Follower with encoder or no ?

[Masbibe]

Hello everybody,
I have a one question for you. If I use two AC induction motors mechanically coupled with 2 ABB drives ACS800 (direct torque control) and if they are working in master follower configuration is it better to put encoder on follower or there is no sense to do this.
Thanks.

 

[Skogsgurra]

If mechanically coupled, I would avoid an encoder on the follower. The reason is that a follower works much better as a torque slave than as a speed controlled slave. If speed controlled, you will have problems with load sharing as soon as anything isn't 100 percent perfect in the system.

Gunnar Englund

http://www.gke.org

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

 

[Masbibe]

Skogsgurra,
OK but I meant that follower should be torque controlled (sorry because I was not precise) and in this case do I need encoder. So the main question is if drive that is torque controlled slave use speed information from encoder in torque algorithm. Concretly I mean on ABB ACS800 drives.
I make this question because as I learn DTC dont use speed information in torque algorithm and when I look behavior of ACS800 drive it seems to use it.
Thanks.

 

[Masbibe]

Also,

In some firmware manual it is written exatcly this:

"Pulse encoders are recommended to be used in all of the torque controlled Followers"

 

[DickDV]

My experience with DTC and ACS800's, an encoder on a torque follower is not needed.

If you need better than 10% of motor nameplate slip speed error, you should use an encoder on the lead drive (speed regulator) only.

 

[DickDV]

In my experience with DTC and ACS800's, an encoder on a torque follower is not needed.

If you need better than 10% of motor nameplate slip speed error, you should use an encoder on the lead drive (speed regulator) only.

 

[DickDV]

Oops! Sorry for the double.

 

[Masbibe]

OK, it is obvius that if you need high precision of speed and if you put encoder on master drive (speed regulated) than you will have high precision of speed because master know exactly how much is speed and its speed regulator will do the job about speed. So because you have mechanically coupled follower it will be on same speed with no need for any encoder.

But,

I just wondering if torque algorithm use speed information than MAYBE if you are on some small speed with high torque required, follower will not do his job correctly ( real torque on shaft will not be as drive think) and then master will need more torque or you will have some oscillations.

So this is my theory why ABB recomend that follower should have encoder. Probably I am wrong, but never know.

I dont believe that they recomend encoder on follower only because some speed limits or in case shaft breaking. For me estimated speed is good enough to do this.

 

[Masbibe]

I only want to add one more thing.

Some of you probably see this situation.
If channels A and B on encoder card on ACS800 are mixed (positive torque generate speed that drive see as negative) and drive is in speed control and you give start command with some positive speed reference shaft will oscilate around zero speed.

This behavior is what is very strange to me.

To me it is logigal that in this situation drive generate positiv torque and motor should start rotating in positive direction and because channels are mixed drive will see that speed is negativ. Then drive should generate more and more positiv torque and speed will be more and more negative (in drive) and at the end you will have some encoder fault.

But drive not react like this. Motor oscilate around zero speed.

So to me this behavior tells me that drive use speed information in torque algorithm.

Correct me if I am wrong.

 

[Masbibe]

In post above I meant that motor is unloaded.