monitoring mesh bed compensation level



  • Is there a method to show, during printing, what mesh bed level compensation is applied ?
    I have a large print and my print bed curves up by about 0.4 mm on the right side based on the mesh map but it appears that although mesh compensation is active and there is some Z movement happening while the printhead is traveling, insufficient compensation is applied.If I could see the amount of compensation applied then it would probably help figuring out what is happening.
    The Z height on the web display of course is constant so that doesn't help.



  • Have you looked at the terrain/heat map uploaded after each mesh run to the web control?

    This should give you a visual +-0.25mm



  • Yes, I can easily display the map but what I am looking for is a way to see the instantaneous correction applied as the printer prints. I can see how the build plate deviates from flat but would like to see the amount of correction applied as it prints because the first layer is not satisfactory. The far right side of the print shows that the gap between nozzle and print base ends up at almost zero with the printed trace being see-thru and the extruder skipping because there is no room for the PLA to extrude. This is with a print thickness on the first layer of 0.4 mm which leads me to believe that insufficient correction is applied from the mesh map.
    Being able to see the instantaneous correction applied would assist me in determining what is going on.



  • If I understand correctly - the first layer is too close to the bed on one side and at the correct height on the other?
    Which is giving the indication the mesh compensation is not matching the real position of the bed?

    If this is the case, a few quick sanity checks spring to mind.

    1. Is the Z-Axis lead screw physically moving when the X position is changing? (assuming this is a cartesian setup)

    2. Do you have a PanelDue? What are the co-ordinates of the Z relative to the X/Y as the print is laying down?

    3. What Z-Probe or sensor are you using?



  • Yes, you understand the issue correctly although I can't say for sure that the correct height is printed on one end. All I can see is that the other side is way too thin.

    1. Cartsian setup and I believe the z axis is moving. Difficult to tell even with touching the z axis to try and feel the movement.
    2. No PanelDuo. The web display shows the same height (0.4 mm) for the full first layer but the area that is printed 'thin' s curled up almost that amount so it appears that almost no correction is applied.
      3)Z probe is a true BLTouch

    I have an uneasy feeling that maybe the nozzle is running into an area that was not probed due to probe offset but am not sure. Being able to monitor the instantaneous correction applied to the Z height would immediately tell me if I am running into an area that was not probed.



  • is your probe offset correct? its easy to get +/- confuesed. this will result in wrong offsets



  • Thinking about it some more, I am pretty sure the printing is happening in a probed area. The print currently running has another 5 hours or so to go at which point I can measure thicknesses and try to measure the gap between nozzle and print bed at various positions. I was hoping that a display of 'actual' Z height rather than 'corrected' Z height would easy the trouble shooting if it was available somehow.



  • To my understanding (and confirmed with trial and error) the polarity is correct. The probe is to the front of the nozzle (negative offset) and to the left (negative offset)
    From config.g: G31 P25 X-13.5 Y-34.81 Z2.65



  • Related question: Suppose my z speed/acceleration/jerk value are too low. Would the Duet slow down x and y so as to not exceed the speed limits of Z or would I see what I am seeing now ? I am assuming x and y would slow as needed but thought I better verify that ....



  • @jens55 Yes. Any print move will be capped to the slowest speed of either XYZ or "E" if any of those speed limits are reached. So a very low maximmum extruder speed setting would also affect the overall speed of any move.



  • @jens55 said in monitoring mesh bed compensation level:

    Related question: Suppose my z speed/acceleration/jerk value are too low. Would the Duet slow down x and y so as to not exceed the speed limits of Z or would I see what I am seeing now ? I am assuming x and y would slow as needed but thought I better verify that ....

    If your Z axis jerk were too low, you'd see stuttering in X Y movement as it tries to apply the mesh compensation.



  • I have confirmed that the model that gave me issues was only printing in the probed area.
    I have also confirmed that the mesh is applied to compensate nozzle height although I do not know if the compensation is sufficient.
    Measuring the gap between nozzle and glass plate is anything but consistent over the build area.

    The options I am left with - insufficient compensation for some reason or other or, more likely, incorrect mesh height sensing.

    I will scan the bed multiple times at the same temperature and compare measurement points to see if anything jumps out.



  • What settings are you using for the BLTouch? There are some tweaks you can implement to make it more reliable and repeatable.



  • Um ... settings ? I am not aware of any settings for the BLTouch other than trigger height which is set at 25



  • More questions than answers .... attached are three bed maps run consecutively with nothing changed. Each run is 100 points I think. A quick perusal shows that the height can be different on different runs by more than 0.1 mm which is not acceptable to me. I do not know if this is a limitation of a BLtouch (comments ?).
    Note that I did not reset z zero between runs!

    If I measure the same spot ten times in a row (this is without moving x or y, I get the following:
    -0.034 -0.034 -0.037 -0.034 -0.029 -0.034 -0.037 -0.037 -0.034 -0.034, mean -0.034, deviation from mean 0.002 ... this is perfectly acceptable
    The second run of the same 10 points shows as:
    -0.054 -0.051 -0.066 -0.066 -0.059 -0.061 -0.059 -0.054 -0.054 -0.061, mean -0.059, deviation from mean 0.005
    The third run shows as :
    -0.056 -0.054 -0.056 -0.059 -0.054 -0.056 -0.054 -0.049 -0.051 -0.054, mean -0.054, deviation from mean 0.003
    In each case the 10 measurements are taken successively at the same spot without changing x or y but there was a 'home' command between each of the 3 runs of 10 where Z was reset and x/y axis moved.

    At this point I do not know what to think. Am I expecting too much ? Is the accuracy of the BLTouch acceptable? Why is there so much of a difference in the three bed scanning runs ?



  • What do you have in your M558 command?

    https://duet3d.dozuki.com/Wiki/Gcode#Section_M558_Set_Z_probe_type

    Here is my BLTouch config for example.

    M558 P9 H3 F60 T6000 A10 R0.75 S0.003 B1 ; P9 for BLTouch, dive height 3mm, probe at 100mm/min, travel 6000mm/s, up to 10 probes, pause 0.5s, heaters off

    The BLTouch can benefit from a slow dive speed as well as a brief pause between moves to allow the probe to settle. I've also tightened the tolerance for acceptable deviations between probes. So two consecutive probe results must be within 0.003 of each other, and if not, it will reprobe up to 10 times. It also disables the heaters during probe moves to remove any potential interference.



  • M558 P5 H5 F500 T4000 X0 Y0 Z1 which is taken from the Duet setup instructions
    Note that if I measure the same point multiple times the readings are ok. It's only when I move x and y between readings do they start to go out of whack



  • I suggest you try my config line then. Yours is rather out of date. P5 is for generic switch type probes. P9 is specific to the BLTouch. The X Y Z parameters are also deprecated. And your dive speed of 500 is too fast for the BLTouch.



  • Thanks, I will re-run everything and see what happens.



  • Slight change, I need to run with the bed heater on since the bed changes substantially between cold and operating temp.
    I will report back.


Log in to reply