Running Pressure Advance on Perimeters Only



  • I am experimenting with trying to only enable PA during perimeters, and disabling for infill, through some custom gcode insertion in SuperSlicer.

    But, running through a range of PA values from 0.0 to 1.5, I saw a uniform (and big) seam bulge (PA is clearly working during the print). So I started troubleshooting:

    • I tried inserting an M400 before the M572 commands to prevent read-ahead from executing the M572 D0 S0 before the perimeter is complete. No change.

    • Then, I manually removed the PA on command right before the external perimeter, since it was duplicating the command given for the internal perimeters.

    This brought the seam bulge down significantly, but it still appeared uniform across a wide range of S values. So, it seems like calling a redundant M572 command (positive S value), between internal/external perimeters, is causing PA to leave an error in the extruder position, causing residual extruder pressure.

    Through some additional logic, I was able to eliminate the redundant M572 command outputs from SuperSlicer.

    @dc42 It seems like having a check in PA that if an M572 command asserts the current value, no action would be taken.

    When setting S=0, PA should "settle up" and return the extruder position to its unmodified position. I'm not sure if that is currently the case.



  • Have you considered hooking a Saleae or similar logic analyzer to the Duet's step/direction signals?

    The recorded data should have very detailed information what is sent to the steppers.



  • @zapta said in Running Pressure Advance on Perimeters Only:

    Saleae

    No doubt. But, there is no way I could justify $400 for diagnostic stepper data.



  • @CCS86 said in Running Pressure Advance on Perimeters Only:

    @zapta said in Running Pressure Advance on Perimeters Only:

    Saleae

    No doubt. But, there is no way I could justify $400 for diagnostic stepper data.

    The £10 from ebay ones plus Sigrok / Pulseview is a good combination, in my experience.


Log in to reply