Experimental Physics and
Industrial Control System

Mark Rivers <[email protected]> · Wed, 24 Feb 2016 12:32:27 +0000

> Do any solutions to this issue exist in software? I would think that one
> could come up with epics channels holding the "real" position data (and
> only changing them when the particular axis is actually used).

I was wondering about that yesterday too.  It's possible that the AG_UCAxis class could keep track of where it really is.  We are only using this controller to position mirrors for alignment, so the absolute position is not important, we just tweak the position.  But if you need absolute position then you are welcome to modify the driver and if you can fix it then issue a github pull request and I will merge your changes.

Mark

________________________________________
From: Gerrit Kühn [[email protected]]
Sent: Wednesday, February 24, 2016 5:06 AM
To: Mark Rivers
Cc: [email protected]
Subject: Re: Motor examples missing devSequencer

On Tue, 23 Feb 2016 16:23:16 +0000 Mark Rivers <[email protected]>
wrote about RE: Motor examples missing devSequencer:

MR> This is from my startup script for an AG-UC8 8-channel controller.  I
MR> only defined 5 of the axes for now.

Thanks again, this looks like it should be easily adaptable to our devices.

MR> One thing you need to be aware of is that the AG_UC8 is a very dumb
MR> controller.  It works fine for doing incremental moves on any motor.
MR> However, it really only has 2 position registers.  That means that
MR> when you do a move on channel 1, all of the odd number channels will
MR> appear to be moving, their position readback changes.  Similarly when
MR> you do a move on channel 2 all of the even number channels appear to
MR> be moving.  So absolute positions are meaningless with this controller
MR> if you use more than 2 axes.

And another big thank you for this warning, otherwise I certainly would
have started wondering about what is going on these, soon.
Do any solutions to this issue exist in software? I would think that one
could come up with epics channels holding the "real" position data (and
only changing them when the particular axis is actually used).

cu
  Gerrit

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System