Y-Axis fake(?) Crash detections - let's debug
That Thread is an offspring of this one
I want to find out how many ppl have it and if we can find out why this is happening to some of us.
So I have 'a lot' of Y-Axis Crash detections and I can't see why. Let's say 10-18 crashed while printing a PLA Benchy at .15 layer height. On the other hand I've printed other stuff without a crash...
What I've seen so far is, that it seems to print just fine and for some reason it detects a crash. There was no obvious bump or "upcurl" or something like that. When I move the Y Axis with my hand it feels fine. Everytime I witnessed it, it happend while printing - not while travelling.
Let's collect some data 
Kit or Pre-Assembled?
[X] Kit / Pre-Assembled [ ]
Firware
3.1.3-245
What Slicers did you use / did it make a difference?
[X] Slic3r
[X] S3D
[X] Pre-Sliced SD
No difference imo
What are the Numbers on your Belt-Tension (Menu -> Support -> Belt Status)
X = 265
Y = 253
Which Version of the Y-axis Assembly / Heatbed Assembly do you have?
I have ne 'new' Version with 3 Screw Cover and textile sleeve with no Nylon Filament in it. http://manual.prusa3d.com/Guide/7.+Heatbed++%26+PSU+assembly+%28textile+sleeve%29/589?lang=en
Only Y-Axis or also other crashes? (look it up in the fail stats -> Total)
Just Y
❗ If you have no crashes, please let us know too, but make sure that this is really the case (look it up in the fail stats -> Total). Don't just assume you have none!
Here's the template
[b]Kit or Pre-Assembled?[/b]
[ ] Kit / Pre-Assembled [ ]
[b]Firware[/b]
[b]What Slicers did you use / did it make a difference?[/b]
[b]What are the Numbers on your Belt-Tension (Menu -> Support -> Belt Status)[/b]
X =
Y =
[b]Which Version of the Y-axis Assembly / Heatbed Assembly do you have?[/b]
[b]Only Y-Axis or also other crashes?[/b] (look it up in the fail stats -> Total)
Re: Y-Axis fake(?) Crash detections - let's debug
[reserverd for some findings]
Re: Y-Axis fake(?) Crash detections - let's debug
My stats:
build = Kit
slicer = PrusaControl
belt tension = x 269, y 266
heatbed assembly = 3 screw cover and textile sleeve
firmware 3.1.3-245
crash x 0, y 204 (about 10 prints so far)
edit - added firmware version
Re: Y-Axis fake(?) Crash detections - let's debug
Everyone should also add to their info the firmware they run.
Re: Y-Axis fake(?) Crash detections - let's debug
Here is a 7-second video of a y-axis crash detect ... they happen seemly at random, on non-challenging objects.
Re: Y-Axis fake(?) Crash detections - let's debug
Kit or Pre-Assembled?
Kit with 3.1.3 fw
What Slicers did you use / did it make a difference?
Slic3r PE 1.39.1
What are the Numbers on your Belt-Tension (Menu -> Support -> Belt Status)
X = 257
Y = 263
Which Version of the Y-axis Assembly / Heatbed Assembly do you have?
3 screws cable cover and spiral sleeve with nylon filament
Only Y-Axis or also other crashes? (look it up in the fail stats -> Total)
9 Y crashes in 240 hours printing time, 12 X crashes.
MK4/MMU3 - VORON 2.4 350 Stealthburner
Re: Y-Axis fake(?) Crash detections - let's debug
Everyone should also add to their info the firmware they run.
You're right. Thx. Edited 😉
Re: Y-Axis fake(?) Crash detections - let's debug
build = Kit
slicer = PrusaControl
Belt Status: 250
Heatbed with Nylon and spiral wrap
Firmware 3.1.3
9 Crashes during a 1.5 hour print. Dissabled since then. No Problems with it dissabled.
Cheers
Oliver
Re: Y-Axis fake(?) Crash detections - let's debug
build = Kit
slicer = Slic3r 1.39.1 Win x64 using MK3 214 settings (Generic PET as filament, 0.2 fast)
Belt Status: X 251 Y 259
Heatbed with new cover (printed in PETG) and sleeve (no spiral and no filament)
Firmware 3.1.3-245
4 Y Crashes in 1d 22hrs 30mins (total print time). Crashes all happened during a print for no apparent reason just as other have described.
Didn't know I could disable it -- looking for that and will disable detection.
P.S. Unrelated but I turned on filament sensor once and it stopped in the middle of a print asking me to remove the filament (PETG). ?? Turned that off again 😉 So, that is also in my crash stats...
Re: Y-Axis fake(?) Crash detections - let's debug
have nearly every print 1-2 fake crash detections, this is a preassembled mk3 with latest firmware send to me end of march using it for two weeks now
and just now he does not find the right height while moving back again after the crash and stopped also moving the material... so i had to stop the whole 6 hour print after 70%,
this is annoying!
How can i disable this in general?
Re: Y-Axis fake(?) Crash detections - let's debug
How can i disable this in general?
You can either print in Stealth mode, crash detection does not work in Stealth, or you can turn it off in your Settings Menu (Settings -> Crash Detection Off)
Take a look in your Handbook that came with the printer. You can find all Settings / Menus and what they do in it.
Back to Topic:
I have 2d11h and 152 Y-Crashes I loosened my Y Belt a bit (261) and the crashes went down...
I also ordered new Bearings and printed new Y-Axis parts from this Thread -> https://shop.prusa3d.com/forum/others-archive--f66/fine-tune-your-mk3-y-axis-t13851.html
I'll try those when I change the bearings.
Re: Y-Axis fake(?) Crash detections - let's debug
I was getting frequent y-axis bogus crashes. I got them very infrequently until I followed support's instructions to tighten the belt. I think the values they are suggesting (near 240) are way too tight. I could feel increased effort to move the platform after tightening this much.
I completely eliminated the crashes by loosening the belt back up (I think it's 272 now), and by loosening the y-axis u-bolts a little. In my case, one of the u-bolts came bent, and although it fit, this was (I believe) causing excess pressure on one of the bearings.
So to those having this issue I would suggest:
Loosen y-axis belt.
Loosen u-bolts - you just want them barely tight enough to prevent movement of the bearing - not the 1/4 extra turn suggested in the instructions.
Re: Y-Axis fake(?) Crash detections - let's debug
ok next print with crash detection off: result heavy layershift with no reason
i am loosing trust in this product... what is going on here?
Re: Y-Axis fake(?) Crash detections - let's debug
ok next print with crash detection off: result heavy layershift with no reason
i am loosing trust in this product... what is going on here?
This looks more like "top layer does not bond to bottom layer" to me. Go slower / hotter / lower leigher height.
Re: Y-Axis fake(?) Crash detections - let's debug
ok next print with crash detection off: result heavy layershift with no reason
i am loosing trust in this product... what is going on here?
This looks more like "top layer does not bond to bottom layer" to me. Go slower / hotter / lower leigher height.
no it is a x and y shift and i was able to print this part one time fully but now 3 times i had to stop the prints always after crashes happend where also the printer unloaded the filament without reason and did not reload before resuming the print so he was printing without filament ...
Re: Y-Axis fake(?) Crash detections - let's debug
the last test i cold look to serveral crashes, it almost happens while printing inner outlines (faster one) lower the speed to 75% does not solved this
Re: Y-Axis fake(?) Crash detections - let's debug
https://vimeo.com/michaelradeck/review/264502305/2e6717d852
here i just startet the video after one crash just in that moment the next happend, he already printed without filament
https://vimeo.com/michaelradeck/review/264502689/28e1380216
the unload resultet in a hang of the filament:
Re: Y-Axis fake(?) Crash detections - let's debug
Kit or Pre-Assembled?
[ ] Kit / Pre-Assembled [x ]
Firmware
3.1.3-245
What Slicers did you use / did it make a difference?
CURA latest
What are the Numbers on your Belt-Tension (Menu -> Support -> Belt Status)
X = 256
Y = 259
Which Version of the Y-axis Assembly / Heatbed Assembly do you have?
I have ne 'new' Version with 3 Screw Cover and textile sleeve
Only Y-Axis or also other crashes? (look it up in the fail stats -> Total)
yes only y-Axis
but also the extruder sometimes unloads the filament after the crash and does not reload before resuming - i have to stop the print
Re: Y-Axis fake(?) Crash detections - let's debug
So, update Time, as said earlier I changes the bearings for Misumi and also printed these modified Y-Axis partsl.
Bottom line, it's better but it's not gone.
However, my test-prints where this one https://www.thingiverse.com/thing:2795629 (have enough benchys already, wanted something different 😉 )
The crashes happened quite fast. I think I printed them with about 18 Y crashes on average.
After the swap I got it printed with 0 crashed two times. However, I printed some other stuff with 1-2 Y crashes still happening. 😥
I watched it print a bit more and still can't figure out what's causing it. When the crashes happen while printing the ship it just looks like it crashing into thin air. However, on occasions, when the print lifts slightly up or the plastic is curling up a litte, the nozzle just rides over it like over a bumpy road an it just doesn't care. So I don't think the crash detection is just too sensitive or something. It's really odd.
So for the others: try to loosen you belt a little bit. Try to use the Manual and you finger. Don't care about the numbers in the menu. See if that helps you.
Re: Y-Axis fake(?) Crash detections - let's debug
So, update Time, as said earlier I changes the bearings for Misumi and also printed these modified Y-Axis partsl.
Bottom line, it's better but it's not gone.
I have had repeated y-crashes for some time. I also upgraded to all Misumi parts and my X-Y axis is super smooth -- absolutely no binding. However, during certain prints (often with honeycomb infill) I still get fake crashes. The prints are high quality so I'm sure all other setting are fine.
Today I started to experiment with changing the Y motor current and sensitivity for crash detection and no crashes thus far...
g-code ref:
M913 -- will read running and hold currents for each axis. Note with the TMC2130 these are not values in mA, but rather a numeric value. See https://shop.prusa3d.com/forum/-f23/einsy-rambo-on-mk2--t6137-s30.html#p74913 for more info.
the default values (as verifed in the firmware) are:
#define TMC2130_CURRENTS_H {16, 20, 28, 36} // default holding currents for all axes
#define TMC2130_CURRENTS_R {16, 20, 28, 36} // default running currents for all axes
M911 -- will set hold current
M912 -- will set running current
e.g. this will up the Y-axis current From numeric value of 20 --> 25
M911 Y25
M912 Y25
M916 (no args) -- will dump the TMC2130 stall sensitivity. By default now all axes are set to '3' (the Y-axis was higher in the original 3.1.0 firmware). I believe from reading the TMC2130 spec sheet this value can be from -63 to +63, with 0 recommended as starting point.
Anyway to change a value pass arguments like this:
M916 Y6
This will decrease the sensitivity of stall detection in the y-axis a little. From 3 --> 6).
Bottom line. If you really think you have some minor mechanical binding you could try upper the currents for the y-axis with M911/M912. If you think the stall detection is too sensitive you could decrease the sensitivity.
I'm currently testing with:
M911 Y25
M912 Y25
M916 Y5
I would be really interested in the results from others.