Struggling with bed level, delta with duet and smart effector



  • Hi

    I've got a delta I built based on DC42's blog and in general it is fine, but Im having a nightmare getting the first layers down, the right and back of the bed have a high point no matter what i do in the config/mesh etc

    It seems like after the probing and mesh creation that its just not getting used, so I wonder if someone could look over my setup and tell me where im being stupid please

    the ally bed plate on the printer is toolplate, so is flat, I'm still deciding if to go with magnetic sheet and perspex or glass and printbite or similar, ideally would love a removeable flex plate, but thats for another post as dont want to confuse this one too much

    config files are below /attached

    3_1566722649408_homedelta.g 2_1566722649408_config-override.g 1_1566722649407_config.g 0_1566722649406_bed.g

    system version is
    firmware wifi : duet 1.02 or later
    version : 2.04rc1 (2019-07-14b1)
    wifi version : 1.21
    web interface : 1.22.3

    bedmesh console info and view
    1_1566722897133_console after bed mesh ran.JPG 0_1566722897133_bedmesh view.JPG

    example gcode (severely truncated)
    0_1566722948499_example gcode produced from slicer .txt

    the basis of my leveling is .....
    run delta calibration min of 3 times till it stabilises on results

    run m666 and m665, take these values and put in config.g as a baseline setting

    then run the bed mesh calibration

    then set nozzle to bed with a4 paper to give estimated 1st layer height
    g92 z0 = set nozzle to zero
    g1 z5
    g30 s-1 = sensor trigger (do it multiple times - check consistent results

    this gives z value for nozzle offset / first layer height

    in config.g file
    g31 z compensation ... put z value in

    after all of this, it starts printing, but is still high on the right size and a point on the back, which is what i thought the bed mesh would resolve

    hope its something stupid

    Thanks for reading
    Phil



  • forgot to add, the current bed is sanded perspex held down with bulldog clips, not ideal long term but its a starting point, all calibration is done with a cold nozzle else it would melt the perspex on probe
    I do check there is no filament on end of nozzle before running the calibration etc

    Thanks
    Phil



  • Can anyone please offer any help / direction i should take please ?



  • What are your results on doing the autocalibration?



  • @phillawrence said in Struggling with bed level, delta with duet and smart effector:

    Hi

    the basis of my leveling is .....
    run delta calibration min of 3 times till it stabilises on results

    run m666 and m665, take these values and put in config.g as a baseline setting

    Nope. Enter M500 from the console instead. This stores the values in config_override.g
    (You already, correctly, have M501 at end of config.g to fetch these at boot. )

    then run the bed mesh calibration

    Good.

    then set nozzle to bed with a4 paper to give estimated 1st layer height
    g92 z0 = set nozzle to zero
    g1 z5
    g30 s-1 = sensor trigger (do it multiple times - check consistent results

    this gives z value for nozzle offset / first layer height

    in config.g file
    g31 z compensation ... put z value in

    Nope. Don't do ANY of that.
    Instead, be sure there is a G29 S1 in your slicer's start g-codes, to invoke the saved mesh map.

    I also run a G32 in my slicer's start gcode, to re-probe (calibrate) the bed before every print. My bed.g does a 16 point, 8 factor, for these pre-print calibrations. Fairly fast, and works great. This is entirely optional.



  • @Danal thanks, it looks like i was missing the g29 s1 in the slicer code, so it was ignoring all the good work before hand lol



  • @oliof thanks, autocalibration was bringing good results back, it looks like the slicer was ignorning the bed mesh, as I had not added the g29 s1 (live and learn)

    Cheers
    Phil


Log in to reply