gr5

I have heard people say that this can happen if the air is too dry. After many months. This is why I prefer my PLA *not* on spools. Or on larger spools.

Don't. This has nothing to do with "play" at this point. If it was play your outer dimensions would be just as bad as your inner dimensions. It has to do with what I call the "rubber band effect". This happens to everyone printing in PLA. I suspect it's not as bad with ABS because ABS hits glass temperature much sooner. With practice you don't have to design everything twice - you get really good at predicting the adjustment the first time. You are not the first person to report this. Most demand that Cura compensates. And maybe it should but the affect is complicated and changes if you change: fan speed, print temperature, bed temperature, color of PLA or brand of PLA. The affect is also a problem for example on gears - the inner part of the tooth - the gap between the teeth. For the same reason. But most people notice it on vertical holes.

I can definitely hear those frequencies. For example when the bed heater goes on and off it makes different noises. Strange noises.

I think you were filming slightly above where the problem is. The belt is rubbing against the wood near the motor I think. Try putting one washer under each of the 4 motor mount spacers to move the belt a tiny bit farther from the wood and the problem should go away. If I am correct your prints probably tilt in the X direction (each layer skips a few steps so that each layer is placed a bit to the side of the layer below). See 4th picture down on left side here and then click on it - the tilted gray robot: http://support.3dverkstan.se/article/23-a-visual-ultimaker-troubleshooting-guide Is that what you see when you print? Tilted prints?

You can set the minimum layer time to zero and then it won't slow down the small tops. But this will just cause other problems (printing on layers that aren't solid yet). I think instead the cool head lift feature should be improved such that instead of sitting over to the side waiting N seconds it should retract extremely slowly to compensate for the heating going on in the head. I suppose I should do a test some day first to see if this works.

Just in case we weren't clear - lower both jerk and acceleration by half or more and see what happens. This will be much gentler on your printer and won't slow down low-poly prints much (but will slow down high poly prints quite a bit). Another solution is to reduce the polygon count: http://www.shapeways.com/tutorials/polygon_reduction_with_meshlab

What firmware are you using? Marlin? Marlin has both acceleration settings and "jerk" settings so you might want to lower both. I'm going to guess you are printing things made with scanning or sculpting software which gives you millions of tiny line segments/polygons and that Cura's limiting resolution is a little bit finer than slic3rs resulting in many more tiny movements that Slic3r discarded but Cura hangs onto. I'm going to guess if you printed something low poly like a mechanical part you won't get these jerky motions.

Regarding your underextrusion the most common culprit seems to be the white teflon piece inside the print head so I recommend you take that out and slide some filament through it to see if you feel lots of friction and maybe drill it out or ask UM support for another one (they have newer ones made of glass impregnated teflon now) if your printer is more than a few months old. Regarding posting - you only need to do 3 posts to have all rights. You have 15. Could you go to that forum, grab the URL from the url bar above and message me with it? It might be that the topic was "closed" by a moderator. I googled dual extrusion in these forums and didn't see any closed topics but there are several topics and I may have missed it. Regarding "header wrap" this seems completely unnecessary and will make taking the nozzle out more difficult. Many people modify their UM only to put it back the way it was - the things work pretty damn well in stock configuration.

I haven't adjusted steps/mm on UM2 but I did it on UM Original - the procedure is simple. Take filament half way out (just pull with head at 90C) and then cut it off 200mm below the feeder. Measure the position of the end hanging down and then move the filament 100mm. Measure the actual movement with caliper. If it moves more than 100mm say 105mm then reduce the steps/mm by multiplying current value by 100/105. If it moves not enough, say 90mm then multiply by 100/90. Or you can do different distances other than 100mm.

Do you know where the existing model is of the existing head? I can link you to it if you want. Why don't you have a UM2 anymore?

I needed to make 500 bracelets/stretchlets so I just cut and paste the stretchlet code 25 times into one big gcode file and I rerun it again when it finishes.

Note that the 3rd fan is ALWAYS supposed to be on. It comes on even before the lights. The side fans can be tested by pretending to do a print - select PRINT, select anything, then go straight to TUNE menu and set fans to 100%. Usually the problem with the side fans (and rear fan) is with the cabling above the print head. You do not need to take anything apart - just slide up the plastic mesh covering as Illuminarti says.

Mine makes very strange noises. Keep it away from anything flammable. Maybe put it on something metal.

Next step is to modify the cad drawing to be the size opening you want. In other words if hole is 1.1mm too small increase in cad by 1.1mm. Your backlash may be gone but you will always have smaller holes than desired for 3 reasons but the main one here is that when the PLA extrudes it starts cooling instantly and is already shrinking and acting like a liquid rubber band and it pulls inward as it is laid down and ends up laid down inside of the desired (commanded) position. This is not a problem for the exterior because the layer below holds it better in that configuration.

Try increasing the temperature by another 10C (but I recommend not going any hotter than 240C) which will reduce the pressure in the nozzle. Also maybe reduce the spring pressure (or increase but more likely decrease) on the feeder.

This is a complicated subject. If you want to just know what to do well there's not much. 100% fan is key before you print any overhangs. Here's the long answer: http://umforum.ultimaker.com/index.php?/topic/4094-raised-edges/

They don't truly overlap - it's more complicated. I believe it will print from 75% to 150% of nozzle diameter so for .4mm nozzle that's .3mm to .6mm per pass. If you ask for a shell that is .7mm it does two .35mm passes. That means it positions the nozzle .35/2 mm (.175mm) inwards from the final finished edge and puts consecutive passes .35mm apart. Another way to think of this is that cura treats your nozzle as though it is a .35mm nozzle and extrudes only what a .35mm nozzle needs (underextrusion) and puts two passes .35 apart just like you would do with a .35mm nozzle. Even though in reality you have a .4mm nozzle. I can see how this would improve adhesion between the 2 shells but you might be better off just increasing the flow by 10% or so. Or fix the belts to remove any backlash or play. Or print slower or hotter because maybe you don't have backlash but instead have underextrusion (too much pressure in the nozzle - plastic can't get out fast enough). I always (almost always) set shell width to a multiple of nozzle diameter (e.g. .4, .8, 1.2, 1.6) as I get much better results that way (in my opinion). The only exception is for thin walls such as when I do text in which case I will often do .3mm shell. This comes out much better. I often simultaneously set the flow to 133%. But this is a special case.

Beware that flexible filaments are very difficult to print with bowden tube but you can do it. You will need to add oil into the bowden (apply to the filament) and you might have to print as slow as 10mm/sec to keep pressure extremely low on the filament. Or alternatively get a different feeder design printed.

Yes, if you set flow to 50% you will get half as much plastic extruded as normally needed. Or 200% will give you double. So for example if you are printing a solid layer and there are gaps between each line and the gap is 10% as wide as the line you can increase your flow to 110% and it should fill perfectly. I haven't changed my flow from 100% since about a year ago. 100% works almost always perfectly for me.

Vous devez niveler la plus proche de la buse. Je recommande à juste tourner les 3 boutons et ne pas utiliser la procédure.

I don't understand the question and what the speed has to do with it. The speed of all the motors change together on every (many) line segments. So they are constantly speeding up and slowing down. The faster you move the XY axis, the faster the E (Extruder ) axis moves. All XYZE moves are linear. To achieve linearity you have to run the E axis very slow compared to how fast is possible. The speed of the E axis is set by the speed of the other axis. So for example if you are moving the head at 100mm/sec with .2mm layers and a .4mm nozzle that is 100*.2*.4 or 8 cubic mm per second. So the E axis moves at the correct rate to extrude 8 cubic mm per second.

Not PLA? PLA is where UM shines the most I think. But ABS is fine also.

Is it the filament that is slipping or is the feeder turning backwards? Check that there isn't ground up PLA in the teeth of the mk8. More information here: http://umforum.ultimaker.com/index.php?/topic/6720-test-report-testing-different-feeder-wheel-grip-patterns/ You might want to stick with the original feeder sleeve.

And as UM grows your shipping department might some day get more sophisticated and ship things that cost less than 10 Euros through a cheaper shipping method. Or Blizz and Didier can combine their orders into one and save on shipping?

shadowfiend you are not clear at all. I dont' know if this is what you are saying but you should be able to put any of the 85 columns in there but should only include columns you need FOR THIS SPECIFIC PRINT JOB. It's a royal pain to check 85 columns times 5 rows when 95% of those values don't need to change. So if you don't care about print speed, leave it out of the table and use the global print speed setting. To know the column names it is simple - look at the ini file (file -> save profile...). That's how it works now - you can put almost any of those 85 parameters in the gcode.start or gcode.end and it will fill in with the appropriate Cura setting. Or the feature where you load the table can list the names of all the potential column headers.