The problem is more specific to this particular model. Because there are so many retractions on the initial layer, it has issues providing sufficient material to get adhesion. Direct drive extruders shouldn’t be retracting beyond 2mm. Otherwise, it takes too much material out, so it can’t reprime as quickly. With a less retraction, you’ll have more material available. If you don’t mind humouring me, can you try it with the lower retraction?
205C nozzle
70C bed
1mm retraction
0.25 layer
Questions.
- stringing?
- upper right square. The inner corner got all messed up. I can’t believe this is a leveling issue.
- Many of the other squares are just messed up.
The squares are very smooth where they are not visually messed up.
yup welcome to my world
@nivekmai and I are discussing if this really matters (over on the very quiet discord) - for example if it only applies to a first layer and one has good adhesion does it matter for most objects.
I think it might as i believe it can exacerbate warping (i had to try printing the same model 6 times before i could stop the warping causing head to catch and make model fall over). also one could just make sure the initial layer is the hidden layer of any model. In my case the poor squares layer are .4mm high and good squares are 0.2mm high (my model and initial layer settings are 0.2.
I have no fixes just thoughts:
- i have seen this model printed by friends on other printers with no issues
- though they have smaller build plates so really need to see this printed on say ender 5 plus for reference on what is acceptable
- for me i get same results on stock plate, glass build plate, energetic pei plate
- no amount of calibration fixes it for me (i have calibrated this machine maybe 100 times at this point, pretty certain i am not making a systematic issue.
- love your use of a mirror
- if this is bed levelling i wonder why ABL can’t cope with it, i note there are github issues and marlin conversation like this around 3 years ago, but they seem to have dropped off since marlin moved to UBL - maybe just coincidence / me grasping as straws
- maybe the test is bad (but see thought 1)
- slicer settings might be pivotal - but i haven’t found any that help, hoping someone finds some…
i currently have my linear modules disassembled (at direction of support) to try and fix once and for all (my cause might be the minor wobble) but @nivekmai doesn’t have any wobble and still sees same above.
tl;dr would love to know if anyone has printed this test and has zero under extrusion on all squares (on snapmaker or other printer).
(PS support did print this test and found no issues on their machines, though they never did the light box test and used white pla)
When I have issues like shown in the test print, the head catches the next time around and pulls the entire thing off the bed.
I’m beginning to think it isn’t a calibration issue as well.
I’ll file a support ticket to see if the support guys can print my STL OOTB.
1 Like
I can see smudging in the mirror. Was the mirror smudge-free prior to starting? Any fingerprint, or other smudging, will prevent the filament from adhering to the mirror. Once they fixed the bug with the mesh bed level, I haven’t had any issues with corner-to-corner, or edge-to-edge, prints. For PLA, the bed should be 50°C, not 70°C. You can probably get away with 60°C for the first layer only, but other than that 50°C is the absolute maximum. Snapmaker is completely wrong in their assessment with that.
I heard back from support. They suggested 0.3mm initial layer height and 300% initial layer width.
This broke luban and cura as they couldn’t handle an edge that was < 300% line width. They just didn’t slice it so I lost my entire perimeter.
I then moved back to the original height sensor and print bed. I did like the original print bed from first layer printing.
I then decided to start from the beginning and recalibrate everything. Turns out at some point my estep calibration got overwritten with the default, so I was too little by 5%. I always felt that one issue is that my initial layers were too thin. Not sure how my printer lost these settings, but they are in there now.
I then printed the x y z calibration cube and it turned out nicely. A small amount of elephants foot, but crisp everywhere else.
I spent a huge amount of time leveling the bed, shimming with paper many of the mounts to the frame. Previously I was hesitant to spend too much time shimming the bed as I would lose it every time I wanted to use the laser head. With someones suggestion to just mount the laser bed to my original/destroyed print bed I was fine perfecting the bed with paper shims.
Adding the linear rails helped beyond belief.
All this to say, it is now printing better than ever and I have a print going right now. I decided to R180 the print such that the first layer is not the last layer and that is the layer that matters the most.
This made is such that I had to print a bit of support structures for the first layer, but they turned out well.
There are still some dribbles from the print head, but I’ll tweak that. I think that is because it is still a smidge too hot. It is not on the 5th layer and the next 42hours are the exact same thing, so I’m happy at the moment.
My final slicing paramters:
MH build series PLA
Head: 205C
Bed: 65C
5mm retraction
15mm/s first layer speed
Lessons: even if you have calibrated esteps, do a M92 and verify 
If this one works, then 17 more to go. I will be printing for another 35 days! Hopefully nothing breaks . . .
Firmware upgrades can reset these, changing/replacing modules/rails can reset it too. Glad you made good progress. Having to do add additional linear rails to the platform IMHO is a bit extreme, but guess whatever gets it to where you need it.
Thanks for all the help. The print finished nicer than any other I’ve done.
And a quick video of its purpose.
I use 200C and 50c for my bed. Perhaps it’s not adhering because it isn’t cooling enough in that spot on your bed.
I do alot of full bed prints.
Nope, my issue is y linear units.
Scyto, you working through it with support?
@Mads0100 he is, they’re going to make sure the replacement rails are fully in spec and working before they send them out to him because the ones he got were sent to him for the previous ones that went bad.
Ahh. It happens. Support is pretty awesome.
yes, two rails arrived yesterday, one looks like this (i only noticed once i mounted it to the bed).
waiting for them to tell me if this is safe to use or not.
tbh i am almost done with this lemon, i get others have no issues, if anyone in the PNW wants this for parts give me a DM.
2 Likes
That is horrible. Can you see any of the copper conductors?
I too had to get a replacement Y rail because it wouldn’t adjust to be within spec. I got my replacement, but haven’t opened it because when I added the linear rails, it fixed everything for me.
no, it seems it is just the black outer, though the pinch on the blue conductor concerns me a little, i continue to look at my two totally disassembled faulty units to see if i can do a cable replacement or not, haven’t figured if it is possible yet.
1 Like
While not ideal, I’d probably just deal with it and wrap it in electrical tape. I wish they’d just send you the new rails for free and call it fair for all the troubleshooting and issues you’ve had.
i held out a small and tiny hope they would give me a better deal on the new rails (like nothing); but as i had 6 rails in total, their ask to combine bits from rails is reasonable… i am actually going to try repairing it at somepoint using a cable off another rail…
1 Like
Yea, I guess on warranty repairs, you’d hope they’d look at the product they’re sending you.
So does the new rail fix your issue? Are you printing at this point?