position calculations.

The 8711 is interesting but you can't independently control the two phases.  If you could control the index along with the torque 9which you can control) it would be pretty good.  Then if you brought out the current sense signals so that I could digitize them it would save me even more.

Thanks for the help.

Gary

Gary,

You may like to consider DRV8412-EVM to control stepper motor using vector control. It comes with C2000 piccolo series MCU TMS320F28035 (support upto-60MHz) and provision for motor phase current sense feedback, which can be read by ADC on TMS320F28035.

In vector control d & q axis current control (being in synchronous reference frame) , is anyway implemented in software. This can be easily programmed in TMS320F28035 and required PWM can be generating to drive DRV8412 on board. Although, we do not have any application note on this topic but I think this kit can be used for vector control of stepper motor without any issue.

Best Regards

Milan-Motor Application Team

Hi Milan,

Thanks, I had overlooked this part.  Having separate ground connections for each output allows for a two resistor and simple differential amplifier current sense.  This avoids sign errors near zero current.  Important if you are trying to accurately model position everywhere.

On resource poor processors I've done the current control down at the winding in analog hardware.  Not ideal, but for slower systems it works fine.

Thanks again.

Gary