Page 1 of 1

Unexpected results from toolpath

PostPosted: Sun Nov 10, 2019 12:53 am
by Soundbender
I began troubleshooting a problem with a Fusion toolpath that was generated using the UCCNC post for a part with a 3D profile, milled on my Bridgeport BOSS converted to servo/eth400/UCCNC. I've been running the machine for some time now and UCCNC has worked wonderfully and produced the expected results time and again. The post on the Fusion site here: https://forums.autodesk.com/t5/fusion-3 ... 505#M71606 identifies the problems and what has been done to date. I have verified that the machine is in perfect working order and after very minor compensation is free of backlash. Nevertheless, the part is being cut very poorly. The toolpath has been gone over carefully and I see no obvious reasons for the part finish that is being achieved. I've noted that the UCCNC path visually follows what the cut part looks like, but the cut part does not appear to follow the toolpath.

This video shows the tool following a scalloped path:
https://photos.app.goo.gl/GAVUoHaeyFVZPCnr6

But a review of the toolpath shows no such features. (image Capture3, attached)
Capture3.PNG
The toolpath viewed in NCPlot


The part does, however.
Capture1.PNG
The mis-cut part

Here you see the scalloped finish as well as obvious gouging in the background - each of these features appears repeatably in the part.

The Fusion forum appears to feel that it's a machine problem. I'm wondering if UCCNC is accurately following the toolpath. The video above seems to support the idea that UCCNC believes it should be scalloping the path for some reason. The toolpath does not appear to follow the visual in the video or the part that gets created. I've attached the toolpath in case someone sees what I don't.

Re: Unexpected results from toolpath

PostPosted: Sun Nov 10, 2019 3:10 am
by cncdrive
UCCNC goes off the programmed path for possible path optimisation with maximum the distance you allowing it to do so.
Check your Constant velocity parameters on the General settings page. Also read the users manual to understand how the path optimisation and these parameters work.

Re: Unexpected results from toolpath

PostPosted: Sun Nov 10, 2019 4:16 am
by Soundbender
Thank you. The machine is configured in Exact Stop mode, so the path should have been followed explicitly. The Gcode does not call G61.

Re: Unexpected results from toolpath

PostPosted: Sun Nov 10, 2019 4:21 am
by cncdrive
In G61 there is no path optimisation, so in this case it should be a mechanical problem of the machine. Or if it uses servos and not steppers then it could be a too softly tuned servo position PID loop.

Re: Unexpected results from toolpath

PostPosted: Sun Nov 10, 2019 4:30 am
by Soundbender
It is a servo machine, and I might agree that the servo PID loop may be tuned too softly, but that would not explain the video showing the tool in the wireframe display of UCCNC tracing a path that was not as described by the Gcode and the resultant machining in the part. I will review the PID on the machine as you have recommended, but the display indicates something else may be wrong. I'm only looking to solve the problem be it my fault or otherwise. There are many settings in UCCNC and I may have set one of them improperly, so I hope only to resolve the problem.

Re: Unexpected results from toolpath

PostPosted: Sun Nov 10, 2019 4:49 am
by cncdrive
The UCCNC has no idea what the servos doing, so ofcourse it is possible that even your servo is disconnected and there is no movement while the UCCNC is commanding it to move along the path when zero movement happens.
The same, if the servos not doing what the UCCNC is telling them to do then the UCCNC does not know nothing about that...

Re: Unexpected results from toolpath

PostPosted: Sun Nov 10, 2019 1:17 pm
by Soundbender
Leaving the machine in Exact Stop mode, I set the path planning parameters to .0001". These settings should not be in effect in Exact Stop mode, however this was the result:
IMG_20191110_071313.jpg
Exact stop with low tolerance in path planner


Thanks for the clue that helped solve the problem.