gr5

If you print the test cylinder above I recommend you look at other people's results here instead but most importantly it is designed to be printed at 230C. If you print at any other temperature it's hard to compare it to another printer. I feel anything that reaches past 5mm^3/sec is a well functioning machine although some people insist on 10: http://umforum.ultimaker.com/index.php?/topic/4586-can-your-um2-printer-achieve-10mm3s-test-it-here/ Well... Marritt might be right. Or she might be wrong in this case. UM started shipping weaker springs recently. I don't know how to tell if you have the original (much too strong) spring or the newer weaker spring. I guess you could do a compression test on a scale. I think you should put it back in the middle and assume the problem was that loose set screw on the knurled sleeve.

Strange! That always worked for me. Try the "keep open faces" checkbox. And if that doesn't work get the free version of netfab as simon says and repair it.

Pictures are very valuable. I didn't realize your test piece had columns so close together!! If you had looked at the part in cura layer view you probably could have seen it wasn't retracting without even printing anything. Blue lines in "layer view" are non-printing moves. If the blue line has a little vertical section then that doesn't mean the head moves up - it is a symbolic representation indicating retraction for that move. So look for those vertical blue lines.\ Regarding the settings above - they might have prevented retraction as those 2 cylinders were so small and close together. "minimal extrusion .04" refers to mm of filament. There is (I believe) a 120X multiplier at .1mm layers so that would be 4.8mm of laid filament. Or maybe I'm wrong - maybe there *was* retraction. Those tiny things between the towers aren't exactly "stringing" but they are caused by tiny tiny amounts of leaking filament - they are easy to cut off with a razor. If the towers were farther apart it wouldn't look so messy as they wouldn't go all the way across. Some filaments (particularly white I have found is bad) string worse than others. The "bonding" issue at 220C is different for each type of filament so I can't comment other than to say if it looks better or feels stronger at 220 then use 220.

minimum layer time isn't an issue for this part as it takes more than 5 seconds per layer. It is difficult to print "holes" or "openings" on a UM printer. I have tried this in the past. One solution is to print vertically - I think it will look much better if you print with the long, thin edge down. You will need brim. Another solution is to use 2 colors. This will come out much better. Print the word "hello" in one color, then end the print, remove filament, put in new filament of contrasting color, then print a simple rectangle on top - you don't need to have the second part know about the letters. Here is an example: To get the two prints to line up I made the words inside a box of equal size such that the second layer printed a rectangle, then I manually removed the second layer from the print with a text editor so that the print would halt before printing the second layer. That way when placed in Cura both prints were aligned even though they were sliced separately. For the upper part it is better if you do a thick bottom layer - say .3mm so that there is more room for the plastic to squeeze around the letters and into the cracks. And print it slow although 30mm/sec is probably fine. If you choose the 2 color method, please post results as I will have suggestions to improve it! Also you can print the second color *on top* instead of on the bottom if the top is flat. This would be most easily done using "pause at Z" and changing filament in the middle of the print.

This is most likely a hardware issue and not a software issue. There was probably a very high acceleration between brim and part. I would mark the 6 pulleys for the Y axis but almost always it is one of the 2 pulleys on the short belt that slipped. Alternatively it could be missed steps and it *is* possible that firmware messed up but very unlikely. You probably just have a setscrew that has worked its way loose over the years. Tighten the hell out of the motor one and the other one on the short belt. Enough so that the allen wrench twists quite a bit. But don't break your wrench.

Check the X axis left homing switch. Bend it out slightly so that when it homes it hits the homing switch .5mm sooner so that you don't scratch the left side. You should use needle nose pliers.

No. I can clearly hear the retraction occurring before it moves over to the side. The problem is that due to thermal expansion as the PLA heats from 20C to 220C and also existing pressure in the bowden (which should be zero due to the retraction). Oh!!! I thought of one more thing! When I first got my UM2, I found retraction of 4.5mm was NOT ENOUGH. I had to go to 5.5mm and that was just enough. If you print something with mega retraction you can tell because you can watch the filament in the bowden and you can see it move up and down touching the top of the arc and then the bottom of the tube at the top of the arc. I could see that 4.5mm was not enough. Part of that extra 1mm was due to the bowden moving up and down on top of the print head. But then later I rebuilt my print head such that the bowden doesn't move anymore. So now I use 4.5mm retraction.

When I was printing robots, I had the most trouble with the antenna. I printed only the top half of the robot head and tried everything. Only when I printed 2 did it work out. You could save time and do what I did - just print the top half of the head. The feature for that is under "advanced" "quality" "cut off object bottom".

Almost same thing for me. 220/65. Also I used Ultimaker light blue filament - same as you I think. Maybe mine *did* leak as bad as yours and I just forgot. I wasn't there when it finished - I definitely remember I had to cut some strings off the very top - it was a print of a building with a clock tower in the middle and the tower had a wall that stuck up higher than the other 3 walls. I look at it now and it looks perfect. Unlike this print, the robot prints I posted earlier didn't have huge strings like this. I don't think they had much of any strings. But of course I didn't use "cool head lift" either.

What was your print head temperature?

I compared my print with yours (I never delete anything) and I had the same settings for: layer height cool head lift cool minimum speed minimum layer time

I don't know why it is leaking so much with "cool head lift". This is a feature that I try to avoid and is only used typically on the top few layers of very small things like robot antenna. Fortunately it is easy to remove excess strings with a razor. I have gotten perfect antennas on the robot without cool head lift - I just print 2 robots like you did. So in other words it's probably best to use one technique or the other but you don't need both. If you go back to my post #24 - I didn't use cool head lift that time. But on other prints I *do* use it (when I don't want to print 2 or there isn't room on the platform to print 2). cool head lift has always left strings for me - but not this badly! The last time I used cool head lift - about 2 weeks ago, I had all the same settings as you but it came out much better. My retraction was max speed and 4.5mm (on the panel).

Unfortunately the video camera is shaking too much to see the bed shake but I believe you. I've seen some bouncing of the bed but it has never been a problem for me. I guess I don't print very often this close to the front edge. Does it affect the shape of the final print? Maybe better to take a photo of the front edge of the print and then another photo of the rear edge. This is a new issue you are bringing up. No one else has mentioned it but I suspect it is "normal" for the UM2. Possibly speeding up or slowing down the test head speed will help change the induced vibrations away from the resonant vibration? If you have a consistent way to measure your vibrations that I can perform then I can compare my machine to yours.

Hey! Try the new cura! http://umforum.ultimaker.com/index.php?/topic/5404-all-hail-gregoire-passault-master-of-the-raft/&do=findComment&comment=47889

It's probably skipping part of your model. Look at the model in cura layer view carefully - especially starting where the problem occurs. Perhaps a wall is too thin to print. Cura will print down to .4, .6 or .8mm wall thickness only. Not sure why there are different limits at different times but the nozzle hole width is .4mm so obviously no thinner than that. Also look at the part in xray mode - anything the color red is a potential problem.

I have found that white PLA is a little more difficult to work with. But colorfabb makes great filament so hopefully this will help.

Everything above is the best advice. But here's also something to think about - temperature versus strings in pictures (which Daid mentions above): http://umforum.ultimaker.com/index.php?/topic/1872-some-calibration-photographs/ I have the second shim robert mentions - the second link. The first one broke after a few hours of printing. The second one has lasted over a year. Also printing slower or thinner layers will help (you don't have to do both!). The slower you print the less pressure is in the head so that when it retracts (which takes time) there is less pressure to start with so less leaks out.

You may have a nozzle clog but from your description I'm going to agree with Illuminarti and go with "ground filament" at the feeder. Look inside the black feeder on the back and see if you can see if the filament is ground up. Regarding the spring tension position - UM might be shipping printers with weaker springs so the center position might be better than all the way up - we (Illuminarti and Ian and I) just don't know at this time as we all have the strong springs.

@LTC - please don't use fileswap. I doubt they will still be hosting these pictures even just a month from now let alone 10 years. Instead click on "gallery" on the top left corner of this page, then click on the blue "upload" button. After uploading start a new post and click "my media" next to the smile icon.

Yes! Yes!

Giving advice is so much easier than fixing it myself! When I have a problem I pretend someone else posts on the forum. I never like the advice I give myself - I think "surely there is an easier way!". Seriously though - I know so much because I read so many posts! Most of what I post is something that someone else learned.

It sounds like some of you have a loose bed that doesn't come back to the same spot every time. Please experiment with pronterface: http://koti.kapsi.fi/~kliment/printrun/ Install pronterface, connect to UM2 with USB cable, and then connect with pronterface. It should be able to pull back lots of great info from Marlin. Now experiment by moving print head up and down: G0 Z0 That should move the bed to where the nozzle touches the glass G0 Z0.1 That should move the bed to where you level with paper G0 Z30 G0 Z0.1 That should move the bed up 30mm and then back to leveling area. Try moving the head around - maybe push on the head with your fingers. Try to get the problem to repeat. I'm thinking something is loose somewhere. I think that is Nicolinux's theory also.

Also you might want to try another SD card. Get one of the same manufacture, and size and type if you can. I don't know which types of SD cards the UM can read. If you print the exact same file twice does it skip in different places of the print? If you put the SD card back into your pc, can you use winmerge or another program to compare that the file on the PC is identical to the file on your computer? It seems like a few lines got deleted or corrupted. http://winmerge.org/

It might actually be all 4 axes that are acting strangely. In the video the printer does some bizarre stuff. I think it's skipping a whole layer and then the extruder tries to over extrude and a fat string comes out. This is because the E values are not relative so that when you skip several layers, now the E position is much farther along and it overextrudes a fat string. The next time this happens it skips maybe 10 layers and another fat string of filament comes out at the same time. Both of these happen after 3:00 in the video - maybe the first one is at 3:10 possibly. The second skip is only a few seconds later. So although anon's comment is clever and good "thinking out of the box", I think in this case, Daid's theory is better. So what to do next? It's time to get your hands dirty TUKL. Simply disconnect both of those gray cables (mark them carefully so you can put them back. Then maybe try to re-arrange them slightly and connect them back in. Just the act of disconnection and reconnection might be enough. Hopefully Daid will come back and inspect your photo - I don't know what the "wrong" way is (red to red? red away from red?).

None intended! Sorry! After I wrote my post I re-read it and I also sensed sarcasm and thought maybe I should rewrite the post. Then I thought that editing my (already posted) post might not help as you might have already gotten it as an email (I follow all posts that I post in). So I left it as is. This is a new problem for me. I didn't know the head was capable of touching the front panel - I thought it would hit the block first like it does on the sides and rear.