No trigger on probe after full movement, but only rarely

Any idea what would cause an infrequent, but persistent:
No trigger on probe after full movement
No trigger on probe after full movement
See the attached video, this is the last minute of a repeated probe accuracy before it fails Before this run I'd done a couple 100 sample runs without issue.
41 Replies
JaminCollins
JaminCollinsOP4mo ago
Here's the console output for the above run.
JaminCollins
JaminCollinsOP4mo ago
here's another attempt, by my count it completed 253 repetitions before failing, in the same way
JaminCollins
JaminCollinsOP4mo ago
and the klippy.og including the most recent attempt
JaminCollins
JaminCollinsOP4mo ago
I see nothing useful in the log, it goes from a successful probe to a failure
cgrr
cgrr4mo ago
In my case that mostly shows up when the nozzle is not clean enough, even a little crud will trigger it. One person also mentioned he got it due to lose wiring (I always get a different warning for that one). Nowadays I try to clean the nozzle as soon as a print is done before it cools down.
JaminCollins
JaminCollinsOP4mo ago
This has nothing to do with the nozzle, it's purely beacon coil probing. If you watch the video you'll see that for several repetitions the bed moves down and fully back up. However for the final failing repetition, the bed moves down but does not move back up the same amount. The bed just stops short for some reason, but claims it fully moved.
samus129
samus1294mo ago
Sorry I don't see a video here but I will say mine did that when I had my beacon upside down by accident :pepehands:
JaminCollins
JaminCollinsOP4mo ago
it's an attachment on the initial post
samus129
samus1294mo ago
Oh the one it wants downloaded 🤦‍♂️
JaminCollins
JaminCollinsOP4mo ago
I'll try converting it to MP4
samus129
samus1294mo ago
Well I would have thought the log would say what RatOS update your on but it doesn't so going to ask if your on the 2.1 RC
JaminCollins
JaminCollinsOP4mo ago
Yes, 2.1
samus129
samus1294mo ago
All good I got it play once downloaded I didn't know it was the video and ya it's triggering as it lighting up red
JaminCollins
JaminCollinsOP4mo ago
Here it is as MP4, if you watch the right hand extrusion you can see that the bed fully eclipses the M6 nuts on a full move. On the final move it stops quite short.
JaminCollins
JaminCollinsOP4mo ago
A normal full move ending:
No description
JaminCollins
JaminCollinsOP4mo ago
The final move where it stops short:
No description
JaminCollins
JaminCollinsOP4mo ago
and again, it works fine for 250-ish repetitions and then just stops short
samus129
samus1294mo ago
Ya mine did that with the calibration even when it passed and every time I home the printer it will stop were it's not triggering - like after it triggers it lowers a little again Does your 3.1 fully power off after your done with it? I ask as VCore 4 doesn't have a power switch anymore after the IEC patch and my minion would get all wonky if I didn't power cycle it if I have done a lot of prints
JaminCollins
JaminCollinsOP4mo ago
Mine stays on most of the time, but has been recently power cycled... as I recently upgraded to 2.1
$ uptime
04:52:11 up 4 days, 8:30, 2 users, load average: 0.55, 0.48, 0.57
$ uptime
04:52:11 up 4 days, 8:30, 2 users, load average: 0.55, 0.48, 0.57
can power cycle it and repro but that'll be tomorrow
JaminCollins
JaminCollinsOP4mo ago
just replicated after a fresh boot:
$ uptime
17:17:48 up 31 min, 2 users, load average: 2.35, 1.56, 1.21
$ uptime
17:17:48 up 31 min, 2 users, load average: 2.35, 1.56, 1.21
Console output from the session attached, 236 iterations by my count (grep -c). For anyone that would like to try this for themselves, this is the g-code I'm running:
[gcode_macro z_stress]
gcode:
G28
Z_TILT_ADJUST
G1 X200 Y200 F4000
PROBE_ACCURACY SAMPLES=1000 SAMPLE_RETRACT_DIST=50
[gcode_macro z_stress]
gcode:
G28
Z_TILT_ADJUST
G1 X200 Y200 F4000
PROBE_ACCURACY SAMPLES=1000 SAMPLE_RETRACT_DIST=50
JaminCollins
JaminCollinsOP4mo ago
Would also love to know if someone can get the 1k iterations to complete without issue.
TheTik
TheTik4mo ago
Running it now, let ya know what happens @Jamin Collins [VC 3.1 400mm] Wait, I'm doing this test with the SuperPinda. That probably won't help you
JaminCollins
JaminCollinsOP4mo ago
@TheTik I think it still will as I don't think this is a probe issue I believe this is a z motion issue it's simply not moving the same amount back up for some reason if you look at the video The last iteration it just stops several millimeters short. So there's absolutely no way the probe could trigger
TheTik
TheTik4mo ago
ah, gotcha. I know you know your machine but maybe the z couplings are loose? Missed steps seem like something you'd notice audibly Interestingly, my probe number has gone way down as it continues. Started at 0.95 and is now down to 0.46
JaminCollins
JaminCollinsOP4mo ago
No noise from it. And I'd think something loose would show over one of the previous hundred or so iterations. Like variance in readings, etc.
TheTik
TheTik4mo ago
No heaters on so I dunno what would be making things grow/shrink
JaminCollins
JaminCollinsOP4mo ago
I'm doing mine could as well. Just motion And if you look through the console output, it's pretty consistent
TheTik
TheTik4mo ago
Mine is consistently dropping. Weird
JaminCollins
JaminCollinsOP4mo ago
Well, completed the sequence on another triple-z printer.... my tri-zero...
TheTik
TheTik4mo ago
Mine is still running, but the z= is now negative. -0.47 and still falling Still going, z=-1.87 Where is the console logged so I can save/share it?
JaminCollins
JaminCollinsOP4mo ago
I've just scrapped it out of the console tab. It's intermixed in the klippy.log with all the other entries. I'm guessing yours has gone beyond the scroll history though
TheTik
TheTik4mo ago
way past
JaminCollins
JaminCollinsOP4mo ago
Glad yours is able to complete the iterations
TheTik
TheTik4mo ago
I can grep it out of the klippy.log its completing it, despite the z value changing by almost 3mm so far it just isn't changing much between attempts
JaminCollins
JaminCollinsOP4mo ago
Wish I had any lead on what would cause the short move though How is your pi powered?
TheTik
TheTik4mo ago
double power double ground pins from Octopus
JaminCollins
JaminCollinsOP4mo ago
Hmm any under voltage warnings? And what gauge wire did you use?
TheTik
TheTik4mo ago
Never had any undervoltage warnings. I think I used 20ga
JaminCollins
JaminCollinsOP4mo ago
I am getting under voltage warnings. Can't draw a complete logic path on why that might matter, but considering it. I mean moves should be handled by the octopus, stepper drivers, and motors. Pi shouldn't really be involved other than making the initial request to move. And can't see a reason the move request wouldn't be for the whole amount. Another difference is 2240 drivers vs 2209s. I guess I could go back to 2209s on it to rule that out. Moving the PI to a dedicated power supply and retrying 259 and counting so far 600 and counting so far may have found the fix, but don't understand the why probably another 30 minutes before it'll complete (if it does) 820+ 🤞 well sheyet:
probe accuracy results: maximum 2.083959, minimum 1.978656, range 0.105303, average 2.058567, median 2.066652, standard deviation 0.079054
probe accuracy results: maximum 2.083959, minimum 1.978656, range 0.105303, average 2.058567, median 2.066652, standard deviation 0.079054
So seems like under-voltage is the cause?!?!? Will run another 1k for verification Well, that's 2x 1k iterations without issues
probe accuracy results: maximum 1.988045, minimum 1.972307, range 0.015739, average 1.980631, median 1.980752, standard deviation 0.005859
probe accuracy results: maximum 1.988045, minimum 1.972307, range 0.015739, average 1.980631, median 1.980752, standard deviation 0.005859
The only change being the RPi power source... would love to understand this more
JaminCollins
JaminCollinsOP3mo ago
RC: incorrect messaging, problem is in fact a communications timeout: https://github.com/beacon3d/beacon_klipper/compare/master...bugfix/comms_timeout_error
GitHub
Comparing master...bugfix/comms_timeout_error · beacon3d/beacon_kli...
Beacon Klipper Module. Contribute to beacon3d/beacon_klipper development by creating an account on GitHub.

Did you find this page helpful?