Cannot get Z axis "perfectly horizontal" during XYZ calibration  

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

Report of testing the printer today:

After seating the left threaded rod yesterday, I have tested the printer today. Yet the bed is not perpendicular like the old times. Left motor mount is still a tiny bit lower than the right one. Left side of bed doesn't print even with -50um compensation. Out of options. I have also reseated the right threaded rod..
Edit: Have redone measurements of the bed and now it looks like that:

Surface to bed top:
Complete rear left 70.5mm
Complete rear right 70.5mm
Complete front left 71mm
Complete front right 71.5mm

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

Did you repeat XYZ calibration? You had to unscrew the frame etc. after all. I'd consider that 0.5-1mm deviation to be completely fine, which might also be caused by the felt pads.

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

I did rerun it. My X axis carriage though, never is perpendicular. Seems like it has left side a bit lower. I would have also expected for PINDA to recover from a mm-off calibration, but apparently not.. Left side of my bed is totally unloved, already ordered new PEI..

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

Don't think a new PEI surface would help. Does the initial problem when trying to hit the top of the Z axis still happen? Try repeating all the steps from the preflight manual just to be sure (you could even try resetting the whole firmware).

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

The Z axis issue didn't change, since it is attached to the frame. And visibly, left rod is at least 1mm shorter. I will do some measurements on them.

Edit: Smooth rods are equal. Left motor mount is 1mm higher.

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

Did you repeat XYZ calibration? You had to unscrew the frame etc. after all. I'd consider that 0.5-1mm deviation to be completely fine, which might also be caused by the felt pads.

again (i wrote it in a previous post):
💡 be aware that the xyz calibration routine only checks if x and y-axis are perpendicular. the routine can't check whether the z axis is perpendicular to the x/y plane or not.

➡ aligning the xy-plane to the z-axis.... or even the x-axis to the yz plane with manual, assembly measurements is important for two major reasons:

1. that your prints will rise correct in z direction
2. that you don't run out of the p.i.n.d.a. sensor range when running the meshbed levelling. if you do so, the printhead might crash into the bed during calibration... which isn't nice.

so "slightly" imperfections in the levelness of the x-axis carriage to the yz plane (due to unprecise holes inside the z-frame) are not severe, as long as you don't run out of the sensitivity-range of the probe during calibration.

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

I had the impression it would also store the z offset for the calibration points or is that exclusive to the pre-print routine? Out of ideas then. 🙁

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

I had the impression it would also store the z offset for the calibration points or is that exclusive to the pre-print routine? Out of ideas then. 🙁
from the information where the printer finds the calibration points (at which x and y values)... and where they should be located (from the architectural desing of the printer), it can detect if the printbed and the x-y movements of the printer are really perpendicular and if it is nescessary to correct an imperfection by additional computing.

and yes, furthermore it stores the z-offsets of the points "measured" during the calibration process !
but all these informations will only be taken in account (computed) during a printing process, not when you move the printhead through the settings menu or pronterface.... and not when you run a 9 point z-calibration.

and this information (especially the "height" of the calibration points) only allows you to correct a slight imperfection at the levelness of the bed (-> meshbed levelling, or better: meshbed level correction).
it won't avoid that you might print a leaning tower of pisa, when expecting a perfect, straight column, if your z-frame isn't perpendicular to the xy plane.

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

Did you repeat XYZ calibration? You had to unscrew the frame etc. after all. I'd consider that 0.5-1mm deviation to be completely fine, which might also be caused by the felt pads.

again (i wrote it in a previous post):
💡 be aware that the xyz calibration routine only checks if x and y-axis are perpendicular. the routine can't check whether the z axis is perpendicular to the x/y plane or not.

➡ aligning the xy-plane to the z-axis.... or even the x-axis to the yz plane with manual, assembly measurements is important for two major reasons:

1. that your prints will rise correct in z direction
2. that you don't run out of the p.i.n.d.a. sensor range when running the meshbed levelling. if you do so, the printhead might crash into the bed during calibration... which isn't nice.

so "slightly" imperfections in the levelness of the x-axis carriage to the yz plane (due to unprecise holes inside the z-frame) are not severe, as long as you don't run out of the sensitivity-range of the probe during calibration.

Thanks for the calibration insight. Yet I don't run out of PINDA range and my left side of the bed prints never get extruded. It's like 3-4CM from the left that is problematic. My dimensions are now 0.25mm different across all the positions of the bed and yet even with 50 micron compensation on the left it doesn't extrude..

Can it be a slicer error? I use Cura 2.5..

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

... and yet even with 50 micron compensation on the left it doesn't extrude...

Can it be a slicer error? I use Cura 2.5..

don't know anything about cura (never used it). but if you like, i can recompile either the 3.0.10 release or the actual beta (3.0.10-11) for you this evening, with a limit of +/- 150µm for the bed correction (but don't blame me, if there are some issues with it).

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

... and yet even with 50 micron compensation on the left it doesn't extrude...

Can it be a slicer error? I use Cura 2.5..

don't know anything about cura (never used it). but if you like, i can recompile either the 3.0.10 release or the actual beta (3.0.10-11) for you this evening, with a limit of +/- 150µm for the bed correction (but don't blame me, if there are some issues with it).

That would be cool, but first I will try using Slic3r. I heard some stories before about Cura/Simplify3D, but I hope these are just gossip.

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

My dimensions are now 0.25mm different across all the positions of the bed and yet even with 50 micron compensation on the left it doesn't extrude..

Might be just a wording issue, but what exactly happens? The filament just won't stick (and instead loop or get pulled or just curl around the nozzle? Or is there nothing flowing through the nozzle?

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

Left side is a complete squishy mess even with maximum 50 micron compensation. Mini purge strip doesn't even extrude from the start. Totally left line cannot extrude normally either.

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

Mini purge strip doesn't even extrude from the start.

Correct me if I'm wrong, but at least that part (assuming you mean the first strip printed on the front/bottom left after bed leveling) is completely normal and that's its purpose: The printer ensures the hotend is filled with filament so there is filament flowing when it actually starts with the actual model.

Thought of something else I didn't consider so far: Have you tried to see whether the smooth rods for the Y axis are on one level? And I'd also try to see how the alignment of the Y axis sled is (i.e. without the heatbed). Maybe one of the bearings isn't properly seated.

The whole thing still confuses me a bit, because I don't see how a skewed bed would screw with the Z axis movement.

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

Mini purge strip doesn't even extrude from the start.

Correct me if I'm wrong, but at least that part (assuming you mean the first strip printed on the front/bottom left after bed leveling) is completely normal and that's its purpose: The printer ensures the hotend is filled with filament so there is filament flowing when it actually starts with the actual model.

Thought of something else I didn't consider so far: Have you tried to see whether the smooth rods for the Y axis are on one level? And I'd also try to see how the alignment of the Y axis sled is (i.e. without the heatbed). Maybe one of the bearings isn't properly seated.

The whole thing still confuses me a bit, because I don't see how a skewed bed would screw with the Z axis movement.

Left side rod appears to be a bit lower indeed: 52.6mm vs 53.2mm. But it's lower, how can the side which is lower, receive ultrasquish mode? 🙂

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

Might be magic. Have you tried printing unicorn horns?

It might be a good idea trying to get the Y axis part perfectly fine first (i.e. check whether diagonals are exactly the same, verify corners are aligned, etc.), but your differences might be within tolerances. It's possible it's everything – i.e. the errors – adding up in an unfavorable way.

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

Might be magic. Have you tried printing unicorn horns?

It might be a good idea trying to get the Y axis part perfectly fine first (i.e. check whether diagonals are exactly the same, verify corners are aligned, etc.), but your differences might be within tolerances. It's possible it's everything – i.e. the errors – adding up in an unfavorable way.

For the Y/X axis, I am getting the best case result during the calibration. I don't even know what to do anymore.. I need the full build platform, not using one fourth of it sounds quite not okay. 🙁

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

That only tests the X/Y orientation, which seems to be fine. Don't worry too much, there's probably some tiny thing we're all missing and it certainly can be fixed. As such try to go through step by step, making sure individual parts are aligned the way they should. For example, once the bed is perfectly leveled, you can start looking at why the X axis is misaligned. Also make sure there aren't any zip ties blocking parts or acting as spacers and use a perfect and hard surface as your underground for now (think you mentioned dampening foam somewhere).

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

That only tests the X/Y orientation, which seems to be fine. Don't worry too much, there's probably some tiny thing we're all missing and it certainly can be fixed. As such try to go through step by step, making sure individual parts are aligned the way they should. For example, once the bed is perfectly leveled, you can start looking at why the X axis is misaligned. Also make sure there aren't any zip ties blocking parts or acting as spacers and use a perfect and hard surface as your underground for now (think you mentioned dampening foam somewhere).

X axis looks misaligned because left motor mount is lower than the right mount by 1mm, frame at the same time is level.

I will probably take some time off again, this is a big headache..

Re: Cannot get Z axis "perfectly horizontal" during XYZ calibration

But calibration should account for that 1mm difference (i.e. rattling on the top). Since you couldn't do that properly so far, you're off. The mount being one millimeter off shouldn't matter.

This is probably not going to work, but some random thing I'd give a shot:

When making the x axis level, there should be a 1mm discrepancy to the top on one side. Print a 1mm thick , 5mm wide and about 30-40mm long bar with 100% infill. Attach it with some double sided tape on top that side's Z axis sliding part to account for the difference. It should cause the slider to hit the screws of the top part a bit earlier. Thinking about the scenario and misalignment, this should adjust the issue.