Printer unable to home Z suddenly (G30 and G29)

jojoscircus

@Phaedrux Hi, thanks for the reply, am using a magnetically attached microswitch, like the Klackender, previously was just using a z endstop, so I don't have the Z endstop defined. As for the Z maxima, will need to readjust it, thanks for spotting that. I'll record a video and post it here soon, maybe that will help with the troubleshooting.

jojoscircus

@Phaedrux Hi managed to solve the Homing issue, but the G29 is still giving me issues.

Here is a video link on what I mean by this, I have some timestamps in the description to skip to the weird probes: https://youtu.be/9va5Om2mgAY

I took your advise and realised I needed to edit my config.g to something like in the G code dictionary's M558 entry:

M574 Z0 P"nil" ; (RRF 3.0 on Duet 2 ONLY) no Z endstop switch, free up Z endstop input
M558 P5 C"zstop" H5 F120 T3000  ; Z probe connected to Duet 2 Z endstop input

droftarts

@jojoscircus Was the video done with M558 H25 set? Certainly looks like it is lifting more than H5, ie 5mm.

If it's reporting inconsistent readings, first try a probe consistency test. See this thread for example macros: https://forum.duet3d.com/topic/35200/probe-repeatability

Ian

jojoscircus

@droftarts Hi, thanks for replying, I actually used your script to test that before, here are my results i just tested. G32 bed probe heights: 0.063 0.070 0.077 0.077 0.077, mean 0.072, deviation from mean 0.005. So I don't think the bed probe repeatability is an issue. As for the h value, yup it is h25, the code I included above was from the Gcode dictionary.

droftarts

@jojoscircus From the video, it looks like all the problematic probes are on the left side of the bed. Try a probe consistency test on that side. It might be something is causing the Z to deviate on that side. Can you post the csv file that is produced, or at least the numbers?

Ian

jojoscircus

@droftarts

yup, here is the heightmap csv generated:

RepRapFirmware height map file v2 generated at 2024-05-22 18:46, min error -0.191, max error 21.656, mean 1.516, deviation 4.513
axis0,axis1,min0,max0,min1,max1,radius,spacing0,spacing1,num0,num1
X,Y,40.00,195.00,40.00,195.00,-1.00,38.75,38.75,5,5
  9.335,  0.045,  0.194,  0.407,  0.622
 -0.182, 21.656,  0.127,  0.299,  0.472
 -0.191, -0.065,  0.080,  0.236,  0.430
  2.877, -0.137,  0.008,  0.171,  0.345
  0.858, -0.160, -0.022,  0.154,  0.335

redid the deviation on the first probe coordinate, which was one of the problematic ones, here are the numbers:
Try 1: G32 bed probe heights: 23.983 11.688 3.748 -0.091 -0.109, mean 7.844, deviation from mean 9.146
Try 2: G32 bed probe heights: 18.055 13.678 -0.088 -0.091 -0.096, mean 6.291, deviation from mean 7.939
Try 3: G32 bed probe heights: 9.372 -0.088 -0.088 -0.088 -0.088, mean 1.804, deviation from mean 3.784
Try 4: G32 bed probe heights: 20.939 -0.081 -0.083 -0.088 -0.091, mean 4.119, deviation from mean 8.410

jojoscircus

After seeing the probe tended to "bounce" on the fast probe I edited my config z probe from

M558 P5 C"zstop" H25 F2000:500 T9000 A2

to

M558 P5 C"zstop" H25 F1000 T9000 A2

Seems to have solved the deviation issue:
G32 bed probe heights: -0.205 -0.204 -0.200 -0.204 -0.204, mean -0.203, deviation from mean 0.002

but when I tried the bed mesh again it generated this:

RepRapFirmware height map file v2 generated at 2024-05-22 21:34, min error -0.360, max error 9.386, mean 0.484, deviation 1.857
axis0,axis1,min0,max0,min1,max1,radius,spacing0,spacing1,num0,num1
X,Y,40.00,195.00,40.00,195.00,-1.00,38.75,38.75,5,5
 -0.205, -0.040,  0.121,  0.301,  0.486
 -0.244,  9.386,  0.070,  0.220,  0.381
 -0.254, -0.099,  0.065,  0.207,  0.371
 -0.310, -0.165, -0.004,  0.136,  0.296
 -0.360, -0.215,  1.611,  0.090,  0.255

Thanks for the help so far!

droftarts

@jojoscircus That looks a bit better, but there does seem to be two erroneous points, with Z 9.386 and 1.611. These feel false triggers. From your video, I noticed your wiring to the hot end was dragging around on the bed; check your wiring to the microswitch for lose connections, and the microswitch itself.

If you repeat the mesh bed probe, do you get these readings in the same place again?

M558 F1000 is still quite fast; I probe using

M558 ... H5 F600:120 T6000 A5

Ian

jojoscircus

@droftarts Will check on the wiring and try again, one thing i noticed is that I had better luck with a lower dive height like the default 5mm. Kinda seems like there are more false triggers with the 25mm I tried, which shouldn't be the case?

will look into the lower feed rate and report back thanks!

jojoscircus

@droftarts Hi after rewiring the probe and ensuring the cables werent dragging on the bed, I think it was a connection issue. Now seems to have no more issues with the probing. Thank you all for your help! Was pulling my hair out trying to solve it over the last few days