Random Z layer underextrusion (UMO with HB upgrade)

[veedee 4]

Hi fellow 3D print enthousiasts,

Normally when one of my printers starts to have issues, I'm skilled enough to fix it by myself. But lately I'm starting to have issues on two printers and can't seem to find the cause and fix it. I really hope someone can help me because those printers are very important in my business. I will explain everything in a listed way to create a better overview of the situation.

Let me first explain the current situation:

• I have four 3D printers. Three UMOs (which are later upgraded with the heated bed from Ultimaker), and one Lulzbot Taz5.

• One of the Ultimakers is 2,5 years old, the others are 1 year old.

• One of the 1 year old Ultimakers received a heated bed upgrade about 10 months ago, so when the machine was 2 months old.

• The other 1 year old and 2,5 year old Ultimakers received their heated bed upgrade about 1,5 months ago, these are the two printers with problems. The problems started after the heated bed upgrades, but I can't remember the exact time it happened for the first time. From now on I will call them the 'problem printers'.

• The first printer where I applied the heated bed upgrade to (10 months ago) doesn't have this issue.

• Both problem printers do have an 0.8mm E3D nozzle (not the hot-end, only the nozzle!), which always worked perfect. They are both printing the same product all day every day.

Let me now explain what seems to happen on both problem printers:

• Those printers are printing the same part all day long. On each build plate fits twelve of those identical parts.

• Most parts come out of the printer like they should, but sometimes (and it looks like it gradually gets worse) a part has one or a few layers of under-extrusion.

• Sometimes it looks like there is a layer of over-extrusion going in front of the under-extrusion layer, but I'm not completely sure about that.

• The under-extrusion layer is often very weak and can be pulled apart very easily, but sometimes there only seems to be a little inward and/or outward dent all around the part.

• As you can see in the pictures, the problem doesn't occur on the same height every time every time. Although it looks like the orange and pink part do have the issue on the same height, the orange part is the sixth out of twelve and the pink is fourth out of twelve parts. Both printed on different printers and running the identical gcode.

• I can run the identical gcode file again and there is a possibility that there is no issue at all, or the issue is on a complete different spot.

• It also has happened that the problem occurred on the same height on multiple parts in the same print run.

• I also found out that pushing down the build platform works, but on the lowest 4-6cm the build platform goes from quite smooth to an almost jam.

• Settings of the file which always worked: 55mm/s, 220Celsius, 0,15mm layer height. Material is colorfabb pla/pha.

Printed with the 1 year old  printer:

Printed with the 2,5 years old printer:

Because the problem is so extremely inconsistent, it is very hard to tell if a possible solution works. What I already tried to do to fix it is:

• Slower printing, down to 30% speed. No effect

• Higher temp, I tried up to 235C. The prints got very ugly due to this high temp, so I didn't test this for very long and thus don't know for sure if this had any effect.

• Cleaned the Z rods and bearings with a cloth and some solvent. Looked like this worked for a short while. But the problem re-occurred. This solved the jam on the bottom few centimeters of the build platform.

• Tried lubricating the rods and linear bearings again with. No effect.

• Tried lubricating the Z screw and rotating Z axis with magnalube. No effect

• Checked the PTFE couplings with calipers on both printers. No weird deformation measurable or visible. Both teflon couplers are a few months old. Exact operation time is unknown.

• Atomic pulls, no debris found and filament comes out nicely 'nozzle shaped'.

• Fresh and clean E3D 0.8mm nozzles. Tightened to the hot-end at 200C without any filament in it (after atomic pull). No effect.

Also a thing that is important to mention: the linear bearings seem to stick/lag or whatever it is called, when I move them up and down the rods slowly. I tried this on all four bearings and rods (both printers) by disassembling it from the printer. When I move the bearings up and down fast, this sticking effect doesn't seem to happen. This effect is also mentionable on the good working Ultimaker.

Another thing I want to mention is that the Leadscrew on the Z axis of both problem printers has some play when I push up the platform a bit. This play is about 0.5 - 0.8 mm-ish, I can't really measure it. The good working Ultimaker with the oldest heated bed upgrade has a lot less play, 0.1 mm at most.

Alright, I hope someone can help me with this problems, because I'm getting frustrated and out of possible options to fix this problems.

Thanks.

Edited January 29, 2016 by Guest

[neotko 1,417]

I got that z play on one of my umo+. Bough one z nut and problem was fixed. Also changed the bearings for more smooth movement.

https://ultimaker.com/en/community/11123-z-axis-layer-error?page=last#reply-130203

Use the ulticontroller to home, them move z slowly. If the bed bumps, you might have that problem.

Edited January 29, 2016 by Guest

[veedee 4]

@neotko

Seems like a possible solution. When did you order/receive your UMO+?

If really this is the problem, I may expect from Ultimaker to ship me a quality Z-nut instead of me buying a Z nut from a third party, right..? Those bed upgrades cost €300 so I expect quality parts.

[neotko 1,417]

That umo+ with the problem was bought 6 months ago or so. The seller did sent me a replacement also but I bought a extra one from pololu. Btw the pololu did fit better than the reseller replacement part.

[neotko 1,417]

Also the 4€ bearings I got from a local seller where much better than the original ones. But also my other 2 umo+ beds are perfect. Must be just a bad batch.

[veedee 4]

Same here, the oldest heated bed has almost no play at all.

I think the combination of having play in the lead screw AND the bearings getting a bit stuck sometimes creates the bed to jams for one layer, which explains the over-extrusion in that layer.

When it moves to the next layer, it forces through the sticky effect of the bearings and skips the previous layer, which explains the under-extrusion in the next layer.

This seems very plausible now that I think of it.

[veedee 4]

@neotko

Temporary solution: I 3D modelled this Leadscrew, printed in ABS and I'll see how this performs. It has no play no play for now, but will probably wear out fast.

[neotko 1,417]

Indeed

[neotko 1,417]

If it breaks, the speed of the fall it's quite scary, trust me...

You could use a spring connected to that printed part and the metal one, to work as a shock absorber, but the spring force should be quite big to impact on the sticky bearings.

Edited January 29, 2016 by Guest

[veedee 4]

I understand your concerns. The final design nut is 5mm wider, quadruple the height of the standard nut, and printed with 100% infill, it's not going to break.

Testing the performance right now. I will post updates asap.

If it works well, I will try to make it from igus tribo filament.

[Nicolinux 288]

Would you mind to share the model? I have Igus Tribo filament here