am-2816a encoder

[wildaburk3160]

Hi guys!

We are hoping to implement encoders in our drive system this year, and I would like to know if the am-2816a encoder from AndyMark is an absolute encoder. If anybody can tell me the type of encoder this is, it would be greatly appreciated.

Thanks,
William

[richwong]

It is not an absolute encoder. The "position" resets to zero each time you restart your robot.

[Kevin Zhang]

Does anyone know what the connector to this encoder is and possibly the part number and distributor site

[Alan Anderson]

Originally Posted by Kevin Zhang

Does anyone know what the connector to this encoder is and possibly the part number and distributor site

I think you want the am-2993.

[wildaburk3160]

Thanks for the help!

[richwong]

AndyMark does carry an absolute encoder now: http://www.andymark.com/product-p/am-2899.htm

[wildaburk3160]

Hey guys, I've been working with trying to get the motor to rotate 90 degrees clockwise and counterclockwise using this encoder, and I can't seem to get it where it will return to a specified position without a significant amount of over travel. And when I have gotten it down to an acceptable range, it has worked, but after 6 or 7 times of initiating a rotation the motor just stops. Has anybody else had issues with this encoder and having over travel issues?

[Ether]

Originally Posted by wildaburk3160

Hey guys, I've been working with trying to get the motor to rotate 90 degrees clockwise and counterclockwise using this encoder,

What does the pronoun "this" refer to? The am-2816a (title of your thread) or the am-2899 (mentioned in the post to which you are responding?

What motor controller are you using?

Are you commanding it with CAN or PWM?

Is the encoder wired to the motor controller or to the roboRIO?

What are your PID gains?

What motor are you using? What gear ratio? What load is the motor driving?

[wildaburk3160]

I am using the am-2816a encoder attached to the am-2161 motor. The encoder is wired into the DIO slots on the roboRIO and the motor is being run using PWM. Right now there is no load being put on the shaft of the motor.

We want the motor's home value to be zero and to be able to rotate 90 degrees clockwise and counterclockwise and to always be able to come back to its home position. However, when the shaft returns to home position, we have not been able to get it within an acceptable tolerance of angles.

I am not using a PID protocol but am scaling down the encoder pulse values returned to determine the output of the motor. Specifically, I get the encoder pulse value, then square root it, then it is further scaled down so that the motor output can never be given a value greater than 1. So basically, as the encoder pulse value gets closer to 0, the greater the reduction of the motor output. However, even with such a reduction, the motor still seemingly has too much momentum and still over travels even when the range of encoder values we have to tell the motor to stop is very small.

I have dabbled minimally with using PID protocols and don't have a full understanding of them currently, but do you believe that a PID system would be better and more precise than the system I have described? Or would another sensor such as a potentiometer be better for our purposes?

[Ether]

Originally Posted by wildaburk3160

do you believe that a PID system would be better and more precise than the system I have described?

Yes. Especially when you decide to put a load on the motor.

You will need to re-tune the PID gains when you put a load on the motor.

Or would another sensor such as a potentiometer be better for our purposes?

If the encoder is working properly, the problem is not the encoder.

An absolute sensor (like a potentiometer or absolute encoder) might be slightly easier to use, because with the relative encoder you need to tell it where zero is.

[wildaburk3160]

Could you point me to an example code or CD thread where PID is shown to be implemented, or a source that can explain the basics of PID?

[Ether]

https://www.google.com/?gws_rd=ssl#q....com+%22PID%22


Also look at the LabVIEW help and examples.

[wildaburk3160]

Thank you for your responses. I will look into this and hopefully end up implementing it in our drive system.