Wave shift pattern on the Z axis


I have a Creality CR-10S Pro V2, I've been printing regularly for one month and then, suddenly, I started to notice waves on the Z axis. They're already noticeable with the test cube but deviation increases with the height and objects are ruined (eSUN Gray PETG 20 mm test cube, printed at VERY low speed):

PET test cube

It was my first attempt to print with PETG then I thought that could be the problem, I cleaned the nozzle (also using eSUN cleaning filament) and switched back to PLA+. The problem is even more evident (eSUN cold white PLA+):

PLA+ test cubes

Note how the "wave" pattern matches between the two test cubes. Even if I was printing from less than a month (and "just" 5/6 hours per day, not every day) I thought it might be time for some maintenance:

  • I fully cleaned the printer.
  • I checked all the belts.
  • I checked all the nuts & bolts. A few were loose, including those 4 that "attaches" the horizontal rail (the one that holds the extruder) to the Z axis threaded rods. Loose enough I could screw them with my bare hands (including those two at the left side which are hidden deep inside and they're a pain to reach with tools even removing a the cover).
  • I fully cleaned the rods and applied lubricant again.

I've got this:

second PLA test cube

I then moved the printer to the ground and tried again, it helped (possibly a lot) and printing PLA I get a cube almost like the PETG one in the first picture.

I noticed the both threaded rods have some freedom of movement in the Y direction (not the rod itself but the bearing ball they go through at the very top of the printer can move within its plastic enclosure); 0.5/1 mm for the left rod and 1.5/2 mm for the right rod. I don't know if it's an allowed tolerance or the sign that something is wrong (and what).

Searching on-line I read as many opinions as many posts I can find; I wouldn't go and buy replacement parts until I randomly solve the problem. It's frustrating enough to have to deal with it (and a huge amount of material wasted in failed prints and tests) after less than one month of activity...

Update: I secured the two rods using two 0.2 metallic feelers and nothing changed. I still have this shift within the first centimetre (it changes position but it does not go away).

Update 2: everything seems perfectly square (I can't swear this for the rods), cleaned and tight. Everything moves smoothly, and both bed and horizontal bar are fairly leveled. Effect is slightly less noticeable but obviously present.

One concerning point: travel speed was 150 mm/s, if I slow down to 130 mm/s then I can see a slightly less sloped shift; that makes me think that something could be vibrating when moving: top and bottom seem straight and they're printed at a much lower speed (20 mm/s vs 50 mm/s for PLA+, 20 mm/s - 15 mm/s for first layer - vs 30 mm/s for the test cube in PETG; travel speed was 150 mm/s for all of them). I guess that to reduce the printing speed to 10 mm/s for the entire object might help but I wouldn't call it a solution...

Adriano Repetti

Posted 2020-11-06T20:23:45.707

Reputation: 115

2No amount of lubrication or flow tuning will improve this. You have a serious mechanical problem, probably with the Z axis roller wheels being unevenly worn/tightened and/or the frame being severely non-square. – R.. GitHub STOP HELPING ICE – 2020-11-07T00:37:23.463

@R..GitHubSTOPHELPINGICE you may have a VERY good point here! Roller wheels on the external right side can be moved by hand (and they they won't move the horizontal support - nor they move when the support goes up) while the others are tighter. I'm not sure what to do with them, it's just a screw and tightening it doesn't move it nearer to the vertical bar they should run in – Adriano Repetti – 2020-11-07T01:14:09.290

On each side, one of the three wheels is mounted on an eccentric nut which can be turned with a wrench to move it closer to or farther away from the V slot. It needs to be tight enough that there's no play, and generally tighter is better except that it will wear more quickly when overtightened. – R.. GitHub STOP HELPING ICE – 2020-11-07T02:14:11.320

@R..GitHubSTOPHELPINGICE spot on!!! I can make it slightly better tightening that eccentric nut but not much, I removed the vertical frame once again and checked if it's in square and...no, absolutely not. The right side (the one on the side of the LCD) gains a couple of mm over 15 cm... Man, you were completely right! Do you know if it's something that I could fix? There are two small bolts to fix the vertical bar (x axis) to the horizontal one (x axis) but they're in-line and REALLY tight. – Adriano Repetti – 2020-11-07T12:58:52.557

3Classical layer shift. Your bed is slipping.That means that its belt is loose or much too stiff, or there's friction. – Trish – 2020-11-08T13:32:06.737

@Trish is right. Since you described it as "wavy" I failed to notice it never actually moves except in one direction. This is almost surely just ordinary layer shift. – R.. GitHub STOP HELPING ICE – 2020-11-08T14:56:46.913

@Trish belt seems tight but I am not sure when tight is too much, I can "pinch" it and it resonates. The friction problem could easily be the one, it seems harder to move than it was. Do you have any suggestion for that? If I untight its wheels they won't touch the rail. I guess that to add oil there isn't a great idea... – Adriano Repetti – 2020-11-08T19:03:54.750

1Awesome!!! That was it!!! It still not perfect but the defect is almost gone, just need to understand how to set it up! Thank you very very much! Would you mind to post it as answer? – Adriano Repetti – 2020-11-08T19:59:39.823



This is classic Layer Shift in the Y plane. This happens in one of three cases:

  • Belt too loose
  • Belt too tight
  • Bed movement hindered

The belt should ring and resonate when tapped, giving a nice tune. I haven't really tuned mines, but my ender3 (250 mm bed movement in total) has about a G major.

The bed movement should be, with motors off, easy and smooth. If there are spots where it hangs, you might need to rearrange cabling or check if there is a bad spot on the linear motion system. If the movement is hard, you might need to lessen the pressure. Since the CR10 is pretty much the same as the ender3, the eccentric nuts should be turned so the bed does not wobble or move side to side but not to a degree the friction slows the movement. Carefully tweak here, use increments of about 10° till the layer shift vanishes.

Be careful where you place your printer: cabling can snag on other items or things can brush onto the bed. That should not happen. On the Z-axis it can unplug the X-axis motors and extruder, or in worst case, yank at the heater cartridge and damage it.


Posted 2020-11-06T20:23:45.707

Reputation: 15 633