Irregular surface printing

[dergoldstein 1]

I've had my UM2 for about 4 months now, and in the past couple of weeks the print layers have become very irregular. With the same materials and settings I used to get perfectly smooth surfaces, so I'm trying to figure out what changed.

Here are 2 photos of a first layer:

Ignore the discoloration of some areas, I had just replaced a black PLA spool so those are just leftovers.

The gaps between certain lines indicates that there's a belt backlash problem, even though this print was made after I tightened the short belts and equalized the tension on the long belts (if there are other operations I should try with the belts please let me know).

What I don't understand is the gradual alternation between relatively smooth lines (about 20 of them) followed by a bunch of very rough, noisy lines. There seems to be a cyclical problem. If all of the lines were noisy I'd think there's a problem with the nozzle, but it comes and goes.

Now here are 2 photos of the same part, several layers into the print:

Same exact problem, except the cyclical nature is even easier to see. A bunch of smooth lines with good spacing between them, followed by a bunch of rough lines that seem to come in "pairs" (two close lines followed by a gap).

In some of the rough areas the lines become so thin that even a pair of close lines aren't enough to close the gap.

If anyone has any idea what's going on please let me know.

[neotko 1,417]

Your nozzle it's too close to the bed. That different density on first layer it's the nozzle dragging filament (you should see your nozzle accumulating filament dragged). Redo the bed calibration and this time make it grap less on the paper on the 3 points. Also remember to calibrate with the bed on to compensate that the glass while hot it's slightly bigger.

Also a first layer with more height should fix that.

[dergoldstein 1]

Your nozzle it's too close to the bed. That different density on first layer it's the nozzle dragging filament (you should see your nozzle accumulating filament dragged). Redo the bed calibration and this time make it grap less on the paper on the 3 points. Also remember to calibrate with the bed on to compensate that the glass while hot it's slightly bigger.

Also a first layer with more height should fix that.

That's actually something I played with a lot, and it's not it. I get the right amount of "squish" when looking at single lines of the first layer (look at the perimeter line on the first layer). If I raise the nozzle any higher I get lines that don't connect to each other no matter where they are.

If it were too close all the time then there wouldn't be patches where there's a smooth surface, it would just drag everywhere evenly. This is why I'm thinking that there's a cyclical problem -- there's a batch of lines that print well, followed by a batch that prints poorly.

[neotko 1,417]

You need to adjust also first layer width to compensate.

[neotko 1,417]

If that fails. Adjust the bed bolts since you might need to calibrate the bed.

[neotko 1,417]

Read this https://ultimaker.com/en/community/19173-lots-of-trouble-leveling-bed-since-heated-bed-install#featured

[neotko 1,417]

Oh well. That looks like underextrusion sorry.

Raise the temp. But first do some atomic pulls.

[shurik 87]

It is not something unusual for the first layer. Either the nozzle is too close to the glass or you print too much volume for the given temperature, so it clicks and gives you the slight underextrusion.

What filament do you use and what are your print settings - temp, 1st layer width, speed? As neotko suggests, you may want to raise the temperature for the 1st layer.

[dergoldstein 1]

I think I managed to solve it, though the solution seemingly has nothing to do with the problem... I went over a long list of possible issues and found one thing I hadn't done -- check the isolator. I'd already removed and cleaned the nozzle a couple of days ago, but I didn't remove the isolator to check if it deformed. It actually didn't, but the part that touched the nozzle was almost black, so I did what I saw on a bunch of different posts and used an exacto knife to shave off the ~0.5mm from the end (so that it's completely white). I also added an extremely slight bevel using a drill bit (manually), so that there's a slight funnel on both ends (almost unnoticeable, I knew not to go overboard with this).

Here's the latest print, a few layers in:

(There's an extra outline on top because I didn't pause right after a layer had finished)

Lines now print smoothly and tightly together. I've been through enough problems with the printer to be too optimistic right away, but for now it appears that it's printing almost as smoothly as when I first got it. I forgot it was able to print such clean surfaces and walls, I just got used to rough surfaces.

My takeaway from this is to avoid temperatures over 230. I'm going to put ABS aside for now and try to remain within 220 for most prints (my standard is 215, depending on the filament).

Thanks for the ideas.

[neotko 1,417]

The new couplers (the ine um2+ & extended +) has will last longer at high temps. So in the future you might be able to have a longer life coupler.

[shurik 87]

That's strange. Even the standard coupler should live much longer than 4 months of whatever normal usage.

If you print PLA, there's no real need to go anywhere over 210C anyway, and with the Olsson's block I usually print at 185-200C.

[neotko 1,417]

Indeed. Maybe the heat sensor it's way off and it's going +15C of the target. It's worth to test it with some tool.

Also since you use abs/pla, remember that it's a must to do atomic pulls from one kind of plastic to another, to avoid underextrusion.

[dergoldstein 1]

That's strange. Even the standard coupler should live much longer than 4 months of whatever normal usage.

If you print PLA, there's no real need to go anywhere over 210C anyway, and with the Olsson's block I usually print at 185-200C.

 

I've been using it for work pretty much around the clock since I got it. I'll sometimes go through a 750 grams spool in a week.

Also earlier on I read somewhere that to really purge the nozzle you should bump up the temperature to the maximum and let it run for 30 seconds, which was obviously a mistake (and I've must have done it like 10 times in the first month).

The lowest I can go with the PLAs I use (mostly colorfabb) is 205, any lower and the feeder starts grinding the filament. But at 205 the squish isn't enough and material can build up around the nozzle.

[yellowshark 153]

Hi, "the lowest I can go.... is 205" is meaningless really unless you relate that to layer height and print speed. I can print colorfabb at 180 (which is meaningless ) and so would say that if your minimum is 205 you have a problem with your printer.

Edited February 14, 2016 by Guest

[shurik 87]

If you want, PM me. Israel is not that big, we can meet somewhere.

[dergoldstein 1]

Hi, "the lowest I can go.... is 205" is meaningless really unless you relate that to layer height and print speed. I can print colorfabb at 180 (which is meaningless ) and so would say that if your minimum is 205 you have a problem with your printer.

I default to 0.1mm layers at 50mm per second, which I think is the cura defaults. While I understand that there's a correlation between the 3 factors (well, at least now I do...), surely there's a specific melting point to specific polymers?

What I mean by 205 is actually not while printing, it's when you use the "insert material" function to run filament through the nozzle. I do that and mess with the temperature until the feeder is no longer clicking (I dial down the temperature very gradually to find the lowest point where it's running smoothly). I then stop the process, disconnect the 2 ends of the bowden tube, and make sure there isn't any unusual friction when moving the filament through the tube manually (possibly by bumps forming when the feeder clicks and grinds a bit of material). That's how I figure out the temperature for a particular filament. For PLA it's often around the default -- 210.

About the sensor, I've disassembled the hot end a couple of times and there isn't a clear stopping point to inserting the thermocouple, it lets you keep pushing it until it's really stuck and hard to get out again, so I don't push it in as far as it can go. This could mean that I'm getting a lower reading than what the actual temperature is. Maybe with the olsson block it's different but I don't have one yet.

[avogra 68]

During insert material ( or.. I only use change material), I too have clicking but printing works fine. I got to conclusion for myself, that Ultimaker intentionally reduce the feeder motor current during insert material in the case of bad filament or so.

[yellowshark 153]

Hi @dergoldstein, sorry I seemed to have missed you last post. Yes you are right, for PLA it is around 180, might be a bit less, my memory fails me. I have certainly manually fed (which would equate I think to your "insert material" on your UM printer) PLA at 190. That makes me suspicious of your 205; but I have a 3ntr rather than a UM so I cannot state absolutely that you have an issue, I just have a suspicion