Print pauses and nozzle touches bed and resumes print

[Shaggy]

My printer will sometimes pause printing and the nozzle goes to the corner of the printing bed and touch it then raises back up to the level of the print, then continues printing.

I feel it is causing/caused by some z-bending/wabbling issues. before the nozzle would collide with the print and do this same motion, but now i do not hear any collision when this happens (not to say its not colliding, but i don’t hear it anymore).

I have changed both the filament settings and printer settings to give a 0.8mm Z hop when retract (the printer is defaulted to have 0.4mm, but i did edit the filament setting override to have z hop when retract to 0.8mm also). I also put both Z Hop Types to be “Spiral”.

I have also changed the print speed to reduce by height by adding modifiers to slow it down to the slowest speed as it gets higher (you can see this in the 3mp i uploaded under objects and when slicing and showing details of speed). HOWEVER, i hadn’t changed the acceleration speed at starting and ending heights, which might help, but not sure if this is the real cause of the problem, but ill change the acceleration for my next prints, just not sure what to change the value to.

it does it when printing the piece at an orientation of 90 degrees, or 45 degrees like this one is positioned on the bed in the video I linked. Print orientation does not fix it.

Any ideas/suggestions to fix it, and what is the cause?

• here is a link to a video of the problem in action: https://youtube.com/shorts/3AV9LYO-Phs

• here is 2 pictures of the same print, but in two different angles
  
  
  
  
  

• here is an upload of two of my sliced models (the custom tree STL supports was to help stabilize it when they get higher, but it happens without these also) Both are printing up to 250mm z-height:
  2X Sup_IZ (250mm version - 3 supporting screws).3mf (60.5 KB)
  3T INF_IZ (250mm version-loose fit-infilled supported-speed adjusted) v1.3mf (69.9 KB)

Bambu Labs A1 printer

• Slicer settings used: see the .3mf uploads for all the settings
  Currently printing with Elegoo PLA+ black (and white) filament. But this happens with any filament I have used so far.

Again, this can happen on a lower height and does, but it tends to do this more frequently at higher heights as the print is printing.

[R3d-Five]

Set the printer back to default and run calibration. When you start changing things “guessing” that they’ll help, doesn’t help. Troubleshooting needs to be one thing at a time. I think you’ll get better prints right after calibration.
Once you have a base you can troubleshoot the problem rather than taking a stab at it by changing settings.

[Bullocks]

Tall wobbling things on A1 can benefit from slowing down acceleration(cap to 2.5k) and jerk (cap to 6, but try just accel first).
Maybe benefit as well as toggling OFF the “Reduce Infill Retraction” setting. You want it to Z hop and retract, to give you movement clearance over the wobble.

E: the nozzle parking and z-traveling to touch the plate is a part of “nozzle clumping detection”. Maybe try turning it off as a test, and digging in further if that changes anything.

[Shaggy]

ive done a calibration right before this current print I recorded the video on, i did a full calibration with all boxes checked.

[R3d-Five]

Have you updated firmware recently?

[R3d-Five]

check timelapse settings, I know it’s here and there but check them, they could be the issue.
Are you sending the file through Handy or BS? If not check the card your using, and it wouldn’t hurt to check the one you are using anyway

[Shaggy]

okay I will cap the acceleration to 2.5k and toggle off the reduce infill retraction.

i saw your edit just now as well… I will try turning off the “nozzle clumping detection” now for my next print and see if it helps.

I am assuming if these changes do help, i would turn nozzle clumping detection back on and do another print to see if the problem occurs again, because it could be possible the acceleration changes might fix it as well.

I also wonder after looking at my pictures if I do have something off with my z-rod because at least in the first picture it seems like there is a pattern of wobbling (but the second picture doesn’t show this same pattern, yet the same model printed, just that both sides were at different 45 degree orientations).

[R3d-Five]

Well z issues will cause the banding. I’m more concerned with the stop. I’m thinking either the SD or timelapse.

As for the Z, take a good look at it, check for lose screws etc. Adjust the tension etc

[Shaggy]

firmware is up-to-date, i have timelapse turned off. I am sending the file through my computer Bambu Studio directly to the printer, not manually with an SD card. Im unsure what Handy is.

The card is the default that came with the printer, though from my time resin printing I know its good to change that out for a new one. If I do change the SD card out, what model of SD card is good for A1 printers?

[R3d-Five]

The same for resin. I have them as well. Just a reputable brand name, FAT 32 etc.

[Shaggy]

yes my bigger concern right now is the stopping like shown in video as well. Sadly these prints are about 10 hour long, so ill print using the suggestion Bullocks gave right now, but will be doing all other suggestions as well for the next prints

[R3d-Five]

I have his replies hidden, so I’ll guess you’ll on it. Good luck.

[Shaggy]

I went ahead and turned off “Reduce Infill Retraction” setting, and slowed down the acceleration to 2.5k from 70mm - 250mm height (250mm is the max height of my print).

I did leave the nozzle clumping detection on, as this is the issue im trying to solve, but i am glad you let me know that the nozzle parking and z-traveling to touch the plate is part of the nozzle clump detection. But the next print if it doesnt solve it, i will have a new SD card installed in the A1 printer, and also update you guys on the issue. I then will do another print with it turned off if it continues.

I will also be printing z-wobble test prints after this current one finishes in 14 hours. (its those basic square prints that are like 254mm height that will have a much shorter print time and save on filament for testing / fixing).

[Shaggy]

speak of the devil, it did the clump detection on either 11th or 12 layer on this print. stopping it now and going to save time/filament and do a z-height test print till this is solved

EDIT: might need to change the acceleration speed for the whole print (starting at layer 1) to the 2.5k cap, instead of waiting till 70mm high

[Shaggy]

Okay so far so good with my test print after i slowed down the acceleration to a cap of 2.5k for every layer (see pictures, the little extra lines coming off giving it defects in the places nozzle clumping detection occured can be fixed by adding a tower [finish reading post to see why]). I did not switch OFF for the “Reduce Infill Retraction” setting on this print, so this is something to consider next time.

After knowing the main function is called “nozzle clumping detection”, i was curious to see what bambu wiki had written about it, and they put this little section at the bottom under note for prints above 70mm (see picture). I will do what they suggest next by adding a tower to my print Next I will add a tower. The wiki also said that if the “nozzle clumping detection” is turned on, " After printing more than 3 layers, the nozzle clumping detection is performed for every 8g of filament consumed " which would be one reason it always is happening. Yet, hopefully the frequency with the prints (not test prints) that it occurs will be reduced to only every 8g of filament used, and not more often due to a higher height.

will update again when my next print is finished (of the actual parts I want to print) finishes. I am hopeful this will fix it.

PS, it is strange how 1 of the 4 sides of the test square is giving me like a ghosting effect with a vertical line, I do believe this is due to the seam because there is 1 vertical seam in this test print going up and down 1 corner (the side I am referring about is in the first picture uploaded). I am uploading the 3mp file of the test print I am posting pictures of as well. (i did post 2 pictures of the same side (the first two pictures), hence being 5 pictures for a 4 sided square, because the lighting gave off a pattern effect that really isnt there)
EDIT: strange the first two pictures dont have the same effect that it shows on my phone, oh well, theres an extra picture of that side with the vertical ghosting effect happening. I am hopeful the problem is solved, but will confirm when i print my actual STL’s I am wanting to print besides the testing high square which is uploaded in this post.

Square 240mm(2).3mf (745.8 KB)

[Bullocks]

@Shaggy

The wiki also said that if the “nozzle clumping detection” is turned on, " After printing more than 3 layers, the nozzle clumping detection is performed for every 8g of filament consumed

Interesting, didn’t know that. And if it uses the prime tower to do its check, yeah it should be dramatically quicker. I admit I just turned it off but on jobs that use a tower, it might be worthwhile turning back on.

Overall looks pretty good.

The effect of the light from pic#1+2 could be a pressure advance slightly out of true but if you dont see it IRL it might not be worth worrying about.

[Shaggy]

Strangly I dont know how to add a prime tower to a single color print, i have the box checked but there is no prime tower. any help regarding this?

EDIT:

@Bullocks

Interesting, didn’t know that. And if it uses the prime tower to do its check, yeah it should be dramatically quicker. I admit I just turned it off but on jobs that use a tower, it might be worthwhile turning back on.

I am unsure if it uses a prime tower to do its checks, it just said if you get the stringy defects, elimate them by adding a prime tower (see last picture i uploaded above this one I am replying to).
EDITx2: ohhh i missed the part of you saying “but on jobs that use a tower […]” , ignore the edits

source: Nozzle Clumping Detection | Bambu Lab Wiki

EDITx3: i found how to add a prime tower to single colors, you need navigate to the “Other” tab, and then enable “smooth” under “timelapse”.

[Shaggy]

I do see the vertical line ghosting from picture 1+2, but not sure if that is what you was referring about (see circled part in picture). but everything else looks great (minus the layers that the nozzle clumping detection occured, but we found the fix according to the wiki already so im not concerned with that, rather I am turning “OFF” reduced infill retraction, because my prints already take alot of filament per piece, and the tower is adding about 50grams extra per piece (and i have MANY pieces i am building, for a storage case for minatures of a tabletop game called warhammer 40k)

[Shaggy]

I just noticed the speed of the test print, was slow from bottom to top (besides layer 1-3), yet the speed i set was wrong (i accidently set the parameters to ending height 25mm, instead of speed at ending height 25mm/s).

but even after taking off the speed parameters, the print speed is very slow on this test model. I dont understand why?

Square 240mm(2).3mf (745.8 KB)

again, yes i messed up on the speed parameters i set, but even after fixing them, the higher speed of 200ms doesn’t want to be 200ms. Just wondering why this is happening? (edit: look at edit number 3)

My current print is printing with the acceleration change to 2.5k we made, it will take another 13 hours so I will update with that piece tomorrow. Though, with this test print speed, i dont know if it really is going to change anything to my normal print, because i messed up on the height (click process > objects and see what i mean under the layers range 0.00-240.00 (mm) > speed. ) I also never changed anything related to the speed/acceleration in the normal “process>global>speed” setting.

its as if my printer is printing this square test print out fine, but my print that is more complex shape it doesnt want to print out as smoothly. Also, this test print square piece outter wall has a very slow speed from bottom to top, like 54ms or something close to that (in speed summary it shows the color thats simliar to 54 mm/s).

ALSO, my print of the original STL i am printing, just did the nozzle clumping detection DESPITE turning “OFF” reduced infill retraction???!?!? it just did it again as I am typing this message.
EDIT: just found out the setting on the printer itself was set to be turned on for it. I just now turned it off… I am hoping it turned it off during the print (like changing the speed mid-print), otherwise ill redue the print again tomorrow afternoon with the setting now turned off.
however I still am curious about the reason for the speed limitation in the square test print above even though the outter wall is set for 200ms?

EDITx2: here is the .3mf of the current print file I am doing that is going to take a total of 14 hours, im pretty sure i did all the acceleration changes correctly.
Comander Z2_05 XL_CNT_INF (250mm version - 3 screw support) clearance v1.3mf (48.6 KB)

EDITx3: look at this fixed version of the square test STL, it starts most of the speed in about 54mm/s (after slicing look at the speed drop down in slicing results), yet i have it set to start the print at 200mm/s. then it increases to a orange/yellow 95mm/s around layer 940, then it goes back down to the 54mm/s speed around layer 1000, then it finishes up to the speed parameter i set which was to make the finishing speed at 240mm to be 20mm/s (dark pink).

Square 240mm - fixed speed parameters but speed summary looks strange.3mf (745.8 KB)

[Bullocks]

Reducing acceleration means it takes longer to get up to max speed (it might not, before having to slow down for the next curve), which reduces force applied to the big slidey plate and the stuff on it, to help it not fall over or become unsteady. This acceleration is the a in “F=ma” that Sir Newton wrote about. Jerk is the third derivative. This is a concern for moving-bed printers, as opposed to the coreXY kinematic printers.

From a “well actually” standpoint it is just the Y acceleration you benefit from throttling, that’s where all the mass is - on the plate. but like … Sometimes simpler is better