torque regulating with an VFD

If u use torque regulation,

then you need todo a motor identification (in sensorless vector)

Is this only necessary for torque or also if you use speed regulation ?

Are you refering to the auto tuning function to identify the motor?

You should always auto tune when using either open or closed loop vector.

You need to have the drive in current or torque mode.

In this case the torque is roughly proportional to the current flowing to the motor. This can be done 'open loop'. If you can sense the current it would be better. However, the a current going to the motor is often chopped so an instantaneous reading is meaningless. What you need to get is what the drive thinks is the instaneous effective current or modeled current. Don't bother to average the current yourself. The phase delay due to filter will be to much.

Modeling will just give you an idea of the ratio of the torque to the inertia. It will not provide the absolute value. If you know the inertia then you can calculate the torque.

As Leadfoot said, you should do a motor identifaction run (torque loop autotune) no matter what mode you run in.

Originally posted by Peter Nachtwey:
In this case the torque is roughly proportional to the current flowing to the motor.

Click to expand...

Be careful with this. While there are some manufacturers out there that will perform sensorless FOC on a BLDC motor, most of the time sensorless FOC is performed on an asynchronous motor. With an asynchronous motor you need more information than just current to calculate torque.

As Peter said, even if you get a calculated torque output from the motor it may not really be the shaft torque. It is really more a reflection of the tangential force on the rotor. This force needs to accelerate the rotor in addition to supplying shaft torque. If you have ever compared the calculated torque of an unloaded motor to its speed you can see what is being talked about. You can see significant bidirectional torque with no readily apparent speed change.

So if your speed is 'relatively' stable and your drive velocity loop is 'relatively' loosely tuned you can have more confidence that the torque output from the drive is equal to shaft torque.

Keith

I don't know what brand environment you guys are working in but, in my world of sensorless vector and DTC, I expect to see zero torque reported on an unloaded motor.

I agree that torque and current are not "roughly proportional" in an induction motor except at and above full load. The error increases rapidly as the motor reduces load until, at no load, the motor is typically drawing 25-30% of FLA.

Originally posted by DickDV:

I don't know what brand environment you guys are working in but, in my world of sensorless vector and DTC, I expect to see zero torque reported on an unloaded motor.

Click to expand...

Take the 'unloaded' motor and accelerate from from 0 RPM to -1750 RPM to +1750 RPM and back to 0 RPM in 1 second, recording the drive's picture of motor torque through that profile. I suspect the drive will not report 0 torque, even though there is no load on the shaft.

Unless ABB DTC drives are significantly different that everyone else, the 'torque' reported by the drive is really more of a force value tangent to the rotor. The drive itself will not subtract the torque required to accelerate the motor's rotor from the total' torque' it is producing.

Keith

Sounds like I misunderstood you. At steady speed I expect the drive to report essentially zero torque. Of course, it takes torque to change the motor speed so I would expect the drive to report that.