Raspberry Pi 4 with Mesa 7c81
05 Apr 2021 23:06 #204983
by JohnG
Replied by JohnG on topic Raspberry Pi 4 with Mesa 7c81
The machine does not lose position.
It is not exactly repeatable in that it does not do the same thing each time but it does do something at the same location of the pass if you are making multiple passes to say cut a part out. I’ll try to get an example picture
It is not exactly repeatable in that it does not do the same thing each time but it does do something at the same location of the pass if you are making multiple passes to say cut a part out. I’ll try to get an example picture
Please Log in or Create an account to join the conversation.
05 Apr 2021 23:15 #204987
by andypugh
Replied by andypugh on topic Raspberry Pi 4 with Mesa 7c81
My first thoughts are that the 7C81 is too "dumb" to do this.
I assume that the backplot does not show this?
There are many possibilities, including mechanical. Check that all your couplings are tight. Though it doesn't feel like a max-force point or a max-speed (backlash) point.
I assume that the backplot does not show this?
There are many possibilities, including mechanical. Check that all your couplings are tight. Though it doesn't feel like a max-force point or a max-speed (backlash) point.
Please Log in or Create an account to join the conversation.
05 Apr 2021 23:20 #204988
by JohnG
Replied by JohnG on topic Raspberry Pi 4 with Mesa 7c81
Definitely not mechanical. It is obvious when watching it that it is commanded motion...
You can see in this photo what it has done after multiple passes. Not exactly in the same spot but same type of error.
You can see in this photo what it has done after multiple passes. Not exactly in the same spot but same type of error.
Please Log in or Create an account to join the conversation.
05 Apr 2021 23:21 - 05 Apr 2021 23:26 #204989
by PCW
Replied by PCW on topic Raspberry Pi 4 with Mesa 7c81
When you say it does not lose position, are you saying
the tool returns to the correct path after the deviation?
Maybe a simpler test code would help diagnose the issue
What are your min_ferror and ferror settings?
large values of these can mask steggen setup errors
Posting your hal and ini files would also help
the tool returns to the correct path after the deviation?
Maybe a simpler test code would help diagnose the issue
What are your min_ferror and ferror settings?
large values of these can mask steggen setup errors
Posting your hal and ini files would also help
Last edit: 05 Apr 2021 23:26 by PCW.
Please Log in or Create an account to join the conversation.
05 Apr 2021 23:22 #204990
by JohnG
Replied by JohnG on topic Raspberry Pi 4 with Mesa 7c81
And correct backplot does not show it. Also verified on standalone gcode simulator
Please Log in or Create an account to join the conversation.
05 Apr 2021 23:29 #204991
by JohnG
Replied by JohnG on topic Raspberry Pi 4 with Mesa 7c81
I wrote a simple gcode program to test. It was a rectangule with rounded corners I used g02 for the curves. Again, perfect plot on screen but the issue you saw in the first photo labeled “A”
Please Log in or Create an account to join the conversation.
05 Apr 2021 23:34 #204993
by JohnG
Replied by JohnG on topic Raspberry Pi 4 with Mesa 7c81
I will look up the configs in just a minute as well post the gcode I ran for the rectangle
Please Log in or Create an account to join the conversation.
05 Apr 2021 23:51 #204997
by JohnG
Replied by JohnG on topic Raspberry Pi 4 with Mesa 7c81
Ferror= .5
Min_ferror= .05
Will post both .Hal and .ini
Min_ferror= .05
Will post both .Hal and .ini
Please Log in or Create an account to join the conversation.
05 Apr 2021 23:54 #204998
by andypugh
Replied by andypugh on topic Raspberry Pi 4 with Mesa 7c81
I agree that it looks like a command motion (even on the part) but if it does not show up on the backplot then I _think_ that can only be backlash comp (bad backlash comp).
Please Log in or Create an account to join the conversation.
05 Apr 2021 23:59 #205001
by JohnG
Replied by JohnG on topic Raspberry Pi 4 with Mesa 7c81
I would agree except that I created the rectangle specifically to make the motors run in only one direction throughout the curve (x only moving in the negative direction and y only moving in the positive direction) to eliminate backlash as the culprit. I really do not believe it is backlash or backlash comp (which is set to zero anyway)
Please Log in or Create an account to join the conversation.
Time to create page: 0.085 seconds