Recently my Snapmaker 2 with dual extruder print head has gone crazy. Printing starts OK but then something goes wrong. The surface is supposed to be completely flat. I am not sure where the artifacts are coming from. You see what happens in the photo. Eventually the print head hits this ridge. Any idea what is going wrong here? BTW, I tried to generate the gcode with Luban as well as PrusaSlice just to make sure the issue isn’t with the slicer.
Simplest way for such a short print would be to observe the print.
I guess you have feeding problems at the cross line, maybe too near to the bed, so it curls up, hard to say…
If the nozzle cross the print there would be a mark or scratch but not like this curled up.
Is it reproduceable?
share the code so we can say that’s not it
just spitballing…there are two things involved here. one being head/gcode the other is the bed.
i have had these type of build sheets i see in the photo have issues if not careful. i had one that wasn’t really magnetic so it didn’t hold well to the heat bed. i had another that appeared to hold well until i put a large abs slab the size of yours on the plate. the abs stuck well enough but when the abs wanted to warp the build plate didn’t stop the movement. ie. the plate was “bowed” by the print instead of the plate winning and holding the print down.
so…do you think the bed could be warping?
It is unfortunately. I am now testing again with Snapmaker’s original build sheet to see if this is the cause. I should know by morning.
The build sheet may be the issue, I am not sure. I used this one for a while but then only for much smaller prints. I am now testing again with Snapmaker’s original build sheet to see if this is the cause. I should know by morning. If it is, I wish Snapmaker made their own powder coated build sheets because I like them a lot more than the smooth ones.
I will post the result of my test.
The build sheet wasn’t the issue. It now happened quickly using the Snapmaker original build sheet. I checked the generated gcodes with their respective viewers and they look good. I also opened the STL in Meshlab and it looks fine.
I am starting to think it may be a firmware issue. The printed object is the max size that fits on the A350, it is exactly 320 mm so I may have hit on a firmware bug?
I shrunk the model to 310 mm to see if it helps. If not, I will shrink it further.
Update: 310 mm was having the same issue. I am now printing it at 300 mm and so far it looks good and clean. If it finishes OK I will give Snapmaker all this information to see if they can replicate it and hopefully fix it.
BTW, I am using PLA so warping should not the issue as can be with ABS.
Link to PrusaSlicer generated gcode
Can’t believe it’s from the gcode, so it has to be a physical issue.
I believe it won’t occur if you print with the single extruder. In this case it’s necessary to check the linear guide rails of the used extruder, do you feel or see any wobble of the nozzle? (note, you have to set the extruder active in Luban by T1 or T0)
Another point to check is, why is the extruder pulled up in the middle, or isn’t it?
Of course it’s only a guess. When I look at your first picture it seems that the print head scrubbed over the material. I had that issue a time with my (single) print head also. After calibrating the Z height using the machine itself it got a lot better.
Another thing is the temperature. In the gcode file it shows 210° for the nozzle and 60° print bed temperature. In case of PLA, try it with lowering both, i.e. 190-200 for nozzle and 50° for the bed. I would also set the first layer speed (better: first 3 layer) extremely slow. In my experiments these settings worked a lot better.
Too much heat on the nozzle can result in minimal clogging on the outside of the nozzle and this material can scrub over your model leaving “nice” artifacts. And the sensor of the print head can scrub over the material and rip off something then. Had that when I was printing over night and at the morning I had a big fat clump of PLA on the print head which hardened and the printer still scrubbed that over the heat bed with a sound you don’t want to hear from your printer… 
(Luckily it has not destroyed my machine…)
Although PLA may not be such problematic in warping it still can happen depending on the PLA vendor. And too much heat bed temperature works like a chimney concentrating the heat at the open surface as the rest is covered with material and there the material can bend up.
And: The thicker you print the hotter your material stays and this can also result in bending the material. You have 0.24 layer height, I got the best results with 0.08 only. Of course it prints a lot longer, but it also cools down better.
In the picture you see on the right side my PLA clump, the rectangle inside is the form of the print head and nozzle…
On the left side this was a test print with a multi-color PLA with too thick output and too much heat, which always resulted in bending up the middle part (which is the longest line length and therefore the hottest), the whole thing should be flat.
And in this picture it is the exact same model, printed with a very fine line strength (of course loosing the multi color effect) with better temperature settings.
Since the problem occurs with to different slicers, and the STL has no structure in it that would explain this effect, I’d also infer that this is something with the machine. My wild guess would be on the bed leveling - the cross-like shape is too obvious to be a coincident. I guess I’d redo bed levelling, potentially with a different pattern spacing. Also, I might try moving the print a few cm left or right - if the cross moves with the print and stays centered on the print, this would indicate that it after all is a slicing issue, but if the cross would stay centered on the print bed (i.e. offset from the object’s center), it would confirm that the machine misbehaves.
I am convinced it is not a machine hardware issue. As I mentioned earlier, I believe it is a firmware issue because of some boarder condition I am hitting due to the size of the object I am printing. I had this problem when it was 320 mm and 310 mm, but when I printed it at 300 mm, the print came out with no issues.
Are you using the vibration compensation beta? I printed larger boxes with no problems at all with the regular firmware…
I don’t use the beta.
I was too quick to claim that it works fine with a 300 mm print. This I tried printing the bottom part of the box and it worked fine for the 1st layer but when it printed the 2nd layer I got this.
If it is a hardware problem then I have no idea what it may be. I can try to switch to the single extruder tool head to see if it solves the problem. I am very frustrated right now.
Do you have quick Swap or Bracing Kit?
Since official Specs you cannot print 320 in Y axis with any of both:
Maybe turn the object by 90° so you print large side in X?
That’s what I did. I also dropped the size to 300 mm but the issue is still there. I am going to try with the single extruder tool head and see what happens.
This looks more like a settings problem - my guess would be either printing too cold, or slight underextrusion, or a combination. It seems like the filament is not sticking well to the layer below and then warps.
The First layer was perfect? Normally this appears in too Low z or on overextrusion. To much Material which is squished next to nozzle.
That was what I thought first, but if you zoom in into the picture, you can see that it is layer lines delaminated and deformed, not squished material! Thus my assumption above…
Upping the temperature didn’t change the result. I’ll check underextrusion next.
I am also wondering if there is an issue with the bed because in the bad area I was able to separate the two layers easily and get two sheets, but not in the area that looks good where the layers fused properly. Maybe a warped bed? Will I be able to see if this is the issue by using a straight ruler?
The Snapmaker 2.0 printing bed is unfortunately infamous for being bumpy - some people have worse, some have better print beds. There are several looooong thread in the forum - this perhaps being a good one to go down that rabbit hole: Bed Leveling - Revisited & Detailed
I do have a nasty, bumpy bed - I’ve got the Kickstarter machine, and those were especially bad. Still, for me with the dual extruder the bed levelling with the densest pattern (I think it was 5x5 - or was it 7x7? Sorry, currently have a job running on my machine, cannot look) yielded good results, but I yet did not print large things with the DX. With the single extruder, I did 11x11 patterns (you need to do that via GCode, not via Touchscreen) and was able to get a really huge good print: Maximum Y Position 347mm? - #9 by Hauke. The DX unfortunately uses a different levelling procedure, and I never tried if via GCode you can go for larger patterns.
What worries me most is: With the single extruder Snapmaker uses standard marlin methods, which means that the bed levelling is not only used for the first layer, but for subsequent also. There’s an option to slowly “phase out” the levelling layer by layer. For the DX, with Snapmaker’s self-made, homebrew levelling I would not think it impossible that they only apply levelling data to the first layer, which would explain your problem… perhaps Snapmaker support can help here? If they really messed this up, the community should push hard on them to use the Marlin method also for the DX!
This is great information, thank you!
Rabbit hole indeed but fascinating. I think I have somewhere a torque screwdriver so I will start with making sure all screws are using the same 5inlbs recommended in the article and see if it helps with the warping. I will also try the 11x11 grid calibration via the console because it is not available via the touch interface for the single extruded.
1 Like