Hello!
I need to learn about the sensorless control of permanent magnet AC (PMAC) motors. Who of you can recommend a tutorial and/or open source code for the sensorless motor control using the
a) TI TMS320 series processor, or
b) Microchip dsPIC33EP128 series processor?
Thank you!

I am not familiar with either a or b above but if you provide more info as to the type and application of the motors it might help.
Are they servo, high speed spindle, automotive 12vdc, miniature, high HP etc?

Michrochip has an app note on the topic. Forgot if it was for pm or im though. Ti must have something too. Simulink has demo. Once understand the implementation is just a matter of convenience.

I don't know Microchip dsPIC series very well.
TI has a long history in motor and power control. The C2000 series has probably the most flexible (and most complex) peripherals.

Although TI has this new "closed" source code for motor control, instaSPIN, there are many old documents, dev kits and examples (Digital Motor Control suite) which are open source and easy to modify.

However, if you are going to develop motor control for a quite long time, I'd take a look at documents other than application notes...

Hi Everyone! I am NICK from Rhymebus Corporation in Taiwan.We have specialized in AC motor inverter for 26 years. Our products are applied in various industrial equipments. Especially, Grundfos pump is also our client.You are welcome to contact with me at anytime.

I think TI and Microchip use sliding mode observer (SMO); ST uses Luenburger Observer (LO). All works fine and has lots of field application. Application notes and source codes can be found on their official website. Among them, Microchip's notes are the best in my view. Hope it helps.

I have used Microchip and TMS320 to develop VFD for various application. To give you a top level view, The TMS320 is a high end solution - a lot of commercial VFD uses TMS320 DSP. Quite a bit of work in programming, but if you go to the Ti web site and look at their new InstaspinFOC and InstaspinMotion, Ti have the entire motor control solution software implemented in ROM, you buy the chip, write a C program to call up their motor control function and you and a VFD. They provide you with a demo kit, PCB and a motor. It take me half a day to get the demo PCB running with my PMSM. Than I copy their design to my own.

The Microchip solution provide you with demo code. I used that before, but it require quite a bit of C programming, and motor tuning take even longer. The demo code and application note are no where near the performance of the Ti solution (I do not work for Ti -so I am not advertising). I take me a week to get my motor spinning with the demo kit from Microchip.

Than there are the International Rectifier solution that is available from many years. The IR sensorless motion control solution have implemented a the FOC motor control in FPGA. So you don't need to write code for motor control. In the chip, it also have a 8051 cpu. You write the program in C; 1 page of code will get a washing machine working. It take me 1 day to get a PMSM motor running with this solution.

I will use the TI solution for high end motor control - such as a US$40,000 dollar, 100HP direct drive PCP used in the oil field.

I will use the IR solution for a water pump, washing machine - things that is a few kw.

I will use the microchip for solution for toys, because Microchip is so much fun to play with.

Hi guys! It was a very informative discussion. Congratulations!

Hi Friends,
I saw lot of literature on Sensorless control of PMSM from stand still to low speed based high frequncy signal injection . But could not find an application note or implementation of this technique .

If any body come across this topic or working in this area , please guide me if there are any application notes / source code from TI or Microchip.

Thanks in advance
Ravi

Ravi,
I don't think you'll find any application notes on this. AFAIK, although they are not recent, few motor drive manufacturers implement this kind of techniques.
However, the schemes in the first papers on this topic (Lorenz) are quite simple and can work for a specific motor and application.

Hi Sandro,
Thanks for quick information .

Is there any reason why these techniques are not impleneted or seen much light in the industry.
Due to complexity ? But the advantages offered by these techniques may over come the complexity. More over these techniques are based saliency of motor they provide rated torque from zero speed.

Thanks
Ravikumar

Ravi,

Yaskawa and ABB both have implemented sensorless PMSM motor control in their standard drive. So they are available in the industry - off the shelve. I have been using the Yaskawa A1000 drive on a surface mount PM motor that has no saliency at all and the drive works just fine.

For low RPM control, I am able to run the VFD (without sensor) to 20 RPM with full torque on a 1000ft-lb 100HP motor. So I say the technology is very mature. That control does not even need the high frequency injection.

Townes