VisSim/ECD example "pmsm2motor-28035"

Altair Forum User

 

Submitted by lukasovich on Thu, 12/08/2011 - 05:38 

 

I have spent huge amount of time to make this example working, but I still don't know how. I use TI PFC & Motor Developers Kit, with Anaheim BLY172S-24V-4000 motors: http://www.anaheimautomation.com/products/brushless/brushless-motor-item... included with Kit.

Problem occurs in sensorless mode. I followed instructions for motor starting, so I first started motor in open-loop until it reached a least 0.25 of max speed. Then, I turned on sensorless mode and then motor stopped still producing noise (I suppose that motor fall out from synchronism).

I tried everything, changed parameters but still got the same error. So, could you write the details about this? I need to know:

1 Did you use TI PFC & Motor Developers Kit with included Anaheim BLY172S-24V-4000 motors?

2 When displaying SMO estimated angle and generated RAMP angle (used in openloop mode) is there phase shift between them or there are in phase? In my case estimated SMO angle and RAMP angle were not in phase, there was phase shift between signals.

If someone experienced the same or similar problem with this example, please answer to this post, it would help me a lot.

Altair Forum User

Submitted by Anders89 on Sat, 12/10/2011 - 04:19.

Sorry to hear you are having difficulties with the pmsm2motor-28035. This diagram works fine in our labs with the TI Multi-Axis DMC EVM Kit. You can start in sensorless mode and the diagram has logic to spin the motors up in open loop then automatically switch to sensorless. As shipped this diagram can only control speed from about .15 to .6 full speed.
TI has a newer motor kit, the DRV8312EVM, which has a newer more robust MOSFET stage. The VisSim diagram for that board has enhancements that make the sliding mode observer gains, Kslf and Kslide, a table lookup based on motor speed. Additionally there is logic to automatically switch to openloop operation below .17 full speed. This allows smooth operation of the motor from full stop to .98 of full speed. The sensorless algorithm starts to lose angle tracking above .98 of full speed after 10-20 secs of operation but can be restored by backing down to below .95 full speed.
There is a small phase shift between the open loop and SMO estimated angle when operating in openloop mode. This should grow smaller with increased motor speed. When operating in sensorless mode there is no correlation between the openloop angle and estimated angle.