Incremental Encoder - misalignment - speed error?

[powersys]

eccentricity error encoder

An incremental encoder is attached to the motor shaft, and the output signals from the encoder is used to calculate the speed of the motor. If there is some misalignment between the shaft and the encoder, do you think the calculated speed will be constant or consistent? Assume that the motor speed is constant.

Besides, if the encoder is wobbling, do you think the calculated speed will be constant?

 

[FvM]

eccentricity encoder

I wouldn't expect errors below a considerable level of misalignment or mechanical interferences.

 

[Jmiller]

how to caculate speed from an incremental encoder

It would depend on the type of encoder. If you are using some sort of modular encoder where disk alignment is dependent on the motor shaft and motor bearing set, the disk may not be concentric throughout rotation. This can cause the pulse cycle width to expand and contract over rotation. The severity of this depends on the eccentricity.

Whether or not this will impact the RPM measurement will also depend on the way the controller/PC calculates speed and how often it updates.

It is good to keep in mind that any cycle width error gained is also lost by that same amount over the full rotation. If the electronics/software average out the RPM over several rotations. Eccentricity error is likely not going to be a factor.

I work for an encoder manufacturer. We make incremental encoders where the disk is held in tight alignment to the sensor by a set of bearings internal to the encoder. this along with the interlaced sensing scheme we use eliminates rotational eccentricity errors that could be caused by misalignment.

Any questions feel free to contact me.

Jim Miller
Application Engineer
Quantum Devices Inc.
jmiller@quantumdev.com

 

[jorgito]

Hello,

It will not depend on the encoder type, but on the coupling type.
If you have a cardanic coupling, the speed will vary along a complete turn, depending on the misalignment angle.
Obviously, the mean speed, integrated in a complete turn will be constant and equal to the motor speed.
If you use an elastic type of coupling, you can have torsional resonances, and the mean value of speed will be (again) equal to the motor speed. However, as the resonances are not ruled by geometric conditions (as in the former case) you can not warrant that integration on a complete turn will give you the exact motor speed.
With a bellows type of coupling, you can assume that the measured speed is the same that the motor speed.

Hope this helps and best regards!