So the Shopbot is having issues to say the least. For the last week I have been tinkering trying to cut accurate parts. .25mm would be just fine if not slightly rough. as it stands, the Shopbot can’t get close to that though. It’s currently cutting poorly on the X/Y axis’
To replicate the issue:
Run a 1/4" 2 flute spiral upcut bit (brand new) 55-85 IPM .25 to .5 depth per pass
the parts BEST dimension on the x axis will be within .75mm of the design
the Y axis will be off by up to 1.5mm.
Circles less than an inch in diameter are visually not round.
Running a compression bit (brand new) on .75 or .5 in full depth passes further worsens the problem.
Materials cut were OSB, Marine plywood, Melamine shelving and solid pine. All had the same issues.
The Y axis has considerable play in it. When the stepper motor is locked in place, the Y gantry is still able to move back and forth a small amount visually. The problem likely lies with the very badly worn rack and pinion gears on the Y and the wobbly motor mount. When jogging the machine at 100 IPM you can see the entire Y axis gantry shaking back and forth significantly. The X does it as well but to a lesser degree. The mounts for the motors are flexible. And visibly shake when jogging or cutting. If forcefully held by hand, they are more rigid and and shake less while jogging.
possible proposed solutions:
Remake the motor mounts from thicker angle iron?
Do away with the spring tensioners and replace with fixed adjustable mounts?
Replace the Y axis rack and pinion. It’s long overdue as far as I can tell
Not too say that the shopbot couldn’t use some maintenance, or that there aren’t some pretty cheesy design choices that the mfr made, but there are a lot of things that can affect cut accuracy, many of which are best practices. Are you trying to take too heavy of a cut for the tools and material? Are you using a spring or finish pass to reach final dimension? Are the tools sharp?
I have tried finishing passes. It makes a difference, but not by much sadly. I have also tried taking some very shallow cuts to reduce the load on the gantry.
All the tools used were brand new.
So most of the problems appear to stem from weak motor mounts flexing under load. I noticed something when looking at a photo of someone elses mounts. Instead of using thicker angle iron you could just weld some bar stock around the edges of the mount and drastically increase its rigidity
It just seems insane that a serious company would design a thin motor mount that is anchored through a single bolt and tensioned by a spring.
Set screws all tight on the motors? There should be no free movement on the gantry from the rack and pinion. Motor moving or set screw? Not seeing how the rack gears could be wearing out.
I only looked at them to see if there was anything loose. The rack and pinion are definitely worn down. One side of each tooth has an enormous high spot and there is corresponding lowspot on the rack. I didn’t bother measuring it but it was probably about 1mm tall. As far as i can tell, this has been happening for a long time now or else people would have noticed metal dust falling on their parts.
It could have been loose set screws on the pinion for sure.
Personally, I am biased to the idea that the motor mounts are a major culprit. If they flex even a little bit, it shifts the pressure unevenly on the gears which would create weird wear patterns. And once its started i think it may be a positive feedback loop. Wear leads to more wear. If the diameter of the gear is smaller than when it was calibrated, logically it should be the reason for parts being too small in that axis?
Sorry, I didn’t see the posts from this afternoon.
Doug taught his shopbot class tonight, after James and I did a little troubleshooting.
We found that the rails for the y axis were out of paralel and the horizontal wheels that lock it down to the angle were a bit loose. We made some improvements on both counts. Though I have not heard back from doug about the machines performance as the class made their rings. The rack and pinion may be worn but works smoothly. I will continue working on it and look forward to finding out from Doug as well as yoursrlf and anyone else if tonight’s adjustments resulted in improved performance.
Haply making!
I will keep you posted. I know the machine runs moderately ok when tuned so hopefully it can last long before needing major maintnence. Maybe wednesday i will have time to do some more dimensional accuracy test cuts. I remember just 8 months ago it was cutting an 80mm part within .06mm of accurate.
What would you consider “good enough” for accuracy on this machine? If any of my parts meet that threshold I will stop trying to improve it with gcode
I’ve recieved good marks from Doug. He says its running “like a dream”.
I would imagine that the olb benchmark of .06mm would be a good target. Though I am not familiar enough to say what tolerances are generally deemed acceptable for this machine.
I suspect that a recent cataclysmic bit break started the alignment troubles and that they’ve just gotten worse since.
Very interested in the results you get when you come in next!
All the best.
| 