Almost always missing layers / underextruding

[Nicolinux 288]

Guys, as soon as I have tightened the metal ring that sits beneath the teflon piece (at the bottom of the head), I have 0 play with the bowden tube. The metal ring has holes that one can use to insert something in there and turn it to the right. I didn't over tighten it, just went so far as it got and didn't force it. I have yet to print something - I'll post as soon as it is done.

 

[braddock 1]

Well, the movement of the bowden tube is a drammatical issue with um2. Can we print a red clip higher so the gray holder pressure is more firm?

 

I don't really like the whole system in general, seems prone to failure. Anyone tried the Bowden clamp mod above?

[Nicolinux 288]

Ok status update. Tightening down the metal ring moved the nozzle up but changed nothing. Still underextruding occasionally. However this time I caught it on camera!

The model looks like this:

For the videos, make sure to watch it in HD

Look to the right:

 

Look to the left (a bit fuzzy, sorry):

 

 

[chrisp 10]

Nicolinux,

Remember I posted here earlier that I had more than one issue causing my under extrusion. The main reason that I believe was the teflon fitting(pics are in an earlier post). It was deformed noticeably where it sits into the milled fitting. I had to use a 3mm drill bit to carefully open up a 2.5mm hole where it was deformed, to a 3mm hole, like it is supposed to be. That was the main cause of my clogs during retractions, which also caused my under extrusion. I started really seeing a difference in my prints once the milled fitting was adjusted along with that. Or maybe I'm just getting lucky again for the next few prints.

5 Hours

 

8 Hours

 

 

[Nicolinux 288]

The thing is - I have never experienced a clog. Neither on the UM1 (which I already sold) nor on the UM2. And this partial underextrusion for just a layer can't be due to a clog right? At least I hope so.

 

[chrisp 10]

Are you using the stock spool holder that mounts on the back of the UM2?

You might want to try out another holder that is free of tension/friction. I recently made one, or there are plenty of other designs. My filament was getting caught on the black netting just before feeding into the extruder.

https://www.youmagine.com/designs/spool-adapter-multi-step

 

[Nicolinux 288]

I tried both. The stock sppol holder (just now) and the spool on a lazy susan. No difference. Still occasionally underextrusion. I printed the same model with the same settings and filament.

 

[chrisp 10]

I'm trying to run through everything I experienced over the past month to help you.

Are your under extrusion lines always in the same general location on the Z axis or are they at random?

Also, from the video you showed above, your nozzle looks like its sitting much higher than mine in reference to the fan bracket at the bottom. What is the gap between the teflon fitting and the milled fitting? The gap looks like it could be 2mm or more from the pic.

 

[illuminarti 18]

Nicolinux, It might be worth taking out your teflon piece, if you haven't, and make sure that it's clear all the way through, and a consistent 3mm diameter. Chrisp's piece had a noticeable lip that wold have made it fit tighter on the filament - that might grab the filament momentarily - esp during retractions, causing brief under-extrusions until the filament frees up and slides forward again?

 

[Nicolinux 288]

Are your under extrusion lines always in the same general location on the Z axis or are they at random?

 

They are in random locations.

 

Also, from the video you showed above, your nozzle looks like its sitting much higher than mine in reference to the fan bracket at the bottom. What is the gap between the teflon fitting and the milled fitting?

 

I turned the metal ring beneath the teflon ring as far as it went. That was just a test to see if it would change anything. Right now the gap is 2.1mm. Before that it was 1.2mm I think.

 

[Nicolinux 288]

Nicolinux, It might be worth taking out your teflon piece, if you haven't, and make sure that it's clear all the way through, and a consistent 3mm diameter. Chrisp's piece had a noticeable lip that wold have made it fit tighter on the filament - that might grab the filament momentarily - esp during retractions, causing brief under-extrusions until the filament frees up and slides forward again?

 

I hoped to get away without taking the head apart. But oh well

 

[chrisp 10]

Just take pics before and mark everything. I've had the hot-end apart about 6-7 times now. The smallest adjustment means the world.

 

[illuminarti 18]

I will also say that while lately I've gotten some stunning prints without any signs of this brief under-extrusion... I have seen it happen before. One of my early attempts at the 'leaf lamp' print (the artichoke looking thing with lots of sharp-edged leaves) basically fell apart when it was done because of all of the under-extruded lines. I think, however, that many (all, I don't know) of the under-extrusion lines I had like this were due to extruder skipping steps - I was getting a noticeable click.

I don't remember what was different about that print, other than it was faster. I think I used different filament from the Faberdashery that I finally used, but I'd used that same filament for other flawless prints later. So I don't think the filament was a major factor by itself.

But all of my good results have come with 0.1mm layers, and speeds below 50mm/s. Which is not to say that I couldn't go faster/thicker... merely that the results that I know were good were in that relatively slow volumetric range (2mm³/s) - and my very best prints were at 30 or 35mm/s (1.4mm³/s).

Looking at Nicolinux's videos, it looks like all the under-extrusions are after retractions; so I wonder if it's just that fast retractions are sometimes skipping steps on the de-retract phase, so the head isn't fully primed. That could be because of extra tension in the teflon, as discussed above. Or for some other reason.

 

[illuminarti 18]

I guess the point was really... is it an issue of detractions being more likely to under-perform when printing at higher speeds, so the filament is being advance quickly into a higher pressure environment?

 

[illuminarti 18]

Or even... can the retraction overshoot, and be more than intended, so it isn't fully recovered.... because there's so much pressure already on the filament pushing it back out? Something similar to what I saw with the UM1 extruder, where the filament could slip back past the knurled bolt at high speeds/pressures.

 

[Nicolinux 288]

Well that's weird. I don't have time right now to take the head apart so I just changed filament and printed again (same settings, nothing touched):

There is still underextrusion, albeit hardly visible. Printed before with Ultimaker PLA (grey metal I think) and now with Ultimaker PLA gold.

 

[illuminarti 18]

Well that's weird. I don't have time right now to take the head apart so I just changed filament and printed again (same settings, nothing touched):

 

There is still underextrusion, albeit hardly visible. Printed before with Ultimaker PLA (grey metal I think) and now with Ultimaker PLA gold.

 

Judging from the crud on the side of the print, it looks like it was printed without retractions?

[Nicolinux 288]

Nothing changed, retraction was on. Here are the complete print settings:

 

• layer height: 0.2
  
• shells: 0.8
  
• retraction: on
  
• bottom/top: 1.2
  
• fill: 20
  
• speed: 40
  
• temperature: 230
  
• bed temp: 65
  
• support type: touching buildplate (15%)
  

 

[illuminarti 18]

Well then it looks like that filament is oozing a lot more than I've seen any other filament on my UM2. What are your in-printer retraction speed settings? And what speed travel?

[chrisp 10]

I've been wondering why everyone keeps their build plate heated to such high temperatures?

I set mine to 55 degrees celcuis and it sticks perfect.

If I set it any higher than 55 degrees celcius, then I get warping at the bottom. Not really warping, but it melts to the plate causing minor deformities.

Maybe it's the temperature difference between where we live?

 

[Nicolinux 288]

Well then it looks like that filament is oozing a lot more than I've seen any other filament on my UM2. What are your in-printer retraction speed settings? And what speed travel?

 

I took your settings and made sure to check them before _every_ print. 5.5mm at 35mm/s

 

[Nicolinux 288]

I've been wondering why everyone keeps their build plate heated to such high temperatures?

I set mine to 55 degrees celcuis and it sticks perfect.

If I set it any higher than 55 degrees celcius, then I get warping at the bottom. Not really warping, but it melts to the plate causing minor deformities.

Maybe it's the temperature difference between where we live?

 

It could be. It is winter here in Germany. I noticed that if I want to have something stick like hell, I'd use the gluestick, smudge it around with a bit of water and then set the heated bed to 75°.

 

[braddock 1]

I've had this exact kind of blockage, a number of times, also having to cut the end off the material.

I've not gone so far as to remove the teflon coupler, I was worried about not getting it all back together.

 

 

 

Ok here we go......

This clog was the worst by far. I started out by trying the change filament option via the controller. Nothing but grinding away the filament is what happened. I tried heating up the hot-end to 250 degrees Celsius and then tried removing the filament via the control panel and with a slight pressure of me pulling the filament with no success. I also tried move material but the filament was already grinded away inside the extruder. So what else is left? Taking apart the hot-end to see what is going on.

I usually start by taking out the 4 side fan screws and I tie up the fans and brackets to the axis rod so I get them out of the way and prevent the wire shielding to get heated up by the heating element.

 

Then I made sure to mark with a pen, the distance the Bowden tube is sitting inside the Teflon fitting. You can see the line marked with 3 dots which is not exactly against the Teflon, but close enough for reference.

 

I then loosened all 4 screws out of the aluminum bracket holding the nozzle in place and every other time so far the Bowden tube has come out without a fight. This time was different. I had to heat up the hot-end starting at 190 and gradually moving up to 250 degrees celcius until I could finally loosen the Bowden tube up enough to pull it out.

This is how the filament looked once I got it out.

 

Then I cut off the filament so I could get the Teflon fitting to go back on the Bowden tube to make sure it is sitting all the way down.

 

It's perfect no issues here.

Next I took the Teflon fitting and observed it very closely only to find that there's a lip inside the end that is seated into the metal hot-end nozzle adapter. I took my digital caliper and measure 3mm in diameter for the opening at the top of the Teflon fitting and 2.5mm for the diameter at the bottom of the Teflon fitting.

 

 

 

 

This I noticed once before but didn't think to measure it, but there is definitely some orange plastic stuck inside where the lip is and extremely hard to remove. I would like to razor knife the lip or drill it very carefully, but I need confirmation before doing so if it's not normal.

I then cleaned out the nozzle and saw light through it, which is hard to see in the pic.

 

 

 

Everything else looks great and I can't find anything wrong here besides the Teflon fitting.

 

 

[Nicolinux 288]

Another update since I still haven't got the time and motivation to disassemble the head.

I printed the exact same object with the same settings twice. Only switched between Gcode and UltiGcode.

The difference is striking. No underextrusion with normal Gcode. So here is another theory. It is not a software problem only but maybe a combination between UltiGcode and the firmware that leads to underextrusion (maybe a problem with the extruder controll).

@braddock and Chrisp: Did you guys solve the problem? If not it might be an easy test to switch to regular gcode and print a test object.

Here are my settings and the link to the model (link)

Regular Gcode:

 

CURA_PROFILE_STRING:eNrtV8Fu20YQvRJGP2KOKRqzS0pK4gi8JLVzaYoAVlHH
F2JFDsWtSS6xu7QsCfr3zi5JibKlNGlyKWAdJPFxZnZm9s1bsuArVHGOYpGbiPmh
t+RFEZtcJHcVak3QG0+hUTwxQlYxVnxeYDRTDXpaFiKNCxdg7xBQjExQjBQrLcwq
CplXyfW6wFiLNVLAsVcrUZlY14hpNGbdpcGyRsVNozAKR0fQMDoCjo6B4x04x/Qg
8KuJp5u6lspEM9kkuagWMG9EkdYFN+jZ70yqMuZpjpoKjv6QFfYucdrwIsYHoxp3
7500ubcUNcZGLlFFV7zQOADie1k0JUbBxJNyTfXnAou0M6Me8RIpxVTQryH30L8Y
P4Vt1U/A0TFwPASzQi6jgDGfDbevbfloMsR4KZvKRBP/AHWldreCV4f3SlHFdHGP
Be32wZ1ElnNqacuPQw9RHvSOEgu9uTRGlgdsG3ly/jcmxA5R3VkrT96jKnjtMiKA
mtmu3RUTTKxRF6klY3uDeCcqx8T22polUhYu/Y61gnZn4mW8J3baMftO0G4XokKq
yHWggxa8jkYUp73qqyqwWpi8y8QGyxpatZupNnyXK2P7q7jkDw7ZJZURSjwlwnRg
jpwGTGSm40w7cUbWwwFsS28RV27XMEu7nrgOb0PvwYcVUUcbXiV2Kl/v8PUQtva1
ULyws9tmMVeiHDaHGq14Ru3gaiEqKthdOQNd88TyIfA7mznX+Gi/97h1WYqUWmnT
cTjNMKqMJ4dO4eObQ08rPw9xLpUStKNxUzmSW1Fb0WDy3fidtpkftaEEZY1VPBdG
HzMgblvFu6eqjTBOWjqzumioMdQtouMielHUwZnf0zxBW0T8EJ0Hj6AVQT/RPijj
LxKZYjS9LkSCKXDzFjYpX23tt0H6sTzenk3fcS0S0GgMra3fwu+W5NDykFyKgdRv
4S8ql2w2h3q/hSuiCsFDCafQn6yagl2H7nV62y561SkOUIPS1rFVoFY8tuUAWrZr
LwZeidRm6GWvKerH4ILB9eaoiG9h+mdFSuP8jQSeprCSjQK5rIBsYWALlhk2HLvY
hfvaUE/CfAgD6D5TUlxFzb7nRYP67ANl29/hc5rLxiDUkppHxKLNOKMEXvcGbk9h
KUwOJkcgQQCZZRT9Ddww+MxsdJphuPn1s82JlAGwSjWNvXZGt6xLwRrdPjUJ4DaY
+AyuNkOh3HYOdsn+mIPU1hlMypIqCOFyV8T+M12jks7JaXdK7W3lzq5zFTIGl6NH
Hp0hjMqSCgOrakCpJXffvgjwBReVW+qwGGJkY8Dtpz3CS6IuXyDICn5//xvoRCFW
1PTgNXzqbHzf96hH/SxdVil8eE9/7daM4fpIVoNa7BTRqCm3UR+DMTvp4OxSR8O9
C7wQmeUV5FQCCPMzdSKAr/hMFdJWkaYckInacXlOLRkx9i/O7gBuWdYPHAcSMMqP
9t/SOnP9sxbtQ9pLyydaFUmQQcsS7RY6zijqMY2C49cvzJ9QDhO4OQ+JsvbrONtu
oam7JTl1vM9olw3eYwWltGH39P9iUadG4yVl6xK1hyYIDZII0uW+5Kuzj2/Gp4Nq
gzXNum4H8eJ0Dken23MTHT7L9A+Q6fH3y3Sr9rPgWKTwP4Viz2fHl86O2Qktm14v
7YOQXcV6hiQAvZ5+z3kTsOPnTcB+6IHj1jrfHH8n2p7N2FcVnQmlzf+p7B90zoYn
DtoZ+8az1voEz+fz8/n8Pedz92Y7fJvagfuHwval+8DIIQMLhaR/CfqJvqcIbtIv
Owrujv92/udoljQcrr6kUcp1s2er7bXbU0J26EtY5uSwGzan8mVTGFEXu4lV2u/G
u6XZ6BVcTRhjPTjb9IZba3CzoRfj+GELn90fehi5dX/W2yeTbsM9ivYPz9WJkg==

UltiGcode

 

CURA_PROFILE_STRING:eNrtV8Fu20YQvRJGP2KOKRqzS0qKnQi8JLVzSYIAVlHH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 

[chrisp 10]

@braddock and Chrisp: Did you guys solve the problem? If not it might be an easy test to switch to regular gcode and print a test object.

I have been smooth sailing for the past couple of days, trying to finish up the Tantillus 3D Printer build that I'm currently working on. So far, so good on my end, but that is still to be determined. I am probably going to go back and try my Nemesis tonight(Orange).

That's odd that the code designed specifically for the UM isn't working as intended. Are you sure you didn't accidentally purge a minor clog from the hot-end?