Large CNC racking issues lately

The large CNC has seemed a bit more prone to getting out-of-square lately.

I talked with @jamesfreeman about this a bit last night and he reminded me of some easy techniques to correct and mitigate the issues, but I think there are 2 things worth mentioning here:

  1. The last two times I’ve come in, the machine was more significantly out-of-square than I had previously experienced. As in, after initial homing, the (X=0,Y=0) corner was visibly offset in the X+ direction relative to the (X=0,Y=60") corner. I didn’t precisely measure, but it looked like between 0.25 and 0.5" off. It was reasonably easy to make gross corrections to that via my usual workflow of resetting the drives and manually jogging X- to push the gantry against the stops. After that, it was better, but…

  2. Last night, I needed to 100% maximize my sheet of plywood, so I was very sensitive to squaring. Even after @jamesfreeman helped align my sheet, the end result was still off by over 0.125" (possibly up to 0.25" across the Y axis). Luckily, that was close enough for this project, but it would definitely affect large-ish joinery projects and such.

So what’s the current expectation? Are others experiencing similar issues?

I’m particularly curious as to why the machine seems to get “extra off” over time. FWIW, my two sessions were about a week apart. I understand that @jamesfreeman and maybe others did some maintenance work on the machine around the same time (thanks!), but not sure if that could be the cause or something else is going on that’s worth looking into.

1 Like

Racking is uncontrolled while the machine is off, so this is expected and the manual un-racking against the stops should have squared it up.

The frame was recently loosened and retightened. It could have been shifted in that process. The stops were calibrated and don’t drift, so if there’s a problem that’s why. I’ll look and see if it needs a stop moved.

1 Like

Yeah, I did part of the work on it, and we haven’t re-calibrated the end stops after the work, so I can believe the endstops that help you reset the rack could be out of calibration. I’m planning on testing it in my free time.

In the past, the rack across the table was below 1/64th of an inch, and it held that calibration over an extended period. If you need the rack really accurate, run it against the end stops, re-home it, then move the X-axis all the way to the soft limits, and compare the gap between the endstops and the gantry. In the past I’ve found that I should be able to fit my fingernail in between both sides. If the gap is not the same between the two, I needed to unhome the machine, and reset the rack. This lets you get it perfectly parallel to the endstops.


Racking is uncontrolled while the machine is off, so this is expected.

Ah, that makes sense, and in retrospect, that’s another “well, duh” moment for me… I guess I had just been relatively lucky in the recent past where it had mostly not been nudged out of alignment while unpowered, so I developed unrealistic expectations that it would maintain squaring.

I’m totally fine with the need to manually align it, I just wasn’t in the habit of seeing it that far out.

1 Like

I’m getting the new version of Linux CNC up on the NUC and this should be automatic