UMO/UMO+ layers printing out of alignment, any tips?

[brictone 0]

Hi All,

 

I've had my printer for over 7yrs and have torn it down and replaced parts and re-built it more than once.  In face I have an UMO+ I purchased a couple of years ago and both have the same issue.

 

They both have the UM2 electronics.  So they are very close to the same.

 

I'm getting these odd layer sizes with features of a print.  I've done everything I can think of.  I tightened all 12 pulley set screws, re-seated the XY motors, tightened belts on one. changed filament. The problem exists on both printers so I don't think it's mechanical.

 

Here's what's happening exactly.  The print is mostly hollow which is typical.  If there is a layer that is more dense, meaning it has an internal support shelf or number of solid layers, the outside of the model with be slightly smaller.  See attached photos.  So it doesn't have a smooth finish.  in some situations it really looks like a set screw might be loose because the anomaly doesn't travel all the way around the model.  It's only on the left or right side because of a design that goes back on the Y.

 

It seems like the thing that helps the most is printing between 50-100% infill. (Not really a solution in this case.) Which makes me thing it's not mechanical again.

 

I'm happy to try anything, or if anyone would like to try my Gcode to see if it exists on their printer too.

 

If anyone has seen this before let me know.

 

Brictone

 

[gr5 2,170]

Mostly those are Z issues.  If the bed moves too far you get some underextrusion (nozzle is too high off the bed) and you get a line that is "inward".  If the bed doesn't move far enough (say a few layers later) then you get an overextruded layer that sticks out.

 

There a few possible causes but most likely you need a thorough cleaning of the Z screw.  If it's a plus the z screw I believe is permanently mounted in the Z motor - remove the Z motor and I think the screw slides out - unplug the z stepper first of course.  If it's a non-plus then there should be a Z coupling you have to loosen first and then I think the Z screw exits upwards.

 

Clean it will - put it on newspaper and spray with WD-40 and use q-tips or paper towels.  Remember it's a triple helix (like dna but triple instead of double) so make sure you get all 3 threads.  After wards add just one drop (size of a pea) of grease.

 

While the screw is out, Z bearings and Z rods should be checked to make sure there are no binding/friction spots when you slide the head up and down.  Try to add some twisting force to the bed (like it's heavy and bending towards to bottom of the printer) while you move it to see if it only binds when the bed is torquing.

 

If this doesn't help post back as there are more things you can do but if it used to work well then it's most likely dirt.  The second most likely thing to "fix" is to replace the Z nut with a more expensive (like 2 dollars or 2 euros lol) Z nut.  

[gr5 2,170]

There may be some loose belt issues combined with bad Z.  I sell very cheap belt tighteners for UMO in USA ($2) or in Europe go to aliexpress:

 

https://thegr5store.com/store/index.php/umo/belt-tensioning-springs.html

 

 

[brictone 0]

Very interesting thought on the Z axis. I'll give the Z rods some attention.

 

 

I just had a thought of increasing the tesselation of the model, right now it's very low poly, to see if it makes a difference. 

 

My thought is if the Z rod was dirty I would expect inconsistent results, right now it's perfectly reproduceable. Even if I lengthen the model on the Z. 

 

I used to have some belt tensioners, I used clothes pins springs. I can try that again.  Although if the belts were loose I would expect see ringing and vibrations in the print. Which I do not. 

 

Thanks for the Z rod suggestion I'll try that tonight. 

[brictone 0]

Well I went and pulled the Z rod out of the newer machine, I suspected it was worse since layers were inconsistent.  It always skips or gets stuck on a layer about 1cm high.  Regardless these consistent layer differences are still in the test piece.  So it did not improve the problem I was having. I also added tensioners to the Large Y Axis belts.

 

Just to clarify, I am getting different outside layers depending on whether or not the layer is solid or is hollow with 20% infill.  They just look different enough on X Y axis.  Solid layers are not as large in the X And Y axis.

 

I'm using the open source Marlin 1.1.8 on the newer printer and 1.1.0  on the older one.  Cura 3.2.1 Cura 15.4.6 was the same.  It's very noticeable on this print because it's white. Both have the same issue so far.

 

Attached are the STL and Gcode if anyone wants to give this a whirl and see if it's different.

MagicBox_7.obj

MagicBox_7.gcode

[brictone 0]

 

 

I have the cooling per layer set to 17seconds, with a minimum speed to 10mm/sec.  I'm testing 5mm/sec now.

 

Since yesterday I pulled the Z rod out of the older machine, cleaned it and re-greased it.  It's actually more stiff than it was.  Before the bed would fall, when the motor is disengaged, under it's own weight to the bottom now it barely falls if at all.  I also loosened and tightened the Y Axis pulley set screws.  I am thinking less and less this has to do with X Y and Z mechanical alignments or calibration.  But I'm happy to revisit everything and anything if anyone has suggestions.

 

Attached is another model that is quicker and easier to print if anyone wants to test this on their machine.  The areas where there is a solid layer, the outer walls come out different than the rest of the print.  Some of it is only after it finishes a feature in some places its around the entire model.

 

• I've been able to affect the outer walls by changing the amount of infill.  This changes the locations of solid infill.
• No change with tightening belts or pulley set screws or motors
• Scaling the model in X Y or Z, moves the anomalies to the same place on the model, where the solid layers and features are.  They are not random.
• Changed firmware to Marlin 1.1.8.  May have gotten worse. I don't have direct comparisons. 

Anyone have suggestions for Jerk, Acceleration, or anything else on an UMO+?  I'm going to try the official firmware here in just a bit and see if that does anything.