Crashes detected in Y-Axis without apparent reason. Resulting in layer shifts.  

RE: Crashes detected in Y-Axis without apparent reason. Resulting in layer shifts.

does the axis move freely by hand?

could your ubolts be too tight?

could your idler bearing bolt be too tight? or the bearing be faulty?
could the motor pulley be rubbing against something?

regards Joan

RE: Crashes detected in Y-Axis without apparent reason. Resulting in layer shifts.

Hi,

I was looking for someone with the same issue. I've assembled a Mk3s+ two weeks ago. The print quality is very good there is nothing to complain. Unfortunately I do get Y-Axis crashes during prints after 1:30 hours. Like two Y-Crashes every time I print a Benchy. And mostly my hotend is not even touching the printed part.

Things I have checked so far:

1. Motor pulley is not rubbing against something.

2. Belt tensions is at 261.

4. Y-Axis runs entirely smooth.

5. Belt idler moves freely

6. Bearings have been oriented as suggested by the assembly manual.

7. My rods don't have any scratches.

Things I have done so far:

1. Lubricated the bearings and rods

2. Checked the bearing clips.

3. Changed the belt tension. There is no difference if loose (belt tension value 280) or tight (250).

Below are my xyz calibration details:

Y distance from min:

-------------------------------------

Left: 10.82mm

Right: 10.47mm

Measured skew: 0.12°

---------------------------------

Slight skew: 0.12°

Severe skew: 0.25°

[0;0] point offeset

--------------------------------

x     0.02 mm

y    0.34 mm

Anything else to check?

Regards

Christoph

RE: Crashes detected in Y-Axis without apparent reason. Resulting in layer shifts.

does the axis move smoothly by hand? 
could the Y axis bearing Ubolts be catching anything? 
Could the heatbed cables be snagging on anything?

regards Joan

RE: Crashes detected in Y-Axis without apparent reason. Resulting in layer shifts.

@joantabb

Hello Joan,

Sorry for my late reply, I didn't got a notification about your reply. Meanwhile, I just disabled the crash detection for my printer - it seems it does not work reliable.

does the axis move freely by hand?

could your ubolts be too tight?

could your idler bearing bolt be too tight? or the bearing be faulty?
could the motor pulley be rubbing against something?

I used the printer now for months, running 24/7 with perfect results and (almost) no issues - except a broken cable for the heat-bed temperature sensor. 

The axis is moving freely and I checked the tension values several times and adjusted the belt tension to almost the exact middle of the recommended range. If the steppers are disabled, the bed moves easily. So, I really found no reason for the detected crashes. 

This year I will replace the PINDA sensor with the SuperPINDA and probably also the bearing clips (and bearings). After this change I will enable the crash detection again and see if this made any difference.

Thank you for your answer and help!

Best,

-LR

RE: Crashes detected in Y-Axis without apparent reason. Resulting in layer shifts.

crash detection should not be caused by the Pinda, unless the Pinda is very low (Large negative live Z value)

regards Joan

RE: Crashes detected in Y-Axis without apparent reason. Resulting in layer shifts.

Posted by: @joantabb

does the axis move smoothly by hand? 
could the Y axis bearing Ubolts be catching anything? 
Could the heatbed cables be snagging on anything?

regards Joan

Hi Joan,

with the printer turned off, I can easily move the axis with my hand (single finger works too)

I've double checked all the screws and made sure not to tighten the bearing clips (have a MK3S+, so no ubolts) too much. My heatbed cable has plenty of space so it does not touch anything and is free.

I've re-assembled the y-axis two times and I can't seem to find any mechanical causes of a crash. Super weird. :S

For now I also disabled crash detection.

Btw: I do have firmware version 3.9.2.

Many thanks for your help Joan!

Regards

Christoph

RE: Crashes detected in Y-Axis without apparent reason. Resulting in layer shifts.

To follow up on something Joan referenced, what is your Z offset?  If the PINDA is quite close to the bed (large Z offset), you can actually get the PINDA running into the surface of the print.  If you have a little bit of ooze as the nozzle lifts, you can get a little spike of filament.  The nozzle will generally knock it down on the next layer and you won't notice, but before that happens, if the PINDS run into that spike, you can get a layer shift.  Once the next layer is printed, the spike is gone and unless you actually see the shift, its really hard to determine what caused it.

Ask me how I know....

RE: Crashes detected in Y-Axis without apparent reason. Resulting in layer shifts.

Posted by: @patrick-mcnamara

To follow up on something Joan referenced, what is your Z offset?  If the PINDA is quite close to the bed (large Z offset), you can actually get the PINDA running into the surface of the print.  If you have a little bit of ooze as the nozzle lifts, you can get a little spike of filament.  The nozzle will generally knock it down on the next layer and you won't notice, but before that happens, if the PINDS run into that spike, you can get a layer shift.  Once the next layer is printed, the spike is gone and unless you actually see the shift, its really hard to determine what caused it.

Ask me how I know....

My current z offset is -1.110. My SuperPINDA has been adjusted as described in the assembly manual using a zip tie. I guess that height should be enough for smaller  spikes of filament.

I've printed some models, mostly 5 or more hours long. Haven't had any layer shifts even though crash detection has been disabled. Currently checking my y-carriage as it seems to rattle just a tiny bit. I noticed that my rods and bearings have very little play. Not sure if this is normal. The heatbed of my anycubic printer is rock solid.

RE: Crashes detected in Y-Axis without apparent reason. Resulting in layer shifts.

I too have this problem. There is a github issue: https://github.com/prusa3d/Prusa-Firmware/issues/2939

This is my 2nd prusa and 4th printer overall I assembled.
My printer prints perfectly. Axis all move very smooth. Belt tension is correct as well as number reported is 260.
Steppers are cold as I added 2 or 3 (40mm x 40mm) heatsinks to all of them.

I get about 2 crashes per hour of printing. At totally random locations.

RE: Crashes detected in Y-Axis without apparent reason. Resulting in layer shifts.

Followed this thread.  I have a further questions.

Just finished assembling a MK3S+.  First thing, I updated the firmware.  And followed the Self Test and Calibration XYZ.

I attempted the Benchy as a test print.  The Y axis crashed and "lost" its calibration.  When it attempted to recover, it would resume elsewhere on the bed. This happened with the first layer.

I re-Calibrated XYZ many times.  I tried the batman emblem as a test print since its a short print.  The machine would continue to crash on the Y axis (as the crash reports indicate) at a very specific place printing the parameter of the print, crash, recover, crash and recover again to finish the layer.

I followed everything on this thread such as belt tension and lubrication.  The problem persists.

There is one thing, however, that is very different from my older MK3S.  When turned off, the Y axis can be moved but does not do so smoothly.  It is as if the motion has these "tabs" to overcome.  The movement of the Y motor has a "jerky-ness" to it.  It does not appear have this issue while printing.  Not sure if the crashing and this "uniqueness" are connected.

Am I missing something here?

Regards, Jack

Crashes but not sure which axis

I have two MK3s printers, Prusa Orange and Prusa Black. Both work pretty well they both have some oddities but no big issues normally. I have just upgraded Black to MK3s+ and Orange to MK3s+/MMU2s . I have had a crash issue with Prusa Black from when I first got it but the Orange, while it has crashed, does not consistently do it. I thought it was the rods (Z axis) but I am seeing a lot about Y crashes. My printer never displayed the axis just that it crashed. It will try to continue but will eventually ask if I want to continue. This has been a significant issue a few times when I really needed to get things done. I was hoping that the upgrade might fix the issue but it started doing it again. It happens on more complex parts usually. I was printing the upgrade parts and just did the group of objects and once it gets to about 20mm it just crashes galore. I have done maintenance on the printer and everything is pretty dialed in for accuracy. It always occurs after 20mm and it may be a bit worse after the upgrade.

Planning to:

1. Upgrade X/Z axis with new rods and mk3s+ parts
2. go over the Y axis again, check all the upgraded parts

Bearings worn out

I forgot about the crash stats on the Prusa. I looked and was having y axis issues on the Prusa Black. My bearings are in bad shape so I will be replacing the rods just in case and upgrading the the misumi bearings. I have gone over a lot of info about packing the bearings before putting them on. I will switch out the bearings on the Orange.

RE: Crashes detected in Y-Axis without apparent reason. Resulting in layer shifts.

I think our problem come from new firmware 13.3.0-6873 , i see where the error was activated, and isn't any reason for this, finally i deactivated the collision detection and now is working without problem, isnt a nice solution, but prusa don say anithing.