gr5

The bowden comes off. Read about how to do this properly. This thread is messed up because the earlier poster is asking about a UM2 and the newer poster is asking about a UM3.

STL files have no units and cura assumes it is in mm. Autodesk I think defaults to inches when you save - somewhere in the settings you can specify the units for STL files - set that to mm. Or maybe it's saving in meters. Anyway set that to mm.

yes.

All ultimaker printers use "3mm" filament which is actually 2.85mm. The um2 with the plus upgrade (around 500 euros?) is identical to the UM2+. You don't need the um2+ to print nylon but you probably will want to print with a 0.8mm nozzle. Some of the older UM2s didn't come with an olsson block so you are stuck with the 0.4mm nozzle only (but getting an olsson block is cheap (< 100 euros) and easy). I only say you may want the larger nozzle because it prints so much faster and your prints are big enough that they may take many days without the larger nozzle. I don't know if all new UM2s come with Olsson block - I think they do. The Olsson block allows you to change nozzles. By the way, printing with nylon is much harder than with PLA so there will be a learning curve. For one thing you need to keep it extremely dry - like 10% humidity which is very hard to do so basically you will have to bake the nylon (for example on the bed set to 100C covered in a box or with a blanket) and if you have print jobs more than an hour (almost every print job) then you may even need to do more drastic things (like I do - not sure it's 100% necessary) -- I like to keep the whole spool in a ziplock with a large - freshly reset dessicant - with only a small hole to let the spool out (to keep humidity out). It's pretty easy once you've done it once. It's harder to get Nylon to stick to the glass - I recommend taulman bridge nylon. Or ultimaker nylon. Also you should cover the front of the printer. Just a plastic bag or sandwich wrap will do. I also like to cover the top but I don't think it's necessary with nylon and set the fan to a low speed (around 30%) or your layer bonding on the nylon will be weak and parts will break at layer lines. Anyway try printing nylon and post in here when you start having your first issues.

It's possible to get rid of the stringing completely by lowering the speed to 30 (for all speeds - we don't want outer shell down to 9mm!) and lowering the temperature another 10C. But I don't think it's worth it. Those strings you have are so tiny and whispy they are easy to remove. If you practice (practice with failed prints) putting them near heat you can make the whispy stuff melt into themselves and disappear or just scrape them off. Bad strings are horizontal. Whispy strings are vertical like yours.

It's not enough to watch the first layer. You have to check that it squished hard. These are photos of 0.3mm thick first layers and how they look with the head starting off at different heights - nominal would be 0.3mm above the bed which doesn't squish hardly at all. 0mm above the bed will give you the most squish and the most sticking. 0.15 is about right for me but if it will be an unattended-multi-hour print I squish more than that.

It's NOT normal. It's just a common problem because people don't know how to get their parts to stick well. I did experiments with many variables (temperatures, glues, glue thicknesses, squish amounts). Not everyone should have to do that. That's why I made the video - to help people understand how to get the parts to stick like hell. The steppers should skip before the part comes off the bed. The part should never come off the bed during printing. Even with 2000 hours of prints. It should be that firmly attached. In particular with PETG I go a little too far sometimes and pieces of glass end up inside the bottom of my print. The key missing piece is usually to squish your first layer more into the glass.

You can remove those clear tubes - easy to do. They are called "bowden tubes". First realize that the two ends are different so don't put the tube in backwards when done. The end on the feeder has an internal taper - like a funnel. To remove first remove the little white clip. then push down firmly on the white ring (collet) - this is best done with a tool such as pliers or screwdriver. While pushing down firmly on the collet lift the tube out. To put it back together push the tube back in and while holding the tube down slide the clip back under the collet. By the 30th time you do this you can do it with your eyes closed.

That works also, yes although I guess we didn't need to add execution to group as well.

This really sucks, man. to fix it you need a heat gun. Hopefully it's PLA as if it's a high temp material you are even more screwed. This has happened to many people and they get through it eventually. Heat the cores to 150C if they'll let you. If one is failing then only heat the other. But mostly you need to get this all out with a heat gun. The root cause is that your part didn't stick well to the glass and was wider than tall so it didn't fall over and instead got carried around the bed like a hockey puck while extruding into a huge blob. The fix is to get your parts to stick better. Pay a lot of attention to how to make your parts stick better and check to make the first layer is squished well into the glass. It's important to use a very very very thin layer of glue and important to do squish. This is talked about in more details here. Think about how much time you waste cleaning this head and how much watching this entire 15 minute video could have saved you. In other words - it's worth watching the whole thing.

It's probably standard if you think. It should be metal (not wood) and that right there is a strong indication that it's from UM. It should have a glass plate with 4 clips in the 4 corners. It should have an extra circuit board underneath the printer that is smaller - maybe 2 inches on a side? I guess the biggest indicator would be the power supply which should be a "Meanwell" and should be 221 Watts. Well if your board is fried then I guess you have to replace it. Well your bed already has a PT100 temp sensor so you only need a PT100 temp sensor for the head. Hopefully that's the only change. I think you can use the existing temp sensor but I'm not sure - it seems to me you'd have to plug it in a different spot than where the PT100 gets plugged in. @tinkergnome - Do you know if the UM2 board lets you plug a UMO temp sensor into a UM2 PCB temp input and just change the firmware? The UMO outputs a voltage where 0V = 0C and 5V = 500C and everything in between is linear. But the UM2 uses an opamp in some way to measure the PT100 but I don't remember what exactly it does. Another choice Gimball - is to replace the temp sensor with a PT100 but I forget what the diameter is for the hole where the sensor goes. It will be an exact mm like 3mm or 4mm. I sell temp sensors that are 3mm in diameter but I suspect the hole in the UMO print head is different. You can get PT100 temp sensors on the internet for other dimensions. There may be other differences between UMO and UM2. The UMO has a bottom fan - you can discard that. The UMO has one side fan expecting I think 12V. The UM2 has 2 fans with 24V among them and they are wired in series so each fan gets 12V - you can't do this with just any fan - if the fan has a computer controller built in (you'd be surprised) then don't use it.' The steppers might be off a bit (steps/mm). X,Y axis will be the same but Z and E will be a bit different. But that's very easy to fix. Actually you have the HBK so Z will be the same also. Oh and the limit switches are different - the UM2 only uses 3 limit switches instead of 6. It's a better way to do things. So you only have to wire up 3 of them (left, rear, and bottom). You should be able to get a UMO PCB equivalent real cheap from aliexpress or ebay. That might be easier.

Absolutely - you need to move it and give it a negative Z value such as -0.5mm. Click on the part and select the move icon on the left edge (towards top). HOWEVER, by default it will not let you change Z away from zero. To do that go into cura preferences (this is a cura-wide setting. Not a profile setting) and there is a checkbox to uncheck - something like "always drop parts to the glass" or "disable z movement" or something. It's kind of obvious I think. And there are only about 10 options in there. I think.

Okay so I just looked at your bench photos. In the hull there are two slits - those look like underextrusion - either the Z moved too far for those layers or a filament tangle or... not sure what. The diagonal lines are a known issue called "zebra stripes" in this forum. Google "site:ultimaker.com zebra stripes" (the first part limits the search to this website) and you can read many pages of information - the problem was figured out 100% and solved with two different possible hardware mods. I sell one of them (tl-smoother) but I don't recommend you modify - for me personally I rarely print things that the zebra stripes show up on. So it's not a big deal for me personally. They are caused because every 16th step doesn't move far enough (steps 15 and 16 are weaker magnetic fields for the stepper). They are most obvious on flatt(ish) surfaces that are *not* parallel to any of the 3 planes (z,y,z) yet are *almost* parallel. Bench's walls are neither plumb vertically (thay slant outward away from the center of the boat) nor are they parallel to any other axis ( put another way the boat house is wider at the top and front of the boat). Now you have a 3rd issue called "ringing". This is when you go around a corner and you see little waves of that pattern emanating next to that. The most common case of this is text on a flat surface. It has to do with the rubber belts vibrating when one axis stops suddenly while the other axis is now moving. You can tell because the ringing fades out after a cm or so. This issue isn't as bad on a UM2 as the UM3 head is heavier. It can be eliminated by lowering the acceleration (and this is done in cura but you can lower it even more - 1000 mm/sec/sec should be enough - at least on a UM2 that removes it completely - maybe google accelration and ringing tests or something (include site:ultimaker.com). Both of these later two problems are most visible on dark shiny filaments as you only see this because the shiny filament is reflecting specular light (direct reflection). Optical reflections can show off tiny imperfections normally impossible to see especially if there is a repeating pattern. Flat (matte) filaments like woodfill and CF fill hide this quite a bit. There aren't a lot of matte filaments out there as PLA tends to be very shiny but there are some. I like the protopasta ones. I'm not sure but I think all there HT PLA are matte.

All I know is "griffin" is the name of the python code folder for the firmware on the UM3. I think "Griffin" was the code name for the UM3 before they went through their laborious 2 month marketing process of picking a name (just joking about the process). Engineering knows to never assume what the name of the next product will be.

Does your UMO have the standard official HBK (heated bed kit) upgrade or a home-grown upgrade?

UMO is a great printer - no need for UM2 mother board. Try printing with it first. The UM2 motherboard expects PT100 sensors in both the heated bed and in the print head. You can use your existing sensors but you will need to create your own custom version of Marlin. Does your UMO come with a spare extruder driver? I forget. The UM2 PCBs shipped by UM - some of them don't have the second extruder chip loaded. I think all of the more recent ones - not sure. Without that spare extruder driver you can't update to the marc2. I'm not sure if there is anything particularly useful on the UM2 mother board that isn't on the UMO. I can't think of anything other than it controls a different display (so you have to get the display board also - well you don't have to - I think you can compile marlin for the old display - I'm sure you can - just get the UMO+ firmware). Really I think the biggest difference is that you don't need the fan anymore and it is compatible with PT100 sensors which require an extra tiny bit of circuitry.

Different types of aluminum have many properties but any type of aluminum will work here. For example there is a type of aluminum that is extra flat and won't warp (MIC 6?). That is not important here as the leveling screws will compensate. Really any aluminum alloy will be fine.

Okay - well that sounds like you don't know linux. Hmm. I could recommend you take a 30 minutes course on linux, folder structures, file manipulation. Hmm.... Well when you start the terminal you have to navigate to the folder where you downloaded to. Type "pwd" to see where you are. Type "cd " followed by a folder name to move down into that folder and then pwd to see where you are now. Type "ls" to list what files and folders are in your curent working directory (pwd = print current working directory). Navigate to where cura is. Most likely that is with "cd Downloads". It's case sensitive so "downloads" won't work. At some point you can type "ls" and you can see Cura is in there. Note that if you type the first few letters of a filename you can hit the tab key and it will auto fill the rest of the file name. "cd ~" brings you back to your primary folder. "cd .." moves you up to the parent folder. This interface may seem horrible to you but it beats the mouse 100X faster if you use it a lot. Once you've located cura with "ls" and you are in the right folder THEN do the commands I showed above but just start with "Cur" and hit tab to let the rest autofill. Same with the "./" command.

Plug usb cable into windows machine. It should make the same "ba bump" musical sound that it makes if you plut in a mouse into the windows machine. If not I think you have hardware problems (with the printer). If it makes the sound then go to device manager, un plug and replug and watch what appears - in the first second it should show up in the usb section then if you have the right arduino windows driver installed it will dissappear from here and appear in the list of COMM ports aka serial ports. If it does then everything is working - not which com port e.g. COM3 or COM7 or even COM20 or whatever. Now try to connect from cura - if it won't work it is probably a non-standard baud rate so don't use cura - use pronterface - which is better than cura and is also free here: http://koti.kapsi.fi/~kliment/printrun/

Now that I see the above post by kman, I'm pretty sure meshlab will fix this. You want to remesh it. This is a pretty rare problem fortunately. STL files contain lots of triangles and nothing else. No data to indicate which triangles go with which - like a big puzzle. All 3 points of each triangle are in XYZ space but sometimes... Anyway after intersecting a plane with all these triangles now cura is left with a bunch of "random" lines which are now in a plane in XY space. Each line segment as two points - two ends - but which line goes with which? Looking at the above photos - where some of these lines come together they are less than 0.01mm apart. The next step in cura is to take these "random" lines and connect them into loops. For example you have 8 paddles so a slice might have 8 loops. Ideally. But there is some error allowed so sometimes cura connects them up in the wrong order and the lines go backwards/forwards and you get twists in the loops. Cura got confused in a few places in this model and cura got some loops twisted up a bit. Not sure if polygon reduction is quite exactly what you want but here is a guide to that feature anyway: http://www.shapeways.com/tutorials/polygon_reduction_with_meshlab @Daid @nallath @ahoeben @ghostkeeper - any comments/suggestions/corrections?

I've never used auto leveling for a few reasons but the way auto leveling works is there is a sensor in the bottom of the print head that detects the distance (capacitance) (very inaccurately which is fine) to the metal below. it can detect tiny changes in distance/capacitance. When the nozzle hits the glass and it keeps moving it realizes that the capacitance hasn't changed it knows that it already hit the bed. It's probably very fast so in milliseconds it can go back up a bit and recheck - not sure - I've never tried it. The point it it's supposed to push into the glass and then push a little farther. This is a feature not a bug. It's also I think why the rear center screw is looser than the other 2. The other two rely on "z wobble" so are okay with the stronger springs.

Oh by the way - so the problem is worse near the heated bed but the heated bed is so amazingly helpful for other reasons it's worth it. But if ALL your parts are smaller than 1cm you could lower the heated bed to 45C. Or even go with blue tape instead (but if you use blue tape you have to wash it first with isopropyl alcohol). Also if you print cooler the problem isn't as bad. I've printed PLA as low as 180C but you have to print much slower as it's more like peanut butter than honey at that temperature. I usually print at 200C on UM3.

This is very common. vertical holes too small. It's much worse with round holes than square holes - with square holes typically it's only the corners that are a problem and typically the sides are fine. 0.5mm shrunk diameter is typical although I usually see closer to 0.3 shrinkage and I typically add 0.4mm to my vertical holes in CAD as a first estimate. To answer your question - this is because PLA cools quite quickly (in milliseconds) as you are printing and so it shrinks a bit so it's under tension - PLA sticks to itself when liquid - like snot - like mucus. This is a very very useful property (try printing with PVA - horrible in comparison) but when you are printing it's like a liquid rubber band and when printing circles it pulls inward or if squares it pulls inward mostly on the corners. There are other reasons why holes are too small - CAD software doesn't output circles but instead polygons with each point on the circle and that makes the polygon *smaller*. This is typically only an additional 0.1mm so not as significant (depends on how many sides to your polygons). Also PLA shrinks by about 0.3% - also not as significant as in your example .3% of 1cm is .03mm.

Very easy and quick to do @korneel - just click and drag the text you want to quote and a popup shows up that says "quote this". Click on that and it adds it to your already in progress reply. I can't help you if you have a phone - phones have horrible user interfaces (in my opinion).

And about the tiny bits of pva ending up in your print and causing holes in the walls - yeah I get that also. A prime tower can help but if you try a prime tower make it at least 1/5 as wide as it is tall (skinny ones fall over) and put it in the back center of the glass - not in the corner (corner towers tend to fall over because the glass is almost always lower in the rear two corners.