Mesa 7i77 and linear encoder tuning PID.

More
13 Aug 2012 16:36 - 13 Aug 2012 16:38 #23095 by ROG
I’m looking for some advice / opinions on the halscope traces below.
This is my first venture into PID tuning / setting up CNC control.
The axis in question is the X axis on a Hardinge HXL lathe.
Mechanically, the axis has been completely re-built with new ball screw bearings and the ballscrew re-balled / checked. I’ve tried a variety of Gib settings. Even belt tensions.
Feedback is from a Heidenhain linear encoder with 10 micron grating pitch via 256 X interpolation giving a resolution of 0.0000390625 mm at 25600 counts per mm.
Single ended encoder output is connected to a Mesa 7i77.
The axis is driven by an Omron R88D 400W AC servo and drive.
I’ve spent countless hours over countless weeks in ever decreasing circles changing parameters on the Omron drive, interpolator, mechanical elements and parameters in ini / hal .
This seems to be about as good as it gets or as close as i can get.
I’ve concentrated on linear moves from 1-100mm / min as these are the sort of feederates it will spend most of its time cutting at.
I have no idea if I should be expecting more or if this is as good as I should expect to get?
No matter what I do, I can’t get rid of the spike at the beginning of the move.
During the cruise period, Is the error waveform likely to be mechanically induced? Or is this oscillation in the PID loop?
The servo drive can output 2048 quadrature pulse per revolution. If I connect this to a spare encoder channel and use this as velocity feedback, would this help? Or would that simply be telling me that my following error was better, when in reality, the problem is still there but just masked?
Any advice or comments really appreciated.

1000 mm/min
Attachments:
Last edit: 13 Aug 2012 16:38 by ROG. Reason: images wrong

Please Log in or Create an account to join the conversation.

More
13 Aug 2012 16:40 #23096 by ROG
100 mm / min

Attachments:

Please Log in or Create an account to join the conversation.

More
13 Aug 2012 16:41 #23097 by ROG
50 mm/min
Attachments:

Please Log in or Create an account to join the conversation.

More
13 Aug 2012 16:50 - 13 Aug 2012 17:07 #23099 by PCW
Its hard to tell scale (since the analog out and error spike are both off scale) but at first glance the cruise portion look good (better than .002 mm or ~1/12000 ") But the acceleration settings are far exceeding what the mechanics/drive are capably of.

scaling HALScope so everything remains on screen would help diagnose the trouble, as would plotting the velocity

Also on the last (100 and 50 mm/min) The initial spike looks like stiction to me since there is no equivalent spike on stopping
Last edit: 13 Aug 2012 17:07 by PCW. Reason: addition

Please Log in or Create an account to join the conversation.

More
13 Aug 2012 17:02 #23100 by ROG
Hi Peter
I left them all at the same scale for comparison.... I’ll post new traces shortly with the correct scales and velocity ....
I’ll also check what acceleration is set to.

Please Log in or Create an account to join the conversation.

More
13 Aug 2012 17:35 #23102 by PCW
Another thing that might help is to zoom in on the spike area especially the large spike when starting a 1000 mm/min move

Please Log in or Create an account to join the conversation.

More
13 Aug 2012 18:03 #23103 by ROG
50 mm/min

Attachments:

Please Log in or Create an account to join the conversation.

More
13 Aug 2012 18:05 #23104 by ROG
50 mm / min zoomed start
Attachments:

Please Log in or Create an account to join the conversation.

More
13 Aug 2012 18:06 #23105 by ROG
50mm /min zoomed end

Attachments:

Please Log in or Create an account to join the conversation.

More
13 Aug 2012 18:07 #23106 by ROG
1000 mm/ min
Attachments:

Please Log in or Create an account to join the conversation.

Time to create page: 0.393 seconds
Powered by Kunena Forum