gr5

You turned on spiralize! Turn that off. I've never used that. That is for only very very specific prints and only works for very very limited shapes and it turns off infill (overrides lots of settings). It will also make your nozzle over extrude by 2X if you ask for .8mm wall because it only does one pass. It's only to be used for cups and vases and only certain shaped ones - For example it will not work for a coffee mug due to the handle. Spiralize is of extremely limited use.

It's the red/yellow/black twisted wire that's the problem. That is your temperature sensor. It's not going through the "F" shaped black strain relief. By now it is too late and there is an open (likely) in one of those 3 wires near the connector (or inside the connector). The simplest thing you can do is swap it with the other unused cable hanging there and this time put it through the F connector. Of course you need to find the other end of these 2 wires underneath your UM and swap them there also. Cura generates a file with the extension ".gcode". Find the file. Open it and look at it. Go here and read about the 5 or so codes found in that file. Read the comments (e.g. layer 1, skirt, etc). Pay particular attention (or search for) the letter Z meaning height. E is extruder. Here are explanations for all the gcodes although only 5 or 10 are actually used to make a print: gcodes for Marlin (and a few other firmwares) great reference: http://reprap.org/wiki/G-code scroll way down to the gcodes on this page: https://github.com/ErikZalm/Marlin/ I plan to one day write up a detailed explanation for how to continue a print. Unfortunately my instructions above are all I can offer now. I need to spend a few hours making screen shots and explaining every little detail even though most of it is very simple to do. The instructions need to work for UM1 and UM2 and for people with or without an ulticontroller and so on. They need to work even if the part is in the way of being able to home (there are more tricks if you don't have room to home without hitting the part). But not today. Sorry.

I also have UM Orange so that's a good choice.

Are you asking what temperature most people print at? Or are you asking about the part itself? It's a PT100 RTD. That means it is 100 ohms at 0C. Here is a table: http://www.intech.co.nz/products/temperature/typert.html

well - or the way the filament is mounted and fed to the extruder maybe? Is that really the extruder that is the issue or the way the filament reaches the extruder? Although your design with the open extruder bottom seems to blame the problem and fix on the extruder. I love that idea by the way. It's just semantics I suppose.

If you do the above test nico I will repeat it on my machine. Publish the temp and speed used in pronterface and please use the UM light blue filament as we both seem to have that. Do you have micrometer accurate to .01mm? I think I have one somewhere that screws.

The photos are very useful but ONLY if you add temperature. Could you please add the temperature you were printing at? If you changed temp throughout the print you could maybe post the final temp where it failed or show with arrow which layer it was at which temp. Your post showing temperature and max print speed is very useful, thanks! I will be referring to it quite a bit! 10mm^3/sec is what I consider very fast. This test is designed to fail. The person who created this test probably didn't expect the printer to reach above 5mm^3/sec. I don't think the UM Original can do much better - many people can't print 10mm^3/sec on the UM Original.

I believe in one of the 20 pages above this post, Sander said that they got more orders than expected and had to double the staff that assembles the UM2. Then in the last month or so I think they hired more and are now at triple the original UM2 capacity they had back in the fall. As far as money to order parts - I doubt that is an issue because they apparently take your money as soon as your order (not when they ship). The bigger issue I believe is interviewing, searching, hiring, and training staff. It's hard to hire good people and it takes a lot of time that you could be spending on building ultimakers, lol.

That's fine. Really every print has different requirements if you want to want to be fast and efficient. I find that 4.5mm is not enough distance for a stock UM - typically you need to either increase that to around 6mm or add this simple part: http://www.thingiverse.com/thing:46157 40mm/sec is fine. If you have the old firmware (with a bug in retraction speed) this will retract at instead 20mm/sec which is a reasonable speed and if you have the latest firmware it caps at 25mm/sec. So 40mm/sec is recommended retraction speed for UM Original. 35mm/sec for UM2. Cura usually does not retract on layer change. This results usually in what some people call a "Z seam" especially if you are printing at high speeds (like 70-150mm/sec and .2mm layers) because it stops while changing Z height and you still have high pressure in nozzle and you get a tiny blob. Usually in the same spot for many consecutive layers. You can improve the Z seam by doubling the Z acceleration that comes default with Marlin. Don't touch the speed - just the accelearation. You will get less noticable Z seam.

Set the temp to 95C and then touch water to the nozzle and make sure it doesn't boil. Push the head around to the 4 corners and check to see if the temp jumps by more than 5C suddenly. Now set the temp to 105C and make sure it *does* boil water. Again push the head around to the 4 corners and make sure the temp doesn't jump suddenly. Hopefully this test will fail and now you have something more direct to debug. The art of debugging is to first get it to fail repeatedly. Then try to get it to work by doing something that is consistent, then repeat the cycle until you understand the failure and a way to fix it. If the above test fails I would turn off the printer, put it on it's side and remove the head temp cable and hook it to an ohm meter. Look up the correct resistances for a PT100 RTD and jiggle the wiring to see if the resistance jumps around. This table is good - room temp is around 108 ohms: http://www.intech.co.nz/products/temperature/typert.html

Not having a picture is really really hard for us. So you mean kind of like printing a box? But round? Where the left and right sides of the box print fine? But the near and far sides don't print at all? That usually happens if there is something in the model that is wrong - maybe the thickness of the cylinder varies slightly and the slicer gets confused.

The print head doesn't move as fast because it has less space to accelerate and slows down at the end of each line allowing the pressure to reduce just enough to avoid slipping. Is the layer height and speed and temperature of the UM Robot identical to this part?

These are fine settings for the second layer and higher, but the bottom layer defaults to extra thick. And it should or you will have other trouble. You want .2mm or .3mm on the bottom layer - you should be able to get away with .2 but if your levelling is a little off, then .3mm is more forgiving. What is the "bottom layer speed" on advanced tab? And under quality what is the initial layer thickness? You are going to probably need to raise the temp to 240C for the first layer. There appears to be no "islands" (gaps between sections of the print) so you don't appear to need about stringing so if it were me I would print the whole thing at 240C but if you want nice quality lower the temp back to 220c or 230C after that first thick layer is done. .06 is damn thin. It's going to take a lot longer at that thickness. Printing .2mm layers is a lot of plastic going through the nozzle so you need high pressures that the extruder just can't quite achieve. Raising the temp makes the plastic flow more like honey than toothpaste so you don't need so much pressure from the extruder. If it still skips, just go to the tune menu and lower the speed % from 100 down to 50 or lower until the extruder stops skipping.

You are talking about the error caused by minimal steps (max resolution of Z axis) but the resolution is so high that it doesn't matter. I looked into this about a year ago. .1 and .2 are fine. Don't worry about this - it's silly. Well - I usually do 150mm/sec because that's plenty fast for me and the non-printing moves tend to be rare - the long moves anyway. On the short moves it never gets up to 150mm/sec. Some people do 250mm/sec. I think the max speed is 300mm/sec. Before going that speed you should test the repeatability using prontrface which is free, and easy to use GUI, and you can do lots of 300mm/sec moves back and forth and see if you lose any steps. Probably you will be fine: http://koti.kapsi.fi/~kliment/printrun/ Doing 80mm/sec for fill and 50mm/sec shell will lower the quality - speed change will result in over extrusion when it slows down. A little bit. But if you don't need "perfect" quality, and if you don't care too much about stringing, I can print at .2mm layers, 100mm/sec or maybe even 200mm/sec if you heat up the PLA to 240C. Here are some photos that explore the relationship between speed, temperature, quality: http://umforum.ultimaker.com/index.php?/topic/1872-some-calibration-photographs/ In general, "perfect" quality is around 30mm/sec. But you can do fine at 100mm/sec if you aren't picky. Another huge speed improvement I have found with zero quality reduction is to not print any infill. The UM is amazing how good it can bridge openings - for example the top layer of a cube with no infill. This means however that you need the top/bottom thickness to be a layer or 2 thicker than if you use infill. Usually 4 layers is plenty. The latest Cura (14.02) is now very accurate about predicting how long a print will take. So you can fiddle with more top layers and less infill and see how much time you will save. The strength of a part comes mainly from the outer edge - that's why bones are hollow and the wood beams in a house can be drilled only in the center (for plumbing/electrical) without sacrificing strength, etc. Speeding up the Z axis won't print any faster but will print with better quality. I doubled the default Z acceleration and it changes layers twice as fast. Didn't need to touch the max Z velocity. This helps the Z seam quite a bit.

Two things: 1) I ABSOLUTELY LOVE Robert's idea of having the filament touch *nothing* before getting to the metal extruder wheel. This could eliminate the wear of tear of getting ABS powder mixed in with your PLA print (the current extruder is made from ABS). Alternatively it could touch only a tube made from the same material as the Bowden as this material is proven to be tough even when rubbed with chewed up filament. 2) No one has mentioned the nut that pushes on the extruder motor. There is a nut that holds the panels of the UM together underneath the extruder motor and it tries to tilt the stepper. Keeping the stepper flat/flush requires more force than normally would be needed. It would be nice if that could somehow be addressed also in the extruder design - maybe make the right 2 screws that hold the stepper be not part of the extruder housing and instead be underneath it. That way you can also remove the extruder housing without the stepper falling. I like many of the ideas stated. A lot! But will refrain from restating them over and over again as many people are tempted to do - I just couldn't resist repeating my point 1 above. Sorry. I don't think we should restrict ourselves to existing extruder parts as long as they are very common parts and easily ordered in Europe and USA.

I agree that we may be tackling the "wrong" issue for some people but I think in this posting we should concentrate on improving the feeder. There is another thread looking for the root cause of some people's underextrusion. There are plenty of improvements to be made on this feeder and they certainly won't hurt the UM2.

Also be aware that if you set your speed over 100% and some of the layers hit the "5 sec minimum layer time" or whatever your layer time is, then you will be overriding this minimum layer time. For example if you had it set for 5 seconds and you print at 200% then the quickest layers will be printed closer to 2.5 seconds and they might not have time to cool before the next layer is printed.

A little background on speed. The speed is called the "feed rate" which is not a 3d printing term but a generic CNC gcode term. The feedrate gcode is controlled with the letter "F". The limiting speed on the UM is mostly X and Y. Most moves do not include a Z component. So on normal extruding moves (not bottom layer, not limited by minimum layer time), the feedrate is set to whatever mm/sec you set in Cura. Once set, Cura doesn't mention feedrate in the gcode until it changes. When you do a Z move, the feedrate is specified because 100mm/sec in Z is too fast (on the UM1 anyway) and if you do extrude-only moves such as retraction, the feedrate is specified (but not ulti gcode mode) also because 100mm/sec in E axis (extruder) is much too fast. Now on the controller for UM1 and UM2, if you change the "speed" you are multiplying by a percentage. So by default is 100%. If your feedrate in Cura was say 60mm/sec then speed=100 is 60mm/sec. If you increase feedrate to 200% you will be printing at 120mm/sec. 10% will be 6mm/sec. If your non extruding moves were 150mm/sec and feedrate is 200% then the non-extruding moves will be at 300mm/sec. However... There is a max speed and max acceleration for all 4 axes. For example, typical XY max speed is 300mm/sec. If setting "speed" or FR (feedrate) to 200% goes over these max values then it will only increase up to the max speed. Changing the FR to 200% does not affect the max speeds or the accelerations. So printing at 200% will never print twice as fast. If you were printing slow enough then it will be almost twice as fast but never quite. Also Marlin continues to be very careful to never violate jerk or acceleration settings and always move in a straight line (linearly). So if the extruder only needs to move 1mm and the X axis is moving 200mm, the limiting speed for the extruder will be set by the acceleration, max speed, etc of the X axis. Changing flow in effect changes the steps/mm of the extruder (without violating the existing max speed of the extruder). Basically Marlin implements all this just the way one would want it to work.

It's in the painting section. Painters use it especially on corners and edges to keep paint off of something. The blue tape is what most people use. Green (frog tape) works okay also I have heard. I also hear white masking tape works but that has a different kind of glue that might be more difficult to clean off the glass. Get as wide as you can. I use 75mm wide. Illuminarti found some 150mm wide! Wide is good for 2 reasons: 1) If lifting on a tiny surface, you want as wide as possible tape as possible to hold down. Skinny tape lifts too easily. 2) It takes less time to apply the tape carefully (getting the edge of one tape perfect with the edge of the next tape). MOST IMPORTANT, clean the blue tape with isopropyl alcohol before printing on it. This removes waxy surface. This alcohol can be found wherever bandages are sold:

I just read it. Nice. Although people can make 2 perpendicular lines using a square. Anyway - thanks for doing this!

Okay I just looked at your picture, nico and calibrated it using both the nozzle outer diameter of 6mm and opening of 8mm. I get that your nozzle hole is smaller than mine by 30% to 60% area. Doing identical comparison, metal to metal, I get that my hole diameter is .49mm and yours .38mm (that's where the edge curves inward into the hole). Comparing the blue filament in my hole - blue edge to edge - comparing that to your nozzle hold metal to metal I get that my hole diameter is .43mm. Either way my nozzle hole seems larger than yours by a significant amount (at least 30%). This technique of course could use some accuracy improvement! It's hard to say exactly where the edge of the hole is. I tried the same technique for both photos, but differences in lighting can change where it looks like the edge is. That photo that showed light coming through the nozzle hole would be the most accurate if one could photograph a ruler at the exact same distance (preferably in the same photo!). I measured 4 different directions (horizontal, vertical and 2 diagonals). If I use the SMALLEST of my 4 "metal to metal" measurements of my nozzle hole, and the LARGEST of the 4 "metal to metal" openings of your nozzle hole I still get .46mm for my nozzle versus .40mm for your nozzle. Which in area is still 30% difference.

Absolutely not. You are confusing UM with Makerbot. They are just a very small company and are very busy. This is something they are looking into seriously. Part of the problem is that it's not clear what the problem is or if there is a problem. I know they are redesigning the extruder but in the end they might decide the current design is better (I doubt it). UM is the kind of company that would offer free extruder upgrades to all owners of the UM2 even though they have no legal obligation to do so. But only if they find there is a serious problem with it. It may be the solution has nothing to do with the extruder. I have 5 good theories for what Nicolinux's problem is. Also it's not clear if 1% of the people out there have extrusion problems or 60%.

It's pretty easy to disassemble - only 4 screws - but then the motor wants to fall so I guess there is a 5th screw to remove the motor cover access. The spring is too strong so everyone uses the loosest setting. I think UM will come out with a weaker spring so there is more control but right now the loosest setting for the extruder is about right. The coolest thing about the UM2 extruder is that the firmware (and gcodes) can control the current. So you can set the current to a threshold below where the extruder grinds the filament. So what happens is if you have too high a pressure, the extruder skips steps backwards instead of grinding the filament. On the downside - it's weaker than the UM extruder. And some people are blaming it for underextrusion issues (completely convinced!) but to me it's not clear where these people's underextrusion issues are coming from so the jury is still out. Also it is made of ABS and gets worn slowly over weeks of use and very tiny pieces of ABS can get into the nozzle which can also cause problems. Maybe. No one knows.

You need to post a video I think.

I disagree! that photo is better than mine. Please: 1) Turn UM2 off, put it on it's "back". 2) Slide nozzle to the top (the front). 3) Tape a caliper to the print head so it is as close as possible to the nozzle. 4) Photograph the same thing but with both ruler and nozzle in the same photo.