Strange layer inconsistency issue

Basic Information:

Printer Model: custom i3 style printer
MCU / Printerboard: SKR 1.4
klippy (2).zip (4.3 MB)

Describe your issue:

On larger parts and only on larger parts there are repeating layer inconsistency issues present. The more detail there is the stronger the inconsistencies are.
Everything is calibrated and I am into this hobby for almost 10 years:
printer frame is true, e-steps, flow, resonance comp, PA, etc.

Due to the inconsistency pattern I am 100% sure it is not ringing and pretty sure it is not hardware related in general.
I am usually very good at troubleshooting but here I am quite stuck.

Printed in ABS 255°C/100°C at very slow 60 mm/s (usually printing 100mm/s)


What was the orientation of the part on the bed please?
Was it right side down on the bed? Which direction was it facing and do you see any differences in X or Y sides?

Is this problem just a one off or is it haunting you for a while?

What has been the history running up to this problem and if more parts were affected, what were they in shape and orientation?

Thank you.

Sorry for wasting your time - I just found out that my Y-idler mount broke, which is why the belt was sagging, causing the issues you see in my pictures.

2 Likes

Not wasting any time at all, it is always a way to learn something new.
I am happy you found what it was and that you came back to tell us about it.

Usually I am the troubleshooting guy but with that one I was stuck until I closely watched the bed move…
Luckily it did not break completely so I managed to design and print a wonky new part that was good enough to print a fine replacement. That’s 3D printing for you…
But I guess 8 years for a printed part under constant load is not too bad.

1 Like

Basic Information:

Printer Model: custom i3 style printer
MCU / Printerboard: SKR 1.4
klippy (2).zip (4.3 MB)

Describe your issue:

I posted this a few days ago and closed it soon after because I thought I had found the underlying issue but sadly this is not the HW issue I thought. Again, I am pretty certain it is a setup issue, maybe to do with PA or resonance compensation.

This issue is only really visible on larger parts with even surfaces so I cant tell how long it has been present. I already tried stepping down acceleration and speeds a lot but nothing changed.
There is also some underextrusion going on I only experience on large parts like this.
I am a calibration/tuning fiend so you can be certain that everything from e-steps, over PA, flow rate, etc. is calibrated as much as possible.

Printed in ABS 255°C/100°C at very slow 60 mm/s (usually printing 100mm/s).
0.3 nozzle, same effect at 0.08 and 0.12 layer heights. Using a Voron M4 extruder with a relatively short bowden tube.
These steps you can clearly see are running parallel to the bed and are equally strong on the faces that run along the X and Y axis.




Not a slicing issue either - the preview shows even layers/surfaces

No, it is not closed.

Imho this print shows that it is not z-banding because I dont know a reason why the leadscrews would only wobble (in a periodic pattern at that) on complex surfaces…

Personally, I’d think it is a mechanical issue, but you can simply try without input shaper and without PA and see if it changes anything.

Am I correct in thinking that each of these ridges are multiple layers?
If so I fail to see how a setup issue could cause this. That would point to a mechanical issue for me and would be the first instance I would check.
But do what Sineos suggested, do the same print with PA and RP switched off and see.
If you still have banding, even if it is different, it suggests a mechanical issue.
If the banding has gone it was a setup issue.
If the banding is still there but utterly different, maybe it is a combination of the two.

I would do this check before touching anything else and see. While checking always change only one parameter to eliminate that from the equation.

This print is 8 hours in, input shaper disabled and speed back up at 100 mm/s because speed does not seem to affect it.
As you see the issue it getting less and less pronounced the smaller the footprint of the model gets which is in line with my earlier observation.

What I find very interesting is that each ridge consists of multiple layers and I looks like we are extruding less and less filament with each layer until at a certain point it starts back at zero. On top of the unexplainable pile is that it seems the issue does not only get less pronounced when the model gets smaller, it also looks like the amount of layers each ridge consists of varies quite a lot.

I just cant think of a HW issue that might cause this behavior. A bent leadscrew should look different and I am using anti kickback nuts to prevent sagging.

Will report back with PA turned off for the next part.

OK, with PA off the issue seems to be less pronounced but I am getting partial underextrusion.
After disassembling the extruder I can confirm that there is no filament grind present and I also rule out a partial clog since there is no underextrusion at all for smaller parts or parts with easier geometry.

I am at a complete loss and would really appreciate any idea!!

In the meantime I tried switching to PLA to rule out warping issues which did not help, I changed the nozzle from 0.3 to a shiny new 0.4 nozzle, I retightend the extruder belt and nothing made a difference. I also did some flow and PA calibration prints which came out flawlessly perfect.

I just cant wrap my head around this: how can it be possible that everything is calibrated, tight and spot on, smaller prints work out nicely (with small I dont mean benchy small but 5x5x5 cm) and only large prints show this issue…

Do you guys maybe know another place where I might find help?

Does it look to be only in one direction or is this photo not showing it correctly?
It looks on this photo that where your red lines are painted in it is much worse then in the direction that is perpendicular. Or is that an optical illusion?

The other obvious place is the machine manufacturer of course. The other thing is the slicer, which slicer do you use?

Have you measured your filament? My brother recently had filament that was so far out of tolerance in its diameter it caused blockage in the filament path.
You mentioned that you had under extrusion and maybe these ridges are over extrusion.

Please tell us what filament you used and measure the diameter along several lengths and directions?

First off, thank you very much for looking into my issue, I am completely stuck atm!
The ridge effect is showing on the x and y faces of the part equally strong and the lines are only running horizontal / parallel to the bed.
The machine was initially a Wanhao Duplicator i3 but I since replaced every part except for the frame (motors, pullys, belts, everything) and I am using Cura as the slicer.

I checked and tried several different filaments (esun ABS+, Overture ASA, esun PLA+, extrudr ABS) which all show the same effect.
I am ruling out filament issues because I have printed enough large parts which all show exactly the same pattern.

OK it looks like I am not stupid and it was not a HW issue after all.
A awesome dude on reddit analyzed the gcode and found that extrusion is completely off for every layer that contains support interface.
Sliced one of the failing models in S3D and and it seems to work.

A stupid bug in cura cost me countless hours of troubleshooting and half a kilo of filament -.-

2 Likes

Thanks for getting back about this. I would never have been able to help since I do not use cura. Great you found what it was though, maybe a good idea to tell the world in a new post what that bug was so people who use cura can look out for it.

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.