I was having a problem with Z axis homing and bed plane compensation. I set up the G32 command and automatic bed plane compensation coordinates as instructed. Before printing I home all axes and then run the bed plane compensation. But when I came to print, there were issues with distance between the nozzle and the bed. On one occasion, the nozzle was touching the bed and at least the first 2 layers did not print. On another occasion, the nozzle was too far above the bed when starting to print, so the extruded filament would not adhere.
While investigating this, I noticed some slight scuff marks on the aluminium tape in the corners of the bed, and I wondered whether they were causing the reading to vary. So I set the height to 1mm and took a series of G32 readings, moving the Y axis in 1mm increments between readings. The readings I got varied from 265 to 438 over 7mm of Y-travel. This confirmed my suspicion that the scuff marks were causing a problem.
I then repeated the test but using white PVC electrical tape in place of the aluminium tape. This time the readings varied between 323 and 332 over 7mm of Y-travel. So replacing the slightly-scuffed aluminium tape with white PVC tape reduced the variability in readings from 173 to 9.
My readings were taken without any Kapton over the aluminium tape or PVC tape. The readings are lower than typical because I adjusted the component values on my proximity sensor board, to stop it saturating and therefore give more accurate readings at small distances.
In summary, I find that white PVC tape gives much more consistent readings than aluminium tape, and the problem I had with incorrect Z-axis height at the start of the print has gone away.
Edited 3 time(s). Last edit at 12/24/2013 02:50PM by dc42.
