Maslow 4 on v1.26.0: Find Anchors fit fails, and bottom-left position error E-stops cuts at about 80%

Hi all, I’ve been fighting this for a few days and could use some fresh eyes.

Setup: Maslow 4, now on v1.26.0 (index.html v1.26.0). Before that it was v1.25.0-24, with the same behavior. The frame is roughly 2.5 m wide.

Problem 1, calibration: The early passes fit well:

  • 6 points: rms 1.25 mm
  • 15 points: rms 1.93 mm
  • 31 points: rms 2.38 mm, max 5.75 mm

Then the next pass fails with: Find Anchors fit failed: maxResidual=22.582mm at anchor=0 measurement=54 (limit 15.000mm). Earlier runs failed with rms=5.157mm and rms=6.246mm exceeds limit 5.000mm. The anchor estimates also drift a lot between passes, for example top right going from about 1965 to about 2459 mm in X.

Problem 2, cutting: Jobs E-stop at the same place, line 8 at 80.5% complete. First Top Left reads about -39 mm and Bottom Left about +24 mm. After that, Bottom Left keeps running 15 to 22 mm behind until it trips Position error > 15mm while running. E-Stop triggered. One run also logged Motor current on Bottom Left axis exceeded threshold of 4000. During Apply Tension I also get “Belt retracted 0.000mm, may not be anchored” warnings on Bottom Right and Top Right, and the next boot says “Previous boot ended in panic.”

Things I’m checking: the bottom-left belt path for rubbing or snags, slack belt dragging, anchor flex, the sled dragging on the sheet, and the arm spinning freely.

Has anyone seen the residual climb like this as calibration expands, or Bottom Left consistently lagging? Is it more likely mechanical, or is there a setting I should look at, like the retraction limit, extend distance, or calibration grid size? Logs are attached. Thanks!