Gap between solid infill

[nux 0]

Hello,

A recently noticed an issue with a print. I could pretty much see the infill below the top solid layers.

I print with 0.1 layer height, 0.8 shell thickness, 0.6 top/bottom infill, and normally at 205C at 30-40/mms.

Here's a photo of the first time I noticed the issue:

http://nux.net/p/um-1.jpg

I then loaded up a "lozenge" that I normally use for testing new filament colors after changing. This has printed like a champ up until now:

http://nux.net/p/um-2.jpg

I don't recall the print speed on the lozenge as I created it weeks ago, but I left the temp at 210 when I normally print at 205.

I found I was able to find a nice balance of speed/temp and have printed many objects. It seems something has changed somewhere with my machine.

My first thought was under-extrusion, that I've heard so much about, so I printed another circle here: http://nux.net/p/um-3.jpg

I was expecting an issue with this, but this tells the extrusion doesn't seem to be the cause.

Leading up to these issues I noticed that on a large first layer, the extruder moter would skip occasionally. I believe this was caused by a string in the tube, as it went away after manually removing the filament and using the atomic method to clean the nozzle.

I am testing this right now with the light blue ultimaker filament that I received with the printer.

I mostly use matterhackers pro pla (which is what the white/yellow is in). I've had success with them up until these recent issues.

I am printing a 20mm box from the "Essential Calibration Set" and it just started the first solid of the top layers. I am seeing not a single line was able to go over the gaps in the 20% infill. Here's a photo of the top of it when it completed: http://nux.net/p/um-4.jpg

Here are some photos I quickly took as it was printing:

first top solid layer: http://nux.net/p/um-5.jpg

second top solid layer: http://nux.net/p/um-6.jpg

This 20mm box was printed at 210 at 50mm/s

Any suggestions?

 

[anon4321 16]

What level of fan are you using? Is the fan used on the calibration mm3/sec cup?

Are the belts tight? Have you lubricated the rods?

Could be that the fan is over cooling things including the nozzle during your prints particularly if the fan isn't used on the cup. Also with the cup, since the harder part is printed further up, even if the fan is on, it might not "backwash" into the nozzle like it would closer to the bed where the easier part is printed. The easier part might be less affected by the backwash.

Might be slightly loose belts or friction on the rods preventing the nozzle from precisely positioning during the layer fill.

 

[yellowshark 153]

Hi my first point is that IMHO using 0.6 solid layer height when using .100 layers is probably not enough. I get the feeling, from my experience but NOT scientifically corroborated that the thinner the layer height the more solid layer depth required. I.E. if you are OK with .6mm when using .200 layers (ie 3 layers) the using 0.6 when using .100 layers (6 layers) is not enough.

If I were printing with .100 layers I would use a top layer depth of at least .8mmand be prepared to go higher. My general rule also is that the higher the infill density the less solid top layers required (and of course vice versa). I also suspect print speed will impact i.e. the slower you go the better those thin solid layers will lay down over the infill.

 

[nux 0]

I've had this machine for about two weeks now. I recall in the manual it says maintenance should be done every 6-12 months so I hadn't thought of that.

The belts all feel the same and I can feel lubrication with a quick run of my finger over a portion of the rods.

The fans have been on their default settings - nothing changed there.

I think my next step is to completely remove the bowden tube and put some vegetable oil on some filament and run it through it a few times to make sure there aren't any strings or anything left in there that aren't visible.

I have changed the filament a handful of times, and take the time to pull the strings out but wouldn't be suprised to find something in there still as some of the strings are long and barely visible.

One note is that this lozenge has printed just fine at least a dozen times.

I slowed the 20mm cube speed to 50% of 50mm/s, so 25mm/s and the results were the same.

 

[IRobertI 520]

Those lines look pretty much perfectly spaced so I wouldn't suspect the belt tension or backlash.

The next thing to check is if the nozzle is clean but you say that you've cleaned it using an Atomic. Do you get a perfectly shaped cone without any debris in it?

Next I would suspect the Teflon isolator but you've only had the machine for two weeks so it seems a bit too early for that...

And you're not printing particularly fast. Plus you tried it at 25mm/s which is definitely not fast.

This issue is really starting to bug the shit out of me to put it bluntly. It's not like you're the first one with this problem, unfortunately.

 

[nux 0]

The cone came out without any debris. I can shine a light through the bottom and see it clearly when looking down.

I did just look at the bowden tube coming up from the extruder. It looks a bit messed up. I've had issues with filament skipping while loading initially. I was able to resolve that by really straightening it before it goes in, and cutting with a pliers to get a crunched cut.

Here is a photo: http://nux.net/p/um-7.jpg

It is a couple MM at the bottom of the tube. Can I cut this off? I am crossing my fingers this is the issue. I need to run to the store to get some vegetable oil to lube this up before I try again.

 

[IRobertI 520]

Yeah I would cut a bit off and get a clean entrance for the filament. In production they widen the end of the tube a bit but it looks like someone used a really dull tool or something on that one.

 

[nux 0]

I clipped off a small portion on the bottom. I put a little vegetable oil on a piece of filament and ran it through the tube a couple times to make sure it was clean and I didn't notice any resistance.

I put it back together, fed some filament in and there isn't much of a change.

The gap in lines seems a bit more random. At first I was excited as it started laying it down well, but within the second layer it got worse. At about layer 3 I stopped it in this image. I pulled it off the bed before it cooled so it twisted a bit, that wasn't related to the print:

http://nux.net/p/um-8.jpg

There's even a decent sized gap between some of the perimeter lines here.

Anyone have any suggestions as to what I could do next?

 

[nux 0]

I was able to get a good print going at 220C and 25mm/s =/

 

[gr5 2,167]

Well it looks like underextrusion. Did you update the firmware recently? There have been changes related to nozzle and bed temperature in recent versions of Marlin.

I suspect you have temperature issues. While printing go into the tune menu and watch the temp carefully - see if it is oscillating. It shouldn't vary by more than 1C from goal temp.

Also consider maybe the thermistor is bad. You can test nozzle temp using this video as a guide:

 

 

[3dmaker4u 0]

How many topics of this kind users will have to open, and some of us to debate around?

I think there are already 2-3 months since this issue has been reported. And in all situation it was about new printers.

The same mysterious behavior that none of the experienced users could come to a solution.

I would normally expect that UM people will post something about, at least something like "we are aware and do our best to come to a resolution, and we'll let you know about the outcome". But... :(

 

[yellowshark 153]

Ah well that is quantum physics for you

 

[nux 0]

I did the temperature test and it looked ok.

I do notice that if I'm printing at 210 and set the temp down to 205, it will often go down to high 190s (197ish) before going back up to 212ish, then down to 202 up to 208 etc etc. It takes quite a while and fluctuates greatly while adjusting. The fluctuation is greater if the adjustment is larger. A 220->200 adjustment would probably reach into the high 180s.

When it does get to the target temperature it holds it well.

I haven't ever updated the firmware. I'm currently on 14.09.0 with the latest version of Cura. I don't see a change long or anything in the github repo on firmware, but I see 14.12 is the latest releases there.

I could upgrade to the latest RC or release version - do you think that could help with my issues? I'm guessing it wouldn't based on the fact that it printed well at first. My guess is it's a hardware issue as nothing changed with the software since I got the machine.

With this being week two of a year warranty, what could I expect Ultimaker to do in this situation? I submitted a ticket yesterday but don't expect to hear back until Monday or so.

Thanks

 

[IRobertI 520]

I would normally expect that UM people will post something about, at least something like "we are aware and do our best to come to a resolution, and we'll let you know about the outcome". But... :(

 

Well I'm not strictly a "UM person" as I don't work for the main UM company but I can assure you that they are aware and they're working on it. There were some fairly in depth measuring done on the whole extrusion system a while back that was posted in the feeder design thread. They're working on a few other things but I'm not allowed to talk about it and I can't say when/if/how that will be made public.

 

[illuminarti 18]

I think what you are seeing is speed/pressure-related under-extrusion. To some extent this is always going to happen (see, for instance my study of the effect with the Ultimaker Original). The extrusion cylinder test shows that your printer has plenty of torque to keep pushing the filament continually - it just doesn't push it as far as it should, hence the thinness of the infill lines. The question is, what is causing it to be so severe on your machine, now?

 

I was able to get a good print going at 220C and 25mm/s =/

 

Do I interpret your comment correctly that the problem goes away if you print hotter? If so, then my first advice is definitely to print hotter. :-)

I'm not sure why the behavior of your printer has changed over time. Has the environment that it is in changed at all? It is now hotter or cooler than it was? And what about the filament that you are using - are you getting further down the roll now, so that the filament has a tighter radius of curvature on it, leading to more resistance in the Bowden tube?

I'm curious if the problem is less severe when you first turn the printer on from cold, and then gets worse with multiple print attempts, or does that not seem to make a difference?

Also, have you tried doing a factory reset? (Maintenance -> Advanced -> Factory Reset). Take the filament out first - since it will make you reload it.

 

[illuminarti 18]

Oh, and that was the initial response to your ticket :-)

 

[nux 0]

Thanks for the input.

I have gone through maybe 2/3 of the spool of UM light blue filament. The first spool I noticed the big issue on was a brand new yellow matterhackers pro pla, so I suspect the tightness of the coil is not it.

I had the printer off until I made the lozenge yesterday. It was as cool as it gets when I started that print. I didn't see any difference between prints with the machine being hotter/colder.

The room the printer in is my work office, and sits around 65-70 degrees with about 30% humidity.

I only moved the printer from my desk to about 5 feet from where it was, to a dedicated stand I got for it. I also put it up on some 35mm feet (http://www.thingiverse.com/thing:563927) with small non slip circle things on the bottom.

As I think back now, I recall seeing mid print that I could see a small ridge/indent in the perimeter lines few days to a week before I noticed this current issue. Looking back I suspect that was the same issue as it was starting. I would just adjust the temperature up a little. This was surprising as I was printing at settings that worked great shortly before with the same filament.

As a stopgap to get this to work I have increased temperature and slowed down the print speed.

I have not done a factory reset, I will try that tonight or tomorrow.

Thanks!

 

[gr5 2,167]

I don't recomend updated to firmware newer than 12.09. If anything most people seem to have *more* problems than less with the latest firmware (although a few people report that their nozzle temp is more steady - other's report that it is less steady).

As far as "yikes lots of people have this problem I hear silence from UM...". The problem is clearly not "one single issue" to me. I mean if you wanted to give it a single description I would call it "underextrusion". But underextrusion has about 20 possible causes and from reading all the posts they seem to be different causes for different people.

The most common:

1) Keep the spool on the floor - having it go into the extruder at an angle grinds up the black PTFE feeder housing causing the part to be the wrong shape, add nozzle clogging gunk to your filament, and add friction.

2) The white ptfe isolator at the nozzle head can get compressed too much (4 screws too tight or spring too tight) and result in a smaller than desired hole for the filament to go through. Consider loosening the 4 long screws a turn and/or drilling out the isolator if you have trouble sliding filament through.

3) The end of a filament being more curved is harder to get through both the bowden and the white isolator especially if it has shrunk.

4) Nozzle temperature fluctuations. Simplest fix is sometimes to add some silver high temp grease between temp sensor and heater and the block.

5) gunk in the nozzle - permanent baked on gunk lining the inside of the nozzle - not causing a complete clog but slowing down the ability to print at speeds that used to work.

All of these 5 issues are underextrusion issues that typically don't show up on the first 10 prints but show up after 10 to 1000 hours of printing.

 

[nux 0]

Thanks gr8!

When I was cleaning the bowden tube the top clip fell down the gab. I took the 4 screws off to get it out. I'm trying not to get my hopes up yet, but so far it looks like the cause of this issue may have been those screws too tight!

Approximately how tight should they be? I didn't see any mention in the assembly instructions beyond "tighten".

I am crossing my fingers that this was it!

 

[gr5 2,167]

I don't know how tight. I think maybe it's not so much how tight are the screws - but how tight is the bowden pushing on the isolator. I just read every damn post on this forum for the last few years and when someone finds a solution (among many solutions that work!) I mention it again.

 

[illuminarti 18]

They should just be barely tight... tighten them until they just start to feel snug.

As George said, the main thing is probably that the Bowden isn't too tightly forced into the teflon coupler, or squeezed between the collet at the top and the coupler at the bottom - that can cause the end to deform.

The screws should be fairly loose - then push the Bowden down into the coupler, and see that it seats all the way down. If you pull up on the bowden (without the blue clip inserted) the collet and tube should move up and down a mm or two. Make sure that the end of the Bowden tube doesn't come up above the top of the teflon coupler as you do that. Then lock it in place by pushing down on the Bowden tube, while lifting the collet up around the tube, and then sliding the the blue clip into place. You should end up with the Bowden seated all the way down into the coupler, but not under appreciable stress. And the collet should barely move in either direction if you tub gently on the Bowden.

 

[nux 0]

In case anyone stumbles across this thread, I thought I'd post an update as to what the problem was. The issue I was having was making any sort of adjustment of the 4 screws without re-seating the bowden tube.

I "fixed" this issue by loosening the screws a lot, but the plastic pieces of the head had decent gaps.

To resolve this issue I removed the bowden tube, tightened the screws so they snug and holding everything together properly, and then placed the bowden tube in like illuminarti stated above.

I have learned if any adjustments/changes are made to the print head (like removing 4 screws to dig out a bowden clip that fell in), you need to re-seat the bowden tube.