So my bed is not flat. No surprise. But the problem is the bed mesh, which is supposed to compensate for that is not working. I’ve been through the configurations, and they all seem right. Any suggestions?
As you can see, the front is laying down properly, but the back of the print is peeling up because it’s too high.
The “mesh” only includes a small number of points. Klipper then creates an interpolated surface to use. You can increase the number of points measured by editing the configuration.
It cannot track sudden localized high/low spots.
Is the back of the plate clean? Is the top of the magnet sheet clean and smooth? The glue residue seen in the picture would be a huge problem if you turned the plate over without cleaning it off.
Upload a picture of your mesh from Fluidd or Mainsail. Perhaps we can recommend a way to solve your issue.
Try printing in one of the front corners also.
I doubled the number of sensor points and added M117 to to the START_PRINT macro and one of them seemed to have fixed the problem. I’ll do some tests to figure out which one.
It might be interesting to see repeated bed mesh graphs with min and max values to see if the bed mesh data are consistent. The probe was loose on my Zero and that was causing inaccurate bed mesh data preventing Klipper from accurately locating the build plate, and the original inductive probe on my SV08 was inaccurate and poorly temperature compensated which prevented Klipper from compensating for the warped build plate. It would be quick and easy to run repeated bed meshes to ensure that you’re dealing with accurate and repeatable data before trying to fix any possible downstream problems.
