How fine tuning a delta printer?

  • Hi,

    how i am close to finish Tevo Little Monster conversion to Duet board… i am thinking in calibrate it.

    Where i can find information of the process of calibration?

    First the calibration of estrusion steps? The geometry?

    the advice and resources are appreciated

    The idea is start for here...

  • Using Duet electronics and an accurate z probe, its quite a simple process, my summary would be:

    Measure your diagonal rods between centre of rotation and centre of rotation, enter that as

    M665 Lyyy.yy where rod length is yyy.yy in mm

    Home and then manually descend until your nozzle just touches the bed (at bed centre), this is your starting height for your M665 command.

    add the M665 command as below in config.g. then reboot.
    M665 Lyyy.yy Hxxx.xx where xxx.xx is height in mm.

    Then define a bed file using the wizard
    Set 6 factors or 8 factors to calibrate (which leaves the diagonal rod length set manually).
    Save it and upload it to the system folder on the sd card in the duet via DWC (or by putting the sd card in a pc).

    Run G32, make sure you are homed first (I put a G28 at the start of my bed.g). Its often worth sending M906 with low motor current values in case anything goes awry and the print head goes on an excursion or crashes the bed.

    This will give you (presuming your mechanics are good, and your probe is accurate) a calibration result, the lower the deviation the better. The new values for M665 and M666 are now in memory and are being used, but are not saved until you send M500, then they are written to config_override.g (so in future remember if you want to manually change values you have to do it in config_override.g for it to be used by the machine on boot).*

    You can now run it again (without homing) to refine your result. After 2-4 runs the result should be the same each time if it varies wildly then something is amiss mechanically with your printer or electrically with your probe.

    If using a probe which is offset from the nozzle you should check your probe's trigger height relative to the nozzle to bed gap, using the paper method at bed centre, and then compare that to the height at your other chosen probing points from bed.g. This tells you if you have a tilting effector. You can enter these height corrections on each line of bed.g using H parameter. If you are using a nozzle contact probe just move to the chosen points at a fixed height (say 5mm) and send G30 S-1 this doesn't change anything, it reports the distance travelled before probe triggered, it should be the same at each point, if your calibration is good.

    Of course there's a lot more to it than this but that's how to get started.
    *some people put m500 at the end of bed.g this can be good as it continually saves your calibration, which I run before each print. However it can also cause problems, if something goes awry during calibration and you do not notice it, the "bad calibration" gets auto saved and then the machine can behave quite strangely. I've suggested as a feature upgrade a pause if calibration deviation exceeds a pre-set threshold, so if you decide that 0.01 deviation is your line in the sand, the machine stops and reports an error if calibration does not meet that standard.

  • Great write-up BTW!

  • Thanks

Log in to reply