Motor Control

Hi rppolicy

We do motion control on a NIOSII (4 BLDC-Motors on a single Cyclone). Because we have rather high requests on dynamics and accuracy we use sensors.

There is a major DSP-Manufacturer having a lot of good application-notes about motor control, also for sensorless operation.

Chris

elan,

Thanks for the reply. I'm not trying to steal anything from you, but I was wondering if you did the commutation in hardware or software? The reason I am asking is that we have a platform that was designed by a mechanical engineering group within our plant. We work in a EE/SW group, and was asked to evaluate thier design. Their design consisted of two DSPs, one for motor control and the other as a general purpose controller. The design is only on paper right now, and none of the motor control has been written. After a preliminary review, we came to the conclusion that 2 DSPs was overkill. Now we are looking for additional ways of BLDC control. We have not ruled out the DSPs - they have their place and are an excellent method of motor control. The NIOS came in as a possible solution for cost and obsolecene issues, and I just doing a little homework for now. I was just trying to get an idea of how many BLDC motor designs use the NIOS.

Rick

Rick,

We are replacing DSPs with FPGA's because we get more flexibility and can use the same HW not noly for BLDC-Motor control - there are a lot of other motor types and sensors/actors in our machines.

<div class='quotetop'>QUOTE </div>

--- Quote Start ---

I was wondering if you did the commutation in hardware or software?[/b]

--- Quote End ---

It&#39;s a co-design: sine-commutation patterns are done in hardware - field vector control is done in software (not necessary in sensorless operation).

Chris

Probably because DSPs keep going obsolete!

FPGAs are much less of a problem because the old design can easily be moved forwards to a new part.

Hi,

until know I only know motor control with FPGA using position sensors. But in fact it shouldn't be a big problem to implement a sensorless control using the back emf. No matter if you are using DSP or FPGA. I mean it is "just" measuring voltage. I think with some efforts you can get this done in hardware.

The commercial soultions I know with positions sensor:

1. motionfire firefighter (Arrow), using Hall sensor for position estimation. Almost all motor control based stuff as current-, speed, position-control and phase commutation is done in hardware. but it is only block-commutated.

2. falconeye (ebv), using Encoder or Resolver for position estimation. Mainly the motor control is done in software using the NIOSII. But the software is partly hardware accelerated (sin/cos) and you will get hardware modules for interfacing your position sensors. This solution is a Field-Oriented-Control (foc) with Space Vector Modulation (svm). So you get higher precision within your control scheme.

3. There are many papers describing the foc and svm implemented completly in hardware. I think alizem (canadian?) selling such IPs. But I am not sure.

Regards

Thanks Pilleman,

Yes, Alizem sells application-specific Motor Control IP on Altera FPGA that can be tested on Arrow's MotionFire and EBV FalconEye.

Please find datasheet and white paper of our latest COTS product for pump and fan applications on Alizem's website.

For more informations, please contact our sales department.

Thanks for your interest in Alizem.

Jack,

You may have a look to this blog post to get an answer to your question.

http://www.design-reuse.com/industryexpertblogs/24744/motor-control-fpga-vs-dsp.html

Regards,