Set up using AMC AB15A100 drives, brushed DC motor Prototrak Plus with Encoders

More
31 Aug 2022 12:07 #250854 by Todd Zuercher
The end result is often a compromise. The art of tuning is finding the compromise that will give you the best possible performance for your machine.

So what effect does adjusting Pot1 have on the plot? It would be useful to see a series of plots, with one or two turns of pot1 clockwise and one or two counter clockwise, all with the same SW2 settings. To try to get a feel for what is going on.
The following user(s) said Thank You: arvidb, new2linux

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

More
31 Aug 2022 13:47 - 31 Aug 2022 13:51 #250857 by new2linux
Todd, Many thanks! Pot 1 setting does not do as much as I was hoping for, attached are many traces (in .zip folder: too large to attach) with SW2 setting as per file name in reference to chart mF. Attached are 3 (1 of each SW2 setting) pics that show very little trace improvement from the traces all ready posted. The slope of the trace (I need experienced eyes to understand) does not improve and the sharp corner does not seem to appear. I found that it was helpful to set both drives to same (SW2 settings) settings, the "x" axes (I am working the "y" axes) was traveling (I had already try Pot4) but after keeping the SW2 the same on both drive seemed to help with this. If you compare these attached files to the ones already posted there is still a spike/hump that is over travel.

Many thanks!

Edit: The RIgol o-scope is out of the system, and all the parts back in place.
Attachments:
Last edit: 31 Aug 2022 13:51 by new2linux.

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

More
31 Aug 2022 14:24 #250863 by Todd Zuercher
That looks like moving in the wrong direction. What happens with a greater number of turns of pot1 (rather than lower)? (Assuming the numbers in the image names represent CW turns from the CCW stop.)
The following user(s) said Thank You: new2linux

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

More
31 Aug 2022 15:14 - 31 Aug 2022 15:32 #250865 by new2linux
Todd, Many thanks. I am going to try the .zip folder (1.4 MB) by its self, see if that helps. If you compare the other traces it gives a better prospective.

Many thanks.

Edit:The folder worked, the Pot settings, I started at mid-point (7 turns) & used paper & pencil (old school) to track the number of turns, so I can reset to mid-point, for next settings..

Edit: This is the process to get trace (in case there is a easier way) I will power back up after setting change; at "terminal" screen (copy&paste) "~/linuxcnc/configs/test/test.demo" & the "test screen" & "HAL Scope will open, resize "test" screen & reset to 0.1Mz; resize the HAL Scope screen; start the "enable servo drive" and wait until the a "change in direction of axes" PUSH the "single" button under "run mode" so the trace will start & be 1 long (as shown) vs. one trace that is broken in the middle. If I have not have it setup correctly say so.

Many thanks!
 
Last edit: 31 Aug 2022 15:32 by new2linux.

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

More
31 Aug 2022 15:51 #250866 by Todd Zuercher
I don't think you uploaded the zip file you were intending. This one looks like it has the traces from the Rigol scope, when you were tuning the current loop.
The following user(s) said Thank You: new2linux

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

More
31 Aug 2022 15:59 - 31 Aug 2022 16:20 #250867 by new2linux
Todd, that was why that folder loaded, it was the wrong folder as you suggested. This is 1/3 of the folder, .many thanks.
Attachments:
Last edit: 31 Aug 2022 16:20 by new2linux.

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

More
31 Aug 2022 16:10 #250868 by Todd Zuercher
The maximum file size (or photo size) for the forum is 1600kB. So a 1.4mb zip file "should" have worked.

Another option would be to upload the images to a sharing site and post a link. That would be more helpful to others looking at the forum than emailing directly.
The following user(s) said Thank You: new2linux

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

More
31 Aug 2022 16:22 #250869 by new2linux
Todd, this is the 2nd of 3 of the files.
Many thanks!
Attachments:

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

More
31 Aug 2022 16:24 #250870 by new2linux
Todd, Many thanks!! This is last of 3 .zip folders.

Many thanks!
Attachments:

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

More
31 Aug 2022 16:46 #250871 by Todd Zuercher
The image "sw1.147pot1 7.0" is certainly the best of that lot.

Is seven turns as far as pot1 will go? If it will go further, you may find that increasing the ohms with SW2 is needed to tame the overshoot caused by a more aggressive setting with pot1.
The following user(s) said Thank You: new2linux

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

Moderators: piasdom
Time to create page: 0.148 seconds
Powered by Kunena Forum