Dim3nsioneer

In addition to the underextrusion I see your skirt lines do not touch each other. This is a typical sign of a to large distance between nozzle and printbed at the first layer. So maybe cross-checking the bed leveling with the famous sheet of paper might be a good idea. To eliminate the underextrusion, precise measuring of the filament diameter is a good starting point (e.g. with digital calipers). Finally it is recommended to have a multiple of the nozzle size as shell thickness (i.e. 0.8mm or 1.2mm).

This is getting interesting... :blink: First Q&A: The infill is printed with the same speed as the perimeters, with 100mm/s (I wrote 50mm/s in my original post by mistake). I haven't yet tried to deactivate combing, but it's on the list (see also below)... Just to make sure we're talking of the same side supposed to be (more) affected by shrink: If you print a cuboid, let's say x=10mm, y=30mm and z=15mm, which dimension would you expect to shrink (more)? The effect I see let would let the x-dimension decrease, but just in the upper 7.5mm... And it's getting even better: - I rotated the part by 180°C around z. The deplacement stays on the same side of the print, but is now on the opposite side concerning the printer (which cancels any theory about cooling from the left side only). - I mirrored the part along the x axis (the displacement occured on the minus-x-wall). After that, it occured still on the minus-x-wall (still the upper half). The plus-x-wall shows no sign of deplacement. - I rotated the part by 90°C (once clockwise, once counterclockwise). There was no deplacement at all! I have to mention that the rotated part were at dfferent places on the print bed than the original print. I'm currently thinking of some strange coincidence of retraction-fault (place-dependent?) and/or some kind of uni-directional backlash effect. I'll do some more tests such as compare the results when printing at different places on the bed, deactivation of combing, etc. I will also have a closer look at the extruder and Bowden tube as my filament has a thickness between 2.90mm and 2.97mm and therefore quite some friction in the Bowden tube (this might be the time to try the oil trick...) Using some other (thinner) filament might be worth a try too. Finally there should be a test on the other extruder as my Ultimaker can play in stereo... :smile: But further ideas are still welcome! Thanks so far!

Some time ago I found that my x and y rods could slip along there axis (one of them by about 2mm). Today I finally fixed this with an adjustable end cap (http://www.thingiverse.com/thing:54075). The play is now something like one or two tenth. And I found the 'z wobble' became much smaller! So it's maybe a good thing to check the rods in order to give the vibrations no chance to produce a displacement. I'll have a look if I find some 'before-after'-pictures...

Hi there Is there such a thing like geometry dependent displacement? I'm currently trying to print this interesting 3D puzzle: http://www.thingiverse.com/thing:23279 And this is the result of part 2 (printed twice, one of them rotated around z by 180°): While the overall print quality is nice (layer height 0.06mm), there is a very nasty displacement in x direction at about half height. It is at the height where the geometry changes significantly. There is a retraction and a movement between the smaller tower (blurred in front in picture 1, on the left in picture 2) and the rest of the structure during the upper half of the print. The movement is carried out with 150mm/s while print speed is 100 50mm/s (edited); temperature is 210°C (pure PLA). Top and bottom shell are 0.8mm, infill is 24%; infill overlap is 10%. My first thought was backlash. But the belts seem tight, the axes cannot slip (anymore). And if it's backlash one would expect the (retraction) gap to be smaller than supposed, but it's actually wider (there is also a very small displacement on the small tower outwards, but not as pronounced as the displacement on the larger part). The funny thing is, that the opposite wall doesn't show any effect at all! The same for the other direction (y). Next thing I checked was the model and the sliced code. The model shows no traces for any displacement. I even checked the x coordinates of the wall in question in the GCODE for different layers (lower half vs. upper half). They are identical to the second position after decimal point. I had such displacements in other prints at heights where the geometry was changing significantly (e.g. the Ultimaker robot; different layer sizes when printing the arms). But they were never so isolated and well pronounced. The most interesting thing is: When printing part 1 of the design (the same shap but mirrored), I don't have any displacement at all... I'm open for ideas and hints...they are highly welcome! EDIT: print speed was higher than written originally (corrected)

A very interesting thought! Brings us back to the question if the pattern is the result of a displacement or a thickness variation... My UM1 is behaving quite well at the moment concerning these wobbles. I'm using 10mm/s and 1500mm/s^2 for the z axis. I also find the wobbles much smaller at 0.06mm layer height compared to 0.1mm.

Wishlist for Christmas? :smile: Which Cura version are you actually referring to? I have the impression Cura 13.12-test allows for very small distances between objects (as long as the gantry height is set to zero). EDIT: Ok, after having tried it again with Cura 13.12-test I have to agree: the manual placement still is delicate...

I had it actually marked for reading a few days ago... :wink: After heaving read it now: Has someone played around with z-jerk in the meantime? Obviously to put it to something like 30mm/s is not a good idea. Where is the threshold where slipping at the beginning of the movement starts? It might be worth a try after Nick's findings with the higher z speed and acceleration. A very good question. If we would transform the transient movement into a power spectrum (by Fourier transform), we would get more amplitude at higher frequencies for the faster movement due to the steeper slopes. Attenuation might be higher for higher frequencies. However, I would expect this effect to be small except we are in resonance somewhere...

Some days ago I had a closer look to http://umforum.ultimaker.com/index.php?/topic/3695-acceleration-in-marlin/ with some very helpful contribution from Daid and gr5. I wonder what influence the z-jerk could have on these lines (has the z-jerk actually the same meaning in z as the xy-jerk in x/y?)

But only if the head is hot... in the (clean!) cold state which is usually used during the (first run of the) bed leveling wizard, it should be 0.1mm higher (that's the amount of thermal expansion between room temperature and around 200°C).

It's usually a trade-off. In this case Poka Yoke would mean to have different plugs for the different fans. That would certainly lead to more different parts for the device (not very attractive due to increased costs) but decreases the risk of an assembly error.

A very small retract at the start of the next skin layer would do the job (decrease pressure within the nozzle), wouldn't it? And the other way round: an additional amount of filament when starting the (fast) infill... But shouldn't that not better be implemented in Marlin? Some kind of latency would have to be specified for that in the Marlin code...

I recommend TweakAtZ 3.0.1 due to changed initial conditions (TweakAtZ 3.0 does not work without special comments in the start.gcode).

What is used for temperature control? Still a thermocouple with transmitter or a NTC?

I can only recommend Ultimaker to introduce Poka Yoke for the UM2 electronics...

The spike effect could be minimized with a low temperature (which you would anyway use for slow printing the last surface). And not all the PLA blends behave exactly the same... But you are so right... this was my major finding when fighting with the dual extruder...

As Christmas is close, I wrote a small present for all of you who wants to get rid of the combing lines on horizontal top surfaces. https://github.com/Dim3nsioneer/Cura-Plugins/raw/master/RetractCombing.py inserts retraction for G0 movements (combing movements are G0 movements) on a specified height. You can specify the minimum and maximum height (e.g. it makes sense to set 0.49mm as minimum and 0.51mm as maximum if you want to have the plugin to work on your 0.5mm layer). You can specify a minimum distance under which no retraction will occur (default set to 2.0mm). Furthermore, retract distance and velocity have to be specified (the only possibility for prints with otherwise no retraction). Optionally, you can also tell the plugin to lift the head (actually lower the z stage) during retraction. I declared the plugin as test version and used the same version labeling scheme as Daid for Cura (hopefully, nobody will be confused by this...) This plugin might also be combined with the TweakAtZ-plugin setting the speed to a lower value for the last layer...which gives especially nice prints...

Must be your magic hands then... :wink: I printed it again this morning. It came out ok (with the exception of the hole in the hat). So it was definitively something hardware-related. After having another look at the failed print, I think it might be caused by partial delamination as the print is significantly thinner at one end (and the material difference most probably distributed over the rest of the print).

I'm looking forward to these comparisons... it would be especially nice to know what kind of filament was used (material (e.g. PLA vs. PLA/PHA), origin of the material, measured filament thickness). Uncoupling the extruder from the frame might be a good idea to reduce vibrations...maybe this is a major difference between UM1 and UM2?

This issue might be worked-around by a Cura plugin...

That's exactly what I was up to... :grin: Thank you very much!

I can confirm that with the ovality of the filament. I think every filament has some? Actually, I get a larger diameter when measuring in radial direction compared to axial direction. I usually take the axial measurements as I think the bending around the reel might make the measurement quite arbitrary. Thanks for the hint with KISSlicer; as soon as I have set it up for the Ultimaker, I'll give it a try. What about the temperatures? Higher, lower or the same as for pure PLA?

Thanks for testing! I'll do another one myself. Maybe it was a big coincidence and a different reason... EDIT: How did you get the space between the hole in the hat and the hat outline filled? Seems to me a bit like a mission impossible with a 0.4mm nozzle...at original size.

Hi there I'm having quality issues with Colorfabb signal red PLA/PHA. I get a lot of knots and clearly visible layer structures when printing this material. This is a typical example: This print was made with 50mm/s, 220°C for the first layer only, then 205°C until the velocity dropped under 25mm/s due to minimum layer time from which I used 195°C. The filament diameter was measured and set to 2.85mm. The layer height was 0.1mm. I do not have these issues with standard PLA. The very same settings result in a very convincing quality with pure PLA. Any hints how the quality could be improved?

Interesting. Are you not afraid of getting a burr on the nozzle inside?

Unfortunately, I was not watching it as my Ultimaker behaves usually... :eek: EDIT: I turn on my fan on layer 2 (at 100%) and let it run until the end of the print... but I'll keep an eye on that temperature during the next attempts.