AC induction motor torque at zero speed with vector control 2

[rulmismo]

Hi,

After some months reading about vector control I have been looking for information about how is the real performace of the main variables of the motor (currents, field synchronous speed) when it is commanded torque at zero speed (i.e. as holding brake).

I would like having an insight in the value of field speed and currents that are needed to keep the rotor at still with any kind of FOC (I mean order of magnitude respect to nominal motor values).

I couldn´t find any waveform or discussion about this particular situation, nor google, nor anywhere.

Maybe if you guys have used some real drive that can do that might help me, thanks!!!

 

[jraef]

I used it once to lift and hold Boeing 777 aircraft, does that satisfy your curiosity? A good high quality FOC or Closed Loop Vector drive can and will provide full torque at zero speed. The current, aside from being dependent on load, can be as high as FLC. For that reason, the motor must have separate cooling (IMHO) if that is your intended application. When I did that project, we did not do any waveform capture etc. so I'm no help there. But a critical feature was what they called "torque proving" which made sure the motor was at full output torque before releasing the mechanical brake. We didn't think dropping a 777 was going to be an acceptable error...

A big application for this is in hoists, you may try doing a search of "vfd drives on hoists" to see more suppliers and information. I used the Magnetek Impulse G3 drives on that application, they worked flawlessly and I think they have the most experience in this application. It is a Yaskawa drive, but they add custom software and the experience.


"If I had eight hours to chop down a tree, I'd spend six sharpening my axe." -- Abraham Lincoln
For the best use of Eng-Tips, please click here -> faq731-376

 

[davidbeach]

The motor is run in the "raise" direction at a speed equal to the slip speed at the torque necessary to hold the load at stand still. In vector control (not in V/Hz control) you can get full torque at any speed. Without an encoder the drive won't know exact shaft speed so the load might drift slowly in either direction. With the encoder there will probably be just a bit of jitter as the load moves back and forth between two adjacent points on the encoder.

 

[rulmismo]

Thanks for your reply. It is really an impressive application.

I have seen very interesting information about the subject in
thread237-149826
thread237-220699

Frome there I read that seems to exists drives that are capable of sensorless full torque at zero speed, but normally an encoder is needed to hold still the rotor position.

Also understand that the flux must be slow rotating, in order to the slip will be enough to induct the needed torque and flux producing component (should be slow slip, but how much? similar to nominal speed <10%?).

Thanks,!

 

[DickDV]

rulismo, it should be noted that not every VFD that claims to have sensorless vector control is capable of developing any torque (much less full torque) right at zero speed. Or, to put it a little differently, there are huge differences in performance among sensorless vector drives especially at zero speed.

In my view, ABB's Direct Torque Control ACS800 drive is the best out there with the most recent version of Mitsibushi's sensorless vector being a close second.

As mentioned, the addition of an encoder pretty much makes the differences go away and most can do a very good job at zero speed.

 

[Skogsgurra]

rulmismo, Yes the slip is identical at all speeds. Voltage and torque are what determines slip. But usually more in the ">0 %" range than in the "<10 %". Modern high efficiency motors are close to one percent.

DickDV, read thread237-220699 carefully. It says something about the ACS800 that may be news to you. There are other drives that behave a lot better at zero RPM.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[rulmismo]

Thanks guys! very valuable information, there are a lot of confusion out there about this subject.

 

[DickDV]

skogsgurra, thanks for that thread reference. I read it carefully and have the following observations:

1. The thread is primarily about using an ABB ACS350 drive in sensorless vector mode on a hoist application. I think we would both agree that this is an unacceptable choice of VFD. Personally, in the ABB drive family, I would only choose the ACS800 with DTC capability for hoisting.

2. I'm not completely clear on your comment about learning something new about the ACS800. It may be that you were directing me to the short discussion of other motor models which I found very interesting. While I have to use the capabilities of the various systems every working day, I am not qualified to comment (much less understand) the significance of these different motor models and how they are conceived. I have to focus on the end results.

3. It may well be that you were directing me to the capabilities of the Optidrive system. I am not familiar with them here in North America but it certainly sounds like they have capabilities that, for hoisting applications, are excellent. While several manufacturers (I am aware of ABB and Magnetek/Yaskawa/Electromotive) providing special software specific to hoists/elevators, I believe that all of them specify encoders for safety purposes regardless of their sensorless performance. I'm going to try to identify a distribution channel for Optidrive in North America and get some more info.

Thanks for the wider perspective. I'll add Optidrive to my "short list" of drives with good zero speed control capabilities.

 

[Skogsgurra]

Yes, I meant that the 'news' about ACS800 was that there are other options.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[cswilson]

In addition to jraef's Boeing 777 hoist application, the hoists NASA uses to lift the Space Shuttle on and off the 747 for transport employ vector-controlled induction motors, and yes, they do care about producing full rated torque at zero speed.

Electromagnetically, the rotor "sees" the slip frequency regardless of its mechanical speed. So if a 4-pole motor produces rated torque at 1740 rpm when operating DOL at 60 Hz (2 Hz slip frequency), it will produce this same torque at standstill with 2 Hz applied (at a reduced voltage that keeps the V/Hz the same).

I do believe you need a true rotor angle sensor to do this reliably. Back EMF detection falls apart at very low speed, and while there is interesting research into other methods of sensing from the electromagnetic properties (reluctance/inductance variation and the like), these tend to be very motor-specific, and real-world users don't trust them.

Curt Wilson
Delta Tau Data Systems