gr5

Well Sander should be back in the Netherlands today and he probably has too many tickets to get through and other business and meetings and things to do but hopefully this week he will put out a new schedule if it has changed since his last schedule.

If you guys are going to report a bug Daid is more likely to look into it if you provide a screen shot of the layer view that demonstrates the problem along with explaining what you set nozzle size and shell to. If you make a little extra effort, then maybe Daid will also.

It can't be. Look at the first photo again. I already tried to explain that it can't be the filament.

I only have the UM original but I'm pretty sure there is no locator. In fact it is recommended not to file any of the rods flat as supposedly that is worse then standard "perfect" rods. So just loosen that grub screw and slide the pulley so that the belt is straighter. Then of course tighten the heck out of it so it stops moving around.

Find out at what temp your heat block boils water. Set it to 90C, then increase by 5C until it boils. It should boil around 95C to 105C. If outside that range something is wrong with your temperature measurement system. It could be the wiring, it could be the small electronics board on you print head. It's usually not the thermocouple - but sometimes even that can go bad. If the print head is just always 20C hotter than you could live with it and always print at 180C instead of 200C. If you go any lower than 180C you will occasionally go below 170C which halts the movement of the extruder. Note that there is a spare cable for the temperature measurement. You can just switch to the other connector (at both ends - the print head and the board underneath).

I agree. That's why I almost always set the speed in Cura to 100mm/sec. The number on the UM2 (and UM original) shows the "feed rate %" which is a multiplier. Normally it is 100% at the start of every print.

But I really doubt that this is caused because it's too hot. I don't see this on my prints and I tend to oscillate between printing fast at 240C and slow at 190C.

So Jet - what is your favorite 3D modeling software?

I'm going to assume that the *back* of the scull was printed first (on the bottom). YOu have some *serious* unsupported overhang there. That's difficult to print. For a bust or skull you probably want to print the flattest possible surface down. And the bottom of things printed often looks worse so for this skull I would consider printing in the "normal" up/down position. The rest of that skull has tons of bumps on it - it looks like crap. Try cutting the print speed in half to see if those go away: See this photo of this pumpkin with similar (maybe?) bumps: http://umforum.ultimaker.com/index.php?/topic/1872-some-calibration-photographs/?p=24010 If you are using these for cold casting you can use sandpaper or metal files to remove bumps, and bondo to fill in gaps. You can use a solvent to hide the layers although like Illuminarti says, that's much easier with ABS. Check out this nice finished print: http://umforum.ultimaker.com/index.php?/topic/467-post-your-latest-print/?p=18661 posting #541 Scroll through all the prints in there - some of them are amazing and you can't tell they were 3d printed It takes a few weeks of printing and experimenting and reading the forum to get your cura settings dialed in for your particular needs. You can certainly do much better than that skull.

How slow? I found no difference between 50mm/sec and 100mm/sec. Did you look at my photo in my post above?

Read about pid controller on wikipedia. It explains how to tune yourself and how it all works. If the temp is unstable you want to decrease P, and/or I or increase D. So increase D or decrease P or decrease I. Usually P is the culprit but it depends how slow these cycles are. For a quick test, cut both P,I in half. Try that. If that isn't much better, try doubling D. These values can be way way off (by a factor of 10) and it might still work pretty well. In fact D isn't really needed as much as the other values and can in theory be 0. The latest version of Marlin and Cura 13.06 or so and newer have these really cool features which graph the contributions from P,I,D so you can see what is going on. But you can't use the Marlin that comes with Cura you need to get it from here: http://marlinbuilder.robotfuzz.com/ And I don't know how to turn show this stuff in Cura. You might have to edit configuration.h and enable something. Not sure.

Why is this pinned? This is pretty old. The UM Original and UM2 have different versions of Marlin now. For UM Original the easiest place to get the bleeding edge Marlin - well it's a month old maybe but it has been tested on UM Original and is much more recent than the version that comes from Ultimaker: http://marlinbuilder.robotfuzz.com/ This replaces Daid's marlin builder but it was written by someone in the community - not a UM employee. The source code for the UM Original and for the above link is from here: https://github.com/ErikZalm/Marlin Although Bernhard K seems to be the main person contributing right now? Anyway ErikZ is basically the primary source from and for the reprap community. The UM2 Marlin source code is in this repository: https://github.com/Ultimaker/SecretMarlin

This is very bizarre. This experiment where you printed slower and faster blows out all my theories (filament gets wider and skinnier, Z screw issues, slicer issues). Maybe you should repeat that experiment in reverse in case the problem had nothing to do with the speed and it was a coincidence. Anyway, assuming it really changes with speed the only thing I can think of is temperature. The time it takes to go from 210 to 190 is much too fast to make any difference in the pattern. The problem is most likely related to flow of plastic. It's extruding less in the skinny sections and more in the fat. Temperature cycles explain this perfectly. Print using the cura window instead of the Ulticontroller and look at the graph of the temp. Time the bumps/waves in the temperature graph and also time how much time it takes to print from one "bulge" on your part to the next. See if they roughly agree. Also to eliminate fan issues you might want to turn it off as an experiment. I would recalibrate the PID values. Actually I would get pronterface and figure out what the current values are. Pronterface is free. When it first connects to your UM it prints out lots of values including the PID values. You can see my values in this screen shot: Oscillation is very common in PID controllers. Larger values of P or I increase oscillations. D decreases. I wouldn't touch D though but it is quite safe to cut P or I (or both) in half. The only downside is it may take a little longer to heat up but probably not. Or you can issue the M303 to recalculate the values: M303 S210 210 is the temp to calcluate the values at. Default is 150C and 210C will give you slightly better results (not much different though).

Or did you mean the Ultimaker2? That's MUCH quieter.

Doesn't sound load to me. Maybe the Z axis - I guess everything about the Z axis is a bit loud. Did you put a drop of green grease on it? Anyway it sounds like mine. I think it is just that your house is much quieter than 3d printing show.

Thinking about this some more - it could be a bad solder joint. I would look at the full schematic and look at the pins with a loupe (or reading glasses). And if any look bad reheat them with a solder iron. Or just do it even if they look good. The full schematic is here: http://reprap.org/wiki/Ultimaker%27s_v1.5.7_PCB There is a zip file at the top. It contains the "brd" file which is the layout. Also the "sch" file which is the schematic. Both files can be opened by eagle software which is free: http://www.cadsoftusa.com/download-eagle/

90% chance it's the arduino then. There's almost zero chance it is the motor itself. I would spend the $20 on ebay and get a new arduino mega2560 (get 2 for $40!). Load it up with Cura before swapping it with the "bad" arduino. But if you are good with electronics you could instead look at this circuit - maybe check the voltage - see if it is a bit low. The signal should be around 0V for one direction, and 5V for the other. If the voltage is more like 3V then it's probably the arduino. A bad stepper driver can damage the arduino. There was a guy posting on this forum in the last few months who went through 3 arduinos until he finally replaced the stepper driver and then the *next* brand new arduino worked. The arduinos were permanently damaged (but only that one pin - X DIR for him). This shows X DIR signal. Your problem is with Y DIR which is similar.

Actually I think it *is* the Y axis. The *leaning* part was removed and replaced onto the bed after printing was over.

Zernu??? Where did you go? I did some bridging tests and posted photos and settings: Post # 17 (for now - unless/until someone(me?) deletes a post above it): http://umforum.ultimaker.com/index.php?/topic/1872-some-calibration-photographs/?p=25304 Temperature was by far the most important factor. I thought .1mm layers would be much worse than .2mm but it was fine! Just as good! This bridging could be better probably if Cura was more careful about not touching the previous bridge lines before they cool.

Bridge test. Only temperature seemed to make much difference. With the exception of the 230C print, all the drooping filament was caused when the hot end damaged an existing bridge string. In other words each string was laid down PERFECTLY. But then a few seconds later the hot end would mess up a previous string. Sliced in Cura 13.10. No special settings. Temperature seemed to be the critical factor. Cura slices the first layer along the length of the string and then prints normally after that: skin/walls first, then diagonal infill. Walls were at .8mm (two passes). stl is 50mm bridge here: http://www.thingiverse.com/thing:12925 As usual, this image is bigger than your monitor. To zoom first click the image, then right click and choose "view image" then click a third time for the FULL zoom. Pan around the image with scrollbars.

This is *very* common. You are slipping or losing steps in X (but not Y). There are several causes. It's probably slipping on a pulley but lets assume losing steps for a minute. Check to make sure if you push the head around in X and Y that it feels about the same. Consider putting one drop of sewing machine oil on each of the rods. Or use a thin oil. More likely a pulley is slipping. Tighten all the set screws. Consider marking each shaft and each pulley to see which one is the slipping one. Most likely it is the short belt which is a bit of a pain as you have to remove the X motor to get to that one. There are 6 pulleys for the X axis alone. SIX! Not 4. Make sure you tighten the set screws on all 6. Your UM probably came with a bag of extra set screws (the silver colored ones). The black ones that come with the pulleys aren't as good (although I'm still using the black ones). Also while it is printing look at the short belt to the X stepper. If it is twisting back and forth in different directions then that means it is touching the wood in the back. This is very easy to fix if you are using the plastic spacers - just add a washer under each of the spacers to move the motor slightly away from the wood.

These filaments are more likely to clog than normal PLA so they are telling you the minimum nozzle size is .4mm. Most people suggest going even larger - say .65mm. This is because there are particles of wood mixed in. The most common filaments by far are 1.75mm and 3mm (although 3mm is actually 2.85mm normally as 3mm would probably get stuck in the bowden feeder tube). Any PLA that is sold in one of those sizes is usually sold in the other size as well.

Ah!! That must be it! Turn off infill.

I've seen this. It seems like a bug to me. To do seperate skirt for each object you can right click on the object in cura and select "split object into parts". Hopefully that will do what it is supposed to do - make sure you view in xray view first as if there is anything red it will split that part up into more parts which would be bad. Then if you set gantry height to 0 in machine settings it will put a skirt around each object. You can see what will happen in advance in layer view (instead of having to print something before you find out it won't work).

Oh - pronterface is a free download. You can use the cura print window also - that works just as well.