moiré on my delta 3D printer

  • Hi community,
    we have issue with moiré effect on flat surfaces. We are sure there isn't any mechanical issue, because we have lot of printers with the same construction running on MKS board with TMC2310 drivers and the are printing almost without moiré effect. I did some experiments with microstepping setings, but smaller microsteps don't have any benefits compared with 16 microsteps with interplation. I think there is something wrong with TMC2660 setings. I tried change HEND parameter, but this change doesn't have any effect. Do you have any idea how to get nice flat surface without moiré? Thanks for your answers.
    ![0_1564041877478_20190725_100139.jpg](Uploading 100%)

  • what printer is it? a delta, a cartesian?

  • It's delta printer

  • what are your stepper motors specifications? and what current are you running them at?

  • And what supply voltage are you using?

    There seems to a common theme of ripples on this forum. I think whatever creates vertical ripples on a Cartesian machine probably makes a moire pattern on a delta.

  • There are motor specifications. I tried currents 800 and 1000mA.

  • We are using 24V power suply.

  • you should be running the steppers between 75% and 85% of the max current.
    so try again with 1600mA

  • You could try stepping very slowly see if you can hear missing microsteps as I did here on an older generation of stepper drivers.

  • administrators

    What layer height are you using? If the microsteps on your motors are not even, that causes moire on flat vertical surfaces unless the layer height is an exact number of full steps. But the surfaces in your photo are not vertical, so this may be less relevant.

  • I think I know what you mean, microteps near hardware fullsteps can't be set correctly (like on my painting).0_1564049974824_Bez názvu.png I did some measurement of coil voltage with osciloscope, but i saw only short spikes, and shorter or longer off-time, not sine wave.

  • @thugmek You need to look at current, not voltage. That is harder to do, I use hall effect current modules that are completely isolated and have no resistance, so can be inserted in series with the coils without having any effect.

    I have also used an isolated scope and a small series resistor but that risks damaging the drivers.

  • Altenratively you can measure the voltage across the sense resistors with a scope but you only see part of the waveform when the current is not being recirculated. See

  • I don't think there is some risk for driver if I use 2 small resistors, but circuit will have different behavior if I increase resistance. I tried measure voltage on sense resistor and I saw square pulse with, off-time and square pulse with negative polarity.

  • @thugmek Even with an isolated scope you can only measure one coil because the probes are normally linked. However you only really need to look at one coil as the the other should be the same with a 90 degree phase shift.

    When looking at the sense resistors there will be periods when it is zero but the on portions should be an accurate reflection of the current.

    The actual current should be a sine wave with a saw tooth ripple. When you look across the sense resistor you see just the rising parts of the saw tooth with gaps in between.

  • I set current 1600mA and it's quite good. Moiré is still visible, but 1000mA benchy is 4x-5x worse. Only problem is motor temperature. We use PLA parts and 60 degrees on motor is too much. TMC2660 should have some sine wave correction. HEND register should be used for this, but I don't have any idea how can I calculate correct value.

  • I'm refering to oficial TMC2660 datasheet, pages 20 and 35-36. But I'm afraid that correction is posible ony in Constant Off-Time Mode 😕

  • @thugmek If it gets better with higher current then perhaps it can't do the low current microsteps, in which case a lower supply voltage might help.

    I only have experience of Allegro constant off time drivers but they definitely don't work properly on 24V because the minimum on time is too long to meet the very small currents needed near the zero crossing.

  • you might be able to isolate the stepper from the pla by using a vibration damper made from cork. this will also reduce the noise of the printer.

Log in to reply