X-Axis Stretched

My Setup

  • Millmage 0.9.00 RC (also 0.8.2 as a backup, not at the same time)
  • X-Carve 12" x 12" (inventables)
  • GRBL
  • Squared, belt tensioned, and other hardware checked.
  • for testing V-bit 45deg cutter.

The Issue

When i do a test cut of squares within squares and circles within circles, the shapes are consistent, closed, and repeating. but for some reason a square that was programmed at 1" x 1" comes out as Y=1.000" and X=1.012"(ish =/- like .005). but it’s consistently bigger in X than programmed. same for circles. but if i jog it 5" in X and 5" in Y they both measure 5.000" with calipers. meaning that something in the between clicking the Process button and the cutting of the object, the X axis is adding distance.

Things i considered

  • belt is stretched / over tensioned
    • but that would mean that jogging would also yield incorrect distances
  • wood grain resistance
    • would be ridiculous at the tool pressures i’m testing at, but tested anyway. not a factor.
  • Machine position Lag.
    • My G-code send stream is consistently multiple lines ahead of my machine position in the Control Mode visual. not sure how much or how this would affect it. maybe accelerations need to be tweaked? again, not sure how acceleration compensation works.
  • Steps / mm firmware adjustment
    • if that were a solve, then jogging would also yield incorrect x-axis results.

UPDATE:

I’ve run a program Profile (spiral down for repetition) cutting a 5.000" square.
results are X=5.012 and Y=4.988
so maybe my steps/mm is off, AND my Y-axis is losing steps?

Summary:

  • @ 1"x 1" => X=1.012 Y=1.000
  • @ 5" x 5" => X=5.012 Y=4.988

solved.

Y2 grub screw was incorrectly positioned. the grub screw was tight, but it wasn’t on the flat of the Y2 motor shaft. so there was some microslipage that i couldn’t detect because it was juuuust tight enough that it didn’t wiggle. i re-adjusted it and things are correct. dimensions cut as programmed.

what’s got more egg on my face is that i checked that the grub screw was tight, too. just didn’t check if it was on the flat. :rofl: