Crooked prints

[daduke]

Printer Model

Siboor V0.2

Extruder Type

Other

Cooling Type

Other

Dear Voron masters,

I built a 0.2 (Siboor kit) and it works beautifully. I have one problem however: some prints (but not all!) come out crooked - they're not straight in neither X nor Y. I don't think it's frame related (since others are perfectly fine), and the combination of X and Y makes me think it's a problem in either A or B. I can't tell which (kind of) prints trigger it, and I've had crooked prints succeed in a second or third attempt. What I've tried/investigated so far:

- A and B belt tension seem fine and very close to each other (both 110 Hz tuning and tensiometer)
- I've swapped out A and B TMCs, no dice
- increased/decreased motor current
- doesn't seem to be print speed related (at least not in an obvious way)
- general mechanical inspection, the motion system "seems fine"

It's as if the printer was missing individual steps, but always leaning in positive X and negative Y.

The pictures show two identical parts back to back and front to front.

I'm a bit at a loss right now and would highly appreciate any ideas where to investigate further. Thanks a lot.

-Christian

 

[vinnycordeiro]

• Check for bindings in the bearings of the motion system at some positions.
• Check if belts are touching something that they shouldn't, if they are riding the bearings/idlers/pulleys at the right position.
• Not that I think this is a reason, but how much cooling are you using? By the way, which material are you trying to print? You didn't mentioned on your post.

 

[daduke]

thanks for your suggestions!

• Check for bindings in the bearings of the motion system at some positions.

I have no reference how easy the head should move, but it seems ok and uniform

• Check if belts are touching something that they shouldn't, if they are riding the bearings/idlers/pulleys at the right position.

yeah I did that and found no problems. I've also redone a good part of the motion system.

Also: shouldn't increasing the stepper current make a difference this these two cases?

• Not that I think this is a reason, but how much cooling are you using? By the way, which material are you trying to print? You didn't mentioned on your post.

it happens for both PLA and ABS. I use little to no cooling for ABS and the usual lot for PLA.

thanks!

 

[daduke]

small update: started with a fresh Orcaslicer profile (was using Prusa before), same issue...

 

[Dave32]

Is the bed mechanically level?

 

[daduke]

Is the bed mechanically level?

yes, I have a clicky and did a mesh level. First layers are fine.

 

[Dave32]

yes, I have a clicky and did a mesh level. First layers are fine.

Clicky and mesh leveling are not mechanical leveling on a Zero. (At least, not unless you have a triple belted or triple lead screw Z conversion.)

If the bed is mechanically tilted, a mesh leveling will make sure your first layer sticks - but your Z travel will not be perpendicular to the plane of the bed - giving X/Y skew to the prints as they grow in height.

 

[daduke]

Clicky and mesh leveling are not mechanical leveling on a Zero. (At least, not unless you have a triple belted or triple lead screw Z conversion.)

If the bed is mechanically tilted, a mesh leveling will make sure your first layer sticks - but your Z travel will not be perpendicular to the plane of the bed - giving X/Y skew to the prints as they grow in height.

you are correct of course. What I meant to say: I used mesh leveling via clicky to adjust the bed as perpendicular as possible. As mentioned in the first post, I also get straight prints from time to time, which tells me it's not a "the whole printer is crooked" problem.

 

[daduke]

I might have a possible lead: maybe I went a bit overboard with the RGB lights. Since disabling a bunch of LED effects, prints seem to be straight again. To be confirmed...

 

[daduke]

after a few successful (straight) prints, I now had another crooked one, with the LED effects off. I really don't know what's causing this. If it were a mechanical problem, I'd expect it to occur in both directions, but it's either perfectly straight or leaning in positive X (and negative Y).

 

[Dave32]

hmmm..... What puzzles me is that it is a steadily growing error. A layer shift or step loss would generally be more sudden and pronounced.

 

[daduke]

I agree. I'm running out of ideas.

 

[daduke]

I'll be damned. Turns out it was a missing "step_pulse_duration: 0.000001" in the A/B stepper config because Mellow boards have a known issue. At least it forced me to tune every last aspect of the printer in the process.
Thanks for your suggestions.