Slight Z Banding / Ribbing

I've been chasing a z wobble / ribbing / banding issue on my printer the past few weeks I'm hoping to get some advice on as I can't seem to find a solution on the forums or other places online.

Here's an example of the issue on a cylinder (0.2mm layer, 45mm/s print, 205C extruder, 55C bed) [imgur.com]

The banding issue repeats every 2mm regardless of layer height which makes me suspect something is off in my z-axis but I can't seem to make the problem better (or worse) with anything I try. The ribbing is causing layers to fluctuate in size evenly around the print.

My printer is a custom built coreXY with

• Duet 3 electronics
• Cantilever z-axis with with igus TR10x2 leadscrew and matching nut
• Leadscrew is supported via thrust bearing & shaft collar. Attached to the motor with helical coupling as the stage is decently heavy
• Print surface is a flat aluminum plate with Buildtak removable print surface on top
• Keenovo silicon heater
• Hemera extruder
• E3D volcano with 0.6mm nozzleX

Usually I print PLA and HTPLA at 60mm/s with 205C extruder and 55C bed

So far I've tried the following without really being able to make it better or worse

• Going up and down in temp (no change at 185C or 225C print temp)
• Changing bed heater from bang-bang to PID
• Printing without heated bed
• Printing with & without mesh bed leveling enabled
• Re-aligning z-axis rods & leadscrew
• Printing with & without z-hop
• Printing with & without part cooling fan
• Printing with & without infill (also trying single walled and vase mode)
• Changing extrusion multiplier (have gone down to 75% Ribbing is still present but less pronounced)
• Changing layer heights (less apparent in 0.4mm prints but still there)
• Trying other filaments (happens with HTPLA and Hatchbox PLA)
• Slicing in other slicer (I usually print with S3D but tried Slicer PE with no change)
• Changing object size / shape / printing in other spots on bed / printing multiple
• Loosening a bearing mount on my z-axis to try to eliminate any potential binding
• Changing motor idle current (normally 30%, tried 60% and 90%)
• Changing motor drive currents

Also have double checked the following so far:

• heated bed PSU voltage (steady 20.86V)
• Duet 3 PSU voltage (onboard voltage probe is between 23.8V and 24V)
• Retuning bed PID (barely moves from 80C setpoint, 55C noted above is measured with IR thermometer on bed surface)
• Retuning nozzle PID (steady to within 1deg C)
• Checking for backlash in z-axis (not measuring any with calipers)

edit: added a few more things I forgot

Any ideas on where to go next?

Edited 2 time(s). Last edit at 04/15/2020 04:10PM by Marty56.

Since no one else responded:-
Given the height variation is repeatable and I believe you have it on different models. I also suspect a mechanical issue affecting Z height.
A picture of the lead screw system would be helpful.
I can't make the numbers add up to something relevant to the lead screw but if you only have one im wondering if its due to the loading being transferred between the threads. Not the pitch itself but which of the threads is on the outside or inside. Im taking this as a 2 start leadscrew with each "thread" having a pitch of 10mm.

A shot of the top of the z-axis. The mounts on the bottom are a mirror of the top: [imgur.com]

Interesting theory on the leadscrew. Mine is only a single start with 2mm pitch. I do have a regular & anti backlash version of the nut (anti-backlash pictured above) so may I'll try back to the regular version.

OK.
Im thinking this is an issue of binding/over constraining. Since you have a spiral coupler the resistance may cause the spiral to wind up/down instead of the platform moving the complete distance.
Is the thrust bearing just resting on the mounting?
Whats holding the motor in place? In theory you only need to stop it rotating not constrain its position.
I would make one guide rod the position control so constrain it top and bottom and the other in a slot left/right in picture so it can always be the same distance from the other.
Wondering if the lead screw is not parallel to the guide rods. Since it looks very constrained. If it makes the thrust bearing (or the housing) not sit perpendicular then it may alter the z height as it rotates.
You could try slackening the fixings holding the nut to the bed carriage so the nut can float to an unconstrained position. It will be tensioned in Z by gravity. However if the fixings are just loose they may vibrate out so you may need to Loctite them in. For a test print you could just keep an eye on them.
Long day.
Hope this helps.

Hi MCCarman, thanks for the suggestions!

Quote
MCcarman
OK.
Im thinking this is an issue of binding/over constraining. Since you have a spiral coupler the resistance may cause the spiral to wind up/down instead of the platform moving the complete distance.

I didn't notice anything with the spiral coupler the first time around, but am noticing the slits spacing changing slightly as it rotates. This might be from me realigning the leadscrew when I changed the out flanged nut for the non-anti backflash version. This changed the appearance of the ribbing I'm getting but didn't really reduce it. t's hard to find a 5mmx x 10mm spiral coupler outside of adafruit, and they don't have much info on the torsional stiffness. Ruland (and I'm sure others) have couplers that appear to be stiff enough for my z-axis so that might be worth a try

Quote
MCcarman
Is the thrust bearing just resting on the mounting?

Yes, it's only secured by gravity, and there is ~2mm of float per side around the lead screw to allow for alignment without scraping the sides of the mount.

Quote
MCcarman
Whats holding the motor in place? In theory you only need to stop it rotating not constrain its position.

Motor is secured by two screws, but there's not much adjustment here, as I planned on adjusting the leadscrew to be relatively aligned with the motor, and then having the coupling handle the rest of the misalignment. I'll make a note to modify this with some more adjustment if I don't make any other progress

Quote
MCcarman
I would make one guide rod the position control so constrain it top and bottom and the other in a slot left/right in picture so it can always be the same distance from the other.
Wondering if the lead screw is not parallel to the guide rods. Since it looks very constrained. If it makes the thrust bearing (or the housing) not sit perpendicular then it may alter the z height as it rotates.
You could try slackening the fixings holding the nut to the bed carriage so the nut can float to an unconstrained position. It will be tensioned in Z by gravity. However if the fixings are just loose they may vibrate out so you may need to Loctite them in. For a test print you could just keep an eye on them.
Long day.
Hope this helps.

Interesting point about the thrust bearing. In a previous incarnation, my motor/ thrust bearing mount used to just be a 1/8" aluminum plate [imgur.com] that would flex considerably when changing directions in Z. The newer design I shared above has largely eliminated that flex, but might be preventing things from self aligning.I'm not totally sure if I had the ribbing issue back then as my bowden setup gave me more headaches than it was worth.

I'm fairly confident in how well the rods are lined up, I can run the stage around 15mm/s up & down without maxing out the super whopper motor. With the leadscrew disconnected the stage falls all the way to the bottom. My drive nut isn't hex shaped, but has a 6 bolt pattern for mounting, so I'm not sure I can get it to float in the current configuration but will take another look at this.

Tl;dr think my plan is

1. Add more adjustment to motor/ thrust bearing mount
2. Find correctly rated coupling
3. re-design nut mount slightly to allow it to float during movement

So I think I solved my ribbing issue... Kind of following my list.

1. Redesigned the motor mount to be able to adjust the alignment of the the motor and leadscrew
2. Took another look at the part that mounts the leadscrew nut to my bed.
   
   Did some kinda crude analysis showing that worst case I could get ~0.4mm of deflection in the part. Over the length of the leadscrew this would more axial misalignment than any rutland coupling was spec'ed for. (Using rutland for comparison since I don't have any info on mine. After realizing this I redesigned the part to be stiffer, switched it from PLA to PETG and reprinted. Got it re-assembled this afternoon and after some still wavy test prints I think I got everything re-aligned.
   
   [imgur.com]
   
   This part is 0.4mm layer heights so it's not as easy to compare to my other images but it definitely looks better than some of my other 0.4mm prints. Probably going to re-design & re-print parts one more time since assembly was a bit of a pain but feeling good about this solving my issue.

If im not seeing your linked photo correctly please forgive and ignore below.

It appears your Z motor is mounted at the top of the machine with leadscrew below? Is the lead screw supported at the bottom, or is it "hanging" on that leadscrew coupler? If it is, consider that your bed is effectively hanging on a spring.

---

Flex3Drive.com