X axis loosing accuracy over long distance

New to OneFinity, new to Fusion 360 (familiar with other CAD), new to CNC in general.

Running the Woodworker 50

I’ve been slowly getting things working and learning this machine over the last month or so. I’ve begun to notice some inaccuracy in the x-axis when tool paths span much larger than 12 in. At first, it was small test projects, but I’ve just finished cutting 9 pockets for T-Tracks, 3" apart, and I’m having issues.

First off, my CAD was set to mill the pockets .77" wide to accept the .75" T-Track rails. However, they wouldn’t fit into the slots. I’ve had to expand the CAD sketch up to .80" before they would fit. Perhaps the plywood I’m using is swelling, not sure. I was ok with that subtle amount of inaccuracy.

However, now that I can fit the first T-Track, I started putting in the next few. The 3rd was very snug, and probably won’t sing completely to the bottom, but the 4th and remaining pockets are not wide enough to allow a T-Track.

I re-zero’d the X-axis 9" to the right and re-ran the tool path for a few slots, now they are wide enough to accept the T-Tracks.

Of course, I can simply widen the sketch and re-run the operation, however, I feel like I need to adjust something in the controllers to prevent the X-axis from losing accuracy.

I did see a posting about checking the Travel per Rev and Max soft limits, those seem to match the recommendations for my machine.

Check the motor to ball screw connection?
Tightening the coupler on the ball screw // Updated for quick access holes // Support - YouTube

Hey Craig,

you say you make 9 pockets, 3" apart, and that after cutting, the width of the pockets is too small. Is the overall distance, between the first and the last pocket, diminished too?

2 Likes
  • I’ve not yet checked the ball screw coupler.

  • I did check the pocket to pocket distance before starting this thread, and they seemed pretty consistently right on 3”. Unfortunately I don’t have my calipers with me to be sure. I’ll put it on my to do list.

It’s difficult without large scale precision measurement capabilities, but I think first you need to figure out if it’s a linear cumulative error (bad steps per unit setting), or a backlash issue possibly caused by something as simple as the x-axis ballscrew nut not tightened quite enough, causing the backlash.

1 Like

If time and material permit, perhaps running the same (partial?) path perpendicular to the first on a scrap board and comparing the results.

I meant overall distance, is the (9 × 3 =) 27 inches distance also diminished?

So I finally was able to run some more long-distance testing. I ran tick marks every .5" with engraving paths down a 30" wide board, and then up a 30" long board. They all look fine.

I have also checked the set screws on the coupler for the ball screw shaft. They are very tight.

Unfortunately, I’m not being very consistent in my designs, from a testing point of view. I’m bouncing all over the place trying to solve/learn other basics. perhaps this issue with fix itself as I progress.

Currently, I’m having an issue where I’m asking it to cut a .75" diameter hole, and it’s coming out about .73" (ranging between .725 and .74). And, trying to get it to cut through a known stock depth is frustratingly difficult. Sometimes it doesn’t make it through, and I’ve got to re-run with a lower Z-axis zero point. Other times it’s so far down that my tabs are non-existent. I’ve purchased, and been using the probe for about 2 weeks now, but still have inconsistent depths. Even tried adjusting the probe thickness in the settings. I’m hoping it’s all just learning pains.

Did you ever figure this out?

I just built my wasteboard and it has Dog Holes in it, w/ a diameter fo 19.05 in my design, but as the machine moves further to the right, the holes seem to be tighter and tighter and for at least the right half of the machine, most of the holes are too small.

I don’t even know what things to consider… any insights here?

If you have a dial indicator you can use it to test the actual movement of the machine vs the commanded movement across the bed of the machine to verify if there is an issue. Set it up parallel to the direction of travel of the axis, zero it out and then move the axis .5 or .75 and validate the result.

FWIW holes will often cut smaller than you specify because of the nature of wood being fibrous and the fact that the machine will turn circles into a series of lines to cut, however this should be consistent regardless of the location of the cut.

Googling dial indicator…

What are the sort of things that might be wrong that I’d learn from a dial indicator?

I’ve created a separate post for my specific issue here btw.

If you command the machine to move .750 and it moves less (eg .744) then you may have a loose coupler screw, calibration issue, losing steps due to a lose connector etc. If you command it to move .750 and it moves more (eg .756) then you have a calibration issue.

There are a few threads on here about calibration, this is one:

1 Like