RE: Z-banding or Z-wobble, do we have a solution yet?
Yep. I did all those things, too. Without taking the bed brackets apart, it looks like it would be possible to open up the hole that the POM nut is held in, either by drilling or reaming, but best by modification of the design. If it can be held in place vertically, but allowed to move in the XY, then it would help reduce/eliminate z wobble. A properly designed isolation would be best, of course, but that could be an interim, or end user mod/fix.
RE: Z-banding or Z-wobble, do we have a solution yet?
I mocked up part of the right hand side of the heatbed (with essential dimensions being measured reasonably well). Then I've added a potential location for WobbleX parts. The green parts are 3d printed. The part in between the green parts is a machined metal assembly available from AliExpress. The upper green part would be remixed to extend up to the underside of the bed, to which it would be screwed using the original holes for the lead nut (not shown). The lead nut would go into the underside of the lower green part (fixings to be remixed).
There is a lot more space on the left side of the bed compared to the right. On the right, the big problem is the screw head that's part of the rotational joint - shown in red. The clearance between the lead screw and the joint screw is probably 1mm or less. It appears to me that if WobbleX was used, the freedom of movement would allow the lead screw to contact the joint screw. And this probably applies to any mod which allows some freedom of movement of the leadscrew in the XY plane relative to the bed subframe. It might be possible to recess that screw, but I don't have a machine shop or the desire to make irreversible modifications to the printer.
RE: Z-banding or Z-wobble, do we have a solution yet?
How much Z would we lose with that? Looks like maybe 10-20mm?
RE: Z-banding or Z-wobble, do we have a solution yet?
Probably something like that.
RE: Z-banding or Z-wobble, do we have a solution yet?
I've done some experimenting with (cheap) dial test indicators (DTI). Initial conclusions suggest that:
- DTI could be used to check for potential banding without doing a print. This would make it easier to adjust things and assess the difference, and also to asses how banding might vary at different bed temperatures.
- Thermal expansion likely contributes to lead screw system alignment issues/stresses and hence banding. Relieving the stress (by the adjustment procedure discussed above) is likely temperature-sensitive. Relieving the stress at room temperature might not help so much with PETG banding where the bed is at 85C.
- The amount of deflection caused by thermal expansion is small - maybe 0.15mm (more tests and test setups needed). But this is enough to perturb the lead screw system.
Assuming that the lead screws are not significantly bent, and given that the apparent unwanted deflection is small (eg, 0.15mm) it might still work to use WobbleX despite the very limited clearance. Also, it might work to have the resting WobbleX arrangement biased somewhat off-centre away from the sides of the printer as this would accommodate thermal expansion while restricting movement towards the problematic rotational joint screwhead. I plan to order a pair of WobbleX parts from AliExpress once things get back to normal after Chinese new year, and see how it works out.
An example DTI setup:
RE: Z-banding or Z-wobble, do we have a solution yet?
I've done some experimenting with (cheap) dial test indicators (DTI). Initial conclusions suggest that:
- DTI could be used to check for potential banding without doing a print. This would make it easier to adjust things and assess the difference, and also to asses how banding might vary at different bed temperatures.
- Thermal expansion likely contributes to lead screw system alignment issues/stresses and hence banding. Relieving the stress (by the adjustment procedure discussed above) is likely temperature-sensitive. Relieving the stress at room temperature might not help so much with PETG banding where the bed is at 85C.
- The amount of deflection caused by thermal expansion is small - maybe 0.15mm (more tests and test setups needed). But this is enough to perturb the lead screw system.
Assuming that the lead screws are not significantly bent, and given that the apparent unwanted deflection is small (eg, 0.15mm) it might still work to use WobbleX despite the very limited clearance. Also, it might work to have the resting WobbleX arrangement biased somewhat off-centre away from the sides of the printer as this would accommodate thermal expansion while restricting movement towards the problematic rotational joint screwhead. I plan to order a pair of WobbleX parts from AliExpress once things get back to normal after Chinese new year, and see how it works out.
An example DTI setup:
Keep us posted. I'm interested in this.
RE: Z-banding or Z-wobble, do we have a solution yet?
Interim update: I'm working on an alternative mechanism for dynamically accommodating small amounts of leadscrew misalignment (but likely not full-blown bent leadscrews). I have a viable standalone proof of concept, now I'm waiting on some parts to integrate it into the printer and test. If all goes to plan, less Z would be lost compared to WobbleX - potentially printable to the full 360, although any final lift beyond z=360 at end of print might be compromised. Hopefully printed parts (depends on adequate stiffness) with a few widely-available vitamins. Fingers crossed that integration goes well, and that it solves banding...
RE: Z-banding or Z-wobble, do we have a solution yet?
This is what I can do:
Extremely disappointing print. This was placed in the front of the printer bed, because it's almost impossible to print at the back (no good first layer).
RE: Z-banding or Z-wobble, do we have a solution yet?
When this problem cropped up on Rat Rigs V-Core printers we tried a number of options including Wobble rings of variouis designs. The most popular solutions are now the Oldham couplers or the VenterMechs.
RE: Z-banding or Z-wobble, do we have a solution yet?
I'm not seeing anything as sharp and random-looking as that. It does not look like classic leadscrew banding to me. If this was a multi-tool print (eg, other parts at the same time with different tools) I'd ask if you've lubricated the coupler pins. Otherwise I'd check if your nozzle is loose (actually wobbly) - the grub screw that holds the long steel tube inside the toolhead, and the V6 adapter/nozzle/heatblock connections if you're using a V6 adapter. Be careful not to overtighten the grub screw as this can pinch the steel tube. Also if you're using the V6 adapter, did you hot-tighten the nozzle?
This is what I can do:
Extremely disappointing print. This was placed in the front of the printer bed, because it's almost impossible to print at the back (no good first layer).
RE: Z-banding or Z-wobble, do we have a solution yet?
It would be great if we could use an off-the-shelf solution. However, the stock steppers have an integrated leadscrew, so there's no easy way to add an Oldham coupler. Also, there's very limited space to add any wobble fix mechanism without reducing the z height of the printer. I'm working on a solution as noted above.
When this problem cropped up on Rat Rigs V-Core printers we tried a number of options including Wobble rings of variouis designs. The most popular solutions are now the Oldham couplers or the VenterMechs.
RE: Z-banding or Z-wobble, do we have a solution yet?
The integrated lead screw would offer no barrier to installing VenterMechs. Z height will also not be as compromised (if at all with VenterMechs) but the diameter of the mechanism may be the issue. If you want more info on this option you can email me [email protected].,
RE: Z-banding or Z-wobble, do we have a solution yet?
I don't seem to have a banding so far.
That doesn't seem like lead screw banding. It's too random.
RE: Z-banding or Z-wobble, do we have a solution yet?
I agree with the others. That doesn't look anything like bent z rods. Is this on a single tool machine, or a tool changer? If the former, check for loose nozzle as above. If the latter, then check that tools are secure to the carrier properly. Greasing the three posts (that look like nuts) on the tools where the carrier mates up is recommended.
This is what I can do:
Extremely disappointing print. This was placed in the front of the printer bed, because it's almost impossible to print at the back (no good first layer).
RE: Z-banding or Z-wobble, do we have a solution yet?
It is a 2 tool head XL, but was printed with one color, single head. Stock 0.6 nozzle, it was assembled in the factory, I did not touch it. Since then I replaced it to 0.4 and it was sitting properly.
I put the print in the front, and I just noticed that the front is vibrating way more than the back. (Frame is just hanging in the air, no support). I'll try to reprint it (same place), and when my new heat bed arrives, will test it in the back as well.
RE: Z-banding or Z-wobble, do we have a solution yet?
I've had a look at the VenterMechs mechanism. It might possibly fit on the left side, but almost certainly not on the right side (the XL is not symmetrical - the left side of the bed is "fixed", the right side has a rotational joint).
The integrated lead screw would offer no barrier to installing VenterMechs. Z height will also not be as compromised (if at all with VenterMechs) but the diameter of the mechanism may be the issue. If you want more info on this option you can email me [email protected].,
RE: Z-banding or Z-wobble, do we have a solution yet?
It is a 2 tool head XL, but was printed with one color, single head. Stock 0.6 nozzle, it was assembled in the factory, I did not touch it. Since then I replaced it to 0.4 and it was sitting properly.
I put the print in the front, and I just noticed that the front is vibrating way more than the back. (Frame is just hanging in the air, no support). I'll try to reprint it (same place), and when my new heat bed arrives, will test it in the back as well.
Assembled... except not. You still had to install the tool docks and stuff, and there's a lot that can go wrong with that. If the docks don't feel so solid that they might as well be welded to the frame, then they need to be tightened some more and their positions recalibrated.
But yeah, your bed should not be visibly moving at all, so that's definitely a problem.
RE: Z-banding or Z-wobble, do we have a solution yet?
Two things I've done to reduce these weird artifacts:
1. Level the printer. Put a good level on the X-Y bar and level it left to right. Put the level on the bed to get forward and back as level as possible.
2. Check the belt tension. This is a super easy thing to do. Prusa provides a web based tool to do this that actually works. You'll need a quiet room to do it properly.
Question: Should belt tensioning be done with the printer cold or hot?
RE: Z-banding or Z-wobble, do we have a solution yet?
I have this z-wobble issues too on a fresh semi assembled XL 5 TH and after closer inspection it is resulting from a slightly bent left leadscrew (only in the first 15cm but that will hit the first 15cm of all my prints 🙁
@tg73 --> Thanks! followed your suggested instructions and got the issue a bit reduced - but it's not a perfect clean print quality on this new expensive machine after all and I'm thinking about my options here...
RE: Z-banding or Z-wobble, do we have a solution yet?
I got sidetracked by other issues (likely bad belts, replacements on their way). This is freshly assembled, I hope to fit it tomorrow for a first trial. It is designed to allow for small a small XY-plane offset between the bed and the trapezoid screw, while having negligible (in practical terms, effectively zero) rotational or vertical slop.
