Top Layers not touching - UM2

[mevander 0]

And here's my test cube (ran the gcode directly at 220C):

This was probably one of the worst prints so far. I also tested a 240 C, 0.2 mm layer and 20 mm/s print with a smaller test cube and it didn't work either. I'm going to take a good look at the extruder motor shaft and see if it's been sliding or something else today.

 

[opit78 0]

Thanks gr5.

I've pretty much the same result as yours once I set filament diameter to 2.55mm instead of 2.85mm ?!

What's going on ? It's possible that my extruder step/mm is that much off ?

There is nothing wrong with Ultimaker filament, correct 2.85mm, nozzle is not blocked and in perfect condition, white nylon insulator as new, extrusion test goes perfect all the way up to 10mm3/s (with filament diameter set to 2.85mm).

 

[opit78 0]

Last night's print with with filament diameter set to 2.55mm

It went pretty well.

 

[mevander 0]

I just spent a few hours with my colleague going over the machine. We tested the extruder to check that it was firmly seated (it wasn't perfect so we adjusted it), check the stop screw for slippage and checked the position of the filament vs the knurled axis. We also tested removing a few meters of filament from the spool, straightened it and ran a test with it like that to check that the increased curvature at the end of the spool didn't increase the friction in the bowden tube too much. We also did another atomic clean that looked very good. We double checked the filament diameter and it's a perfect 2.85. We also were advised to see if we could try to increase the current to the motor but apparently it's capped at 1300 mA so we couldn't test that.

The tests we did still resulted in the same kind of structures we've had before. The only thing that seemed to make a difference was if we manually pushed the filament pretty hard into the bowden tube while feeding it (to help the motor). I'm guessing that this would be the equivalent of increasing the flow or lowering the diameter.

I'm guessing that this means that there needs to be more material flowing but that it's a calibration/setting somewhere that is wrong?

Here's the last test we did with the ultimaker filament after all the cleaning and alignments:

 

[opit78 0]

I just took our printer head apart and I noticed that nozzle hole is off centre, it's off a lot !

Literally is drilled at the angle.

Look at the filament ! That's ridiculous.

I'am pretty sure that my issue is here.

 

[mevander 0]

Hmm, I have a similar shape on my filaments after pulling them but didn't think much of it

 

[opit78 0]

Filament is hitting the edge of the hole instead of going straight into the hole.

Our UM2 is only one month old so we've opened a ticket with Ultimaker technical support, hope they gonna help us.

That's manufacturer fault rather than user error or lack of experience.

BTW check the white nylon coupler as well.

Our is well wear off on the inside probably because extra friction created by faulty nozzle.

 

[mevander 0]

We checked our hotend as well but as far as I can tell, it looks pretty nice. We also swapped to a new glass plate to rule out a non-flat surface and did a factory reset and realignment of the bed. Still same result :(

 

[Nicolinux 288]

Nice find opit78. The filament shape looks very familiar to me. This could also explain some weird underextrusion problems that seem to be affect only some users while others are fine.

Maybe this is the explanation of the great mystery regarding underextrusion on the UM2!

 

[gr5 2,189]

@mevander

1) Pushing on the filament should only help the extruder not slip/skip. It shouldn't change the flow by more than 1% which is not visible. Better to just raise the temperature to 240C as this reduces viscosity by a huge amount. But just as a test.

2) I recommend you calibrate Z and E axes.

2a) To calibrate Z move the Z axis by 10mm or even 100mm and see if it really moved that far with a ruler. Maybe lay a flat pencil or ruler on the bed hanging over the edge and tape a ruler vertically to the corner and watch where the flat lines up with the ruler as you move it up and down.

2b) To calibrate E axis tell the UM to retract 10mm of filament, mark the filament with a marker. Retract another 100mm and make sure it really moves 100mm.

To have control like this on the axes you should connect a computer through USB to the UM2 and issue move commands. I like to use pronterface here:

http://koti.kapsi.fi/~kliment/printrun/

 

but you can use Cura in pronterface mode (I've never used cura in pronterface mode so don't know exactly how to turn that mode on) which simulates a few (but enough) of the pronterface features. Do Cura print window.

 

[gr5 2,189]

) Pushing on the filament should only help the extruder not slip/skip. It shouldn't change the flow by more than 1% which is not visible

 

Of course if you feeder is slipping or skipping then pushing can make a huge difference.

 

[kreativgeek 0]

Just to update my previous reply ... I also have the issue with other materials like Faberdashery:

I will check my hotend as well!

Just checked it, the hole is nice and center, so I don't have the same issue:

 

[opit78 0]

Your hotend looks pretty good for me.

Did you tried to print it little bit hotter yet?

I haven't got any reply from Ultimaker just yet but in the meantime I'am printing with 235C-240C and material flow set to 124%, filament diameter is back to 2.85mm for now.

 

[mevander 0]

Thanks gr5, I'll try to run some more tests today. I did a long run over the weekend at 230 C, 0.2 mm/s, 40 mm/s and 120% flow. The item looked better than we've seen before but still a lot of gaps between the lines.

I was thinking, is there an idea to try to measure/calibrate the nozzle size as well? If this would be a bit on the small side it would also push things in the wrong direction.

 

[mevander 0]

Right! I managed to get pronterface running and tested the z-motor and the extrusion. The z-axis seemed to perform well enough but for the two tests I did with the filament retraction I ended up getting 97 mm instead of 100 mm. It's a pretty small error though so I'm not sure it's responsible for me problems.

 

[gr5 2,189]

3% error is tiny. Not noticable in the final part.

Did you do this test? I forget:

http://umforum.ultimaker.com/index.php?/topic/4586-can-your-um2-printer-achieve-10mm3s-test-it-here/

 

[gr5 2,189]

I did a long run over the weekend at 230 C, 0.2 mm/s, 40 mm/s and 120% flow.

 

How about a short test instead at:

240C .2mm 30mm/sec 100% flow

 

[mevander 0]

3% error is tiny. Not noticable in the final part.

Did you do this test? I forget:

http://umforum.ultimaker.com/index.php?/topic/4586-can-your-um2-printer-achieve-10mm3s-test-it-here/

 

Hullo.

Yup, I get nice 10 mm^3/s cylinders at 230C and above.

 

[mevander 0]

How about a short test instead at:

240C .2mm 30mm/sec 100% flow

 

I did test 240C, 0.2 mm, 20 mm/s and 100% flow here: http://umforum.ultimaker.com/index.php?/topic/5375-top-layers-not-touching-um2/?p=71121

Today I did a had the same issue with a 0.1 mm layer run at 40 mm/s as well. I have a feeling that the problem is getting worse ...

 

[gr5 2,189]

@mevander - sorry - it gets confusing when 3 people have posted in the last 3 days with similar but quite different issues and I'm too lazy to go back and read all your posts again.

I feel like we ruled out underextrusion because you can print the cylinder just fine. That leaves Z axis. You said it was fine but - I think that on the top layer of your cube the Z must be moving too far. Really nothing else makes sense. The top is underextruded. The gcode is fine. You can print the cylinder fine so and your extruder is accurate (within 3%) so the onloy thing left is that the Z axis moved too far on the last few layers.

This is a common issue on UMO but not common at all on UM2. I guess I'm kind of stumped. Is your Z screw twisted or bent? How did you do the Z movement test? I guess I would do the following:

Get control of your Z axis again with bed and head at room temperature. Move Z axis off the surface of the glass by same amount as your part - or a little less. Then slip a stack of paper under the nozzle. Measure the thickness of 10 sheets of paper with a micrometer and divide by 10. Then move the Z axis by this amount many times and slide one sheet of paper under the nozzle each time. Feel the friction of the paper as it slides in each time. Are the Z movements consistent? Or are they variable? Repeat over a few mm (at least 20 sheets of paper. 30 sheets might be better).

I'm expecting the Z to occasionally move 2X as far as intended (and other times maybe 1/2 as far as intended).

If this test passes then I want to go back to underextrusion and see a closeup photo of your test cylinder that you say printed fine.

 

[gr5 2,189]

@mevander - can you send me the gcode for that box you printed that had underextrusion on the top layer?

 

[mevander 0]

@mevander - sorry - it gets confusing when 3 people have posted in the last 3 days with similar but quite different issues and I'm too lazy to go back and read all your posts again.

I feel like we ruled out underextrusion because you can print the cylinder just fine. That leaves Z axis. You said it was fine but - I think that on the top layer of your cube the Z must be moving too far. Really nothing else makes sense. The top is underextruded. The gcode is fine. You can print the cylinder fine so and your extruder is accurate (within 3%) so the onloy thing left is that the Z axis moved too far on the last few layers.

This is a common issue on UMO but not common at all on UM2. I guess I'm kind of stumped. Is your Z screw twisted or bent? How did you do the Z movement test? I guess I would do the following:

Get control of your Z axis again with bed and head at room temperature. Move Z axis off the surface of the glass by same amount as your part - or a little less. Then slip a stack of paper under the nozzle. Measure the thickness of 10 sheets of paper with a micrometer and divide by 10. Then move the Z axis by this amount many times and slide one sheet of paper under the nozzle each time. Feel the friction of the paper as it slides in each time. Are the Z movements consistent? Or are they variable? Repeat over a few mm (at least 20 sheets of paper. 30 sheets might be better).

I'm expecting the Z to occasionally move 2X as far as intended (and other times maybe 1/2 as far as intended).

If this test passes then I want to go back to underextrusion and see a closeup photo of your test cylinder that you say printed fine.

 

Hey, we're extremely glad that you're here and helping us at all. You're allowed to be both confused and lazy

I can check the z-axis again. We normally see that the layers start separating from the 4th layer or so, so it's not only the top surface that's wrong. This leads to the shell not being attached to the infill and the parts are not very rigid. By the way, I have what is called an indicator on wikipedia, a high resolution single-axis profilometer with a dial that's normally used for testing and calibrating milling machines. I can see if I can attach it to the print head somehow and then try to step it and see how it moves. It'll be fairly small steps but I think it has a range up to a few mm.

I've attached a photo of the latest extrusion test we did about 2 weeks ago at 250C.

 

[illuminarti 18]

I just took our printer head apart and I noticed that nozzle hole is off centre, it's off a lot !

Literally is drilled at the angle.

 

Look at the filament ! That's ridiculous.

I'am pretty sure that my issue is here.

 

Actually, I'm not convinced that your nozzle hole is drilled off center - at least not from your photo. I just looked at a 100 nozzles, and most of them basically look like that. But it's an optical illusion caused by the threading on the column. The start of the thread cuts away part of the top surface, making it look like the hole is closer to one side of the top in most lights, because the threading reflects light differently from the top surface. If you actually screw the isolator nut down around it, and look at the alignment of the hole compared to the ring of the nut, it's actually ok.

 

[gr5 2,189]

@mevander - that's one beautiful cylinder. I don't think I've yet printed one quite that good.

First of all when you say it starts separating around 4mm, well by default the fan comes on slowly and by 5mm it is at 100% fan. The PID controller is supposed to keep the nozzle at the same temp the whole time but stil...

Did you try printing the cube at half speed and at 240C? Maybe you should check your nozzle temp using this simple technique in this video:

 

 

[gr5 2,189]

@mevander my next thought is that your Z screw is messed up. I mean that has been a coninuing thought. That cylinder looks damn good though - I would expect a repeating vertical patern of horizontal stripes where the layers are too close followed by too far.

Your plan of using an indicator is excellent. It doesn't have to be attached to the nozzle. Any part of the frame that can then touch the bed is fine. You want to do similar movements to your test part so if your part was .2mm layers than .2mm movements in Z. Repeat as long as possible.

Maybe you should just post the gcode for a box that you've printed with bad top surface. Maybe it's the gcode.

This is kind of stupid - but how thick is your top surface again? Maybe you didn't do enough? Did I already ask this? You want about 4-6 layers for a really good top surface so top/bottom thickness=.6 for .1mm layers or 1mm for .2mm layers is plenty. Each layer should get better than the last.