Denvi / Candle

GRBL controller application with G-Code visualizer written in Qt.
GNU General Public License v3.0
1.37k stars 547 forks source link

Heightmap process seems to gradually lower Z flying height during processing until bit touches PCB during XY-move #626

Open poes8943 opened 7 months ago

poes8943 commented 7 months ago

Versions

PC info

Describe the bug Heightmap process seems to gradually lower Z flying height during processing until bit touches PCB during XY-move

To Reproduce using standard probe command: G21G91G38.2Z-30F100; G0Z1; G38.2Z-2F10 doing a manual Z-probe first using Z-probe control button followed by a Z-move UP 0.5mm starting a 3x3 heightmap (20x25mm PCB area) Zt:2.00 Zb:-1.00 automatic results using heightmap feature: 1.964, nan, nan 1.976, 1.966, 1.559 0.400, 0.784, 1.179 flying height of the probe gets lower after every probe until the probe touches the PCB heightmap stops after 7 probes with ERROR:9 manual results: (doing the first 3 probes manually with Z-probe G command 'G38.2Z-2F10' and jog Z-up 0.5mm after probe) Grbl 1.1f ['$' for help] G38.2Z-2F10 < [PRB:0.000,0.000,0.009:1] G38.2Z-2F10 < [PRB:10.000,0.000,-0.004:1] G38.2Z-2F10 < [PRB:20.000,0.000,-0.018:1] The manual process gives reasonable values ... so stepper motors and spindles seem to be fine and I am not loosing any step counts. ... no ideas why this happens ... any clues or known issues?

Expected behavior There should be no difference between 'manual' and automatic heightmap.

poes8943 commented 7 months ago

further test steps using Z-probe control button: 1) push Z-probe control button and perform a Z-zero after that (work coordinates now at 0.000 0.000 0.000) 2) jog Z-up 0.5mm and pushing Z-probe control button again (Z axis work coordinate at 0.399 now) 3) jog Z-up 0.5mm and pushing Z-probe control button again (Z axis work coordinate at 0.807 now) 4) jog Z-up 0.5mm and pushing Z-probe control button again (Z axis work coordinate at 1.199 now) ... there seems to be roughly an error of 0.4mm after each step and values adding up ...

same test with manual commands: 1) push Z-probe control button and perform a Z-zero after that (work coordinates now at 0.000 0.000 0.000) 2) jog Z-up 0.5mm and executing 'G38.2Z-2F10' (Z axis work coordinate at 0.009 now) 3) jog Z-up 0.5mm and executing 'G38.2Z-2F10' (Z axis work coordinate at 0.006 now) 4) jog Z-up 0.5mm and executing 'G38.2Z-2F10' (Z axis work coordinate at 0.007 now)

mar0x commented 7 months ago

Hello, Please follow the long discussion in #550 maybe you find something useful.

poes8943 commented 7 months ago

thanks - was reading #550 some hours ago but seems not to solve my issue (revisiting the thread again)

Hostaler commented 7 months ago

How can I put all this to work? Thank you very much.

El El dom, 11 feb 2024 a las 13:38, poes8943 @.***> escribió:

further test steps using Z-probe control button:

  1. push Z-probe control button and perform a Z-zero after that (work coordinates now at 0.000 0.000 0.000)
  2. jog Z-up 0.5mm and pushing Z-probe control button again (Z axis work coordinate at 0.399 now)
  3. jog Z-up 0.5mm and pushing Z-probe control button again (Z axis work coordinate at 0.807 now)
  4. jog Z-up 0.5mm and pushing Z-probe control button again (Z axis work coordinate at 1.199 now) ... there seems to be roughly an error of 0.4mm after each step which is adding up ...

same test with manual commands:

  1. push Z-probe control button and perform a Z-zero after that (work coordinates now at 0.000 0.000 0.000)
  2. jog Z-up 0.5mm and executing 'G38.2Z-2F10' (Z axis work coordinate at 0.009 now)
  3. jog Z-up 0.5mm and executing 'G38.2Z-2F10' (Z axis work coordinate at 0.006 now)
  4. jog Z-up 0.5mm and executing 'G38.2Z-2F10' (Z axis work coordinate at 0.007 now)

— Reply to this email directly, view it on GitHub https://github.com/Denvi/Candle/issues/626#issuecomment-1937617909, or unsubscribe https://github.com/notifications/unsubscribe-auth/A6RZWF2OC6Q5B2QCKY63BNDYTCUTHAVCNFSM6AAAAABDDOIYIWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMZXGYYTOOJQHE . You are receiving this because you are subscribed to this thread.Message ID: @.***>