Manual z-probing & bed leveling


I just switched from Marlin to Repetier and gotta say, I am VERY impressed with the software!!! I think I will stick. :D

When fiddling around with my Delta printer (Folgertech Kossel 2020 RevB) I came to the conclusion that the inductive z probe is aaaawfully way of. It measures my bed with a deviation of ~ 0.3 mm (compared to manually lowering the nozzle to the bed with paper method - at the same position the sensor was measuring). I think its because of the PCB below the aluminium bed or something like that. So no z probing, auto bed leveling and bed correction for me right now. I want to switch to a retractable JCRocholl style z probe later on, but I still think the following would be helpful to a lot of people without z probes (and/or to confirm auto measurements):

Why not implement a "manual z probe" mode? I.e. the head moves to the measurement point and the user lowers the nozzle until it barely touches the bed and confirms to take the measurement now (in software, as G-Code or on the controller board). Rinse and repeat for other points. This should result in a way more precise leveling / correction compared to inductive probing.

Hence, just a proposal. I gotta admit, I am VERY new to Repetier and didn't even donate yet. I am gonna but want to try it out a little bit more before doing that.



  • With inductive you need a complete covered metal first. Wires from PCB will disturb. Even more if they are enabled and create their own magnetic field.

    You might simply add a thin metal sheet between pcb and glas to fix that and get more accurate results. (be careful not to create a short).

    Manual is not really a solution if you have so many probes. Even with automatic probe it takes quite a while especially with the new features in 0.92.8 for better average planes.
  • Well, I have an 2mm thick aluminium bed (no glas yet - will come) but still have issues - or do you mean another metal like copper or steel? Tbh and imho those probes were never meant to do high-res distance sensing. Manual switches are way more reliable.

    I do see your point against manual. Thank you never the less! :)
  • Not sure what works best. I have a inductive sensor that is definitely more precise then 0.3mm. But I guess you have sensors and sensors. Also I have only 3 sports with metal for measurements but there it works reliable and precise.

    With mechanical switches you get easily 0.02mm precision.
  • Good to know, thank you! (I guess my inductive sensor is a pretty cheap one)
  • edited January 2016
    My cheap inductive 4mm blue tip probe gives me 0.01 precision. But only if it has zero offset (mounted right under the hot end tip). All delta printers have dimensional imperfections (diagonal rods are not equal/parallel, carriages/mounts are not ideal). In this case an effector will always have slight tilt and z-probe with 10-20mm offset, like in FT Kossel will always return different reading for different coordinates even for ideally leveled bed. This is the main problem, not the probe.
    Good to know, thank you! (I guess my inductive sensor is a pretty cheap one)

  • edited January 2016
    AFAIK Repetier firmware has an adjustment for such cases, allowing to compensate the effector tilt (and z-probe error) near towers (correct me if I'm wrong, I never used this feature).
  • No, we can correct for the bumps you measure but do not correct effector tilt. So we fix symptoms not the cause.
Sign In or Register to comment.