gr5

@Achim - to post a picture click "gallery" on the top left of this page, then click the blue upload button. When done uploading start a new post and click "my media" next to the smile face. Only once per layer? Could that be when the Z axis moved? Are most of the "blobs" mostly on the same side of the object? If so it's probably the Z seam. Regardless what caused the hesitation blobs, you can improve those by printing slower because the difference in filament flow from moving to stopped is smaller so the blobs are smaller. In other words half speed will give you half sized blobs.

I added a "issue" to marlin here: https://github.com/Ultimaker/Ultimaker2Marlin/issues/23 I doubt the retraction bug exists after looking quickly at the documentation as those are supposed to change the feedrate and then change it back. I think. Hell I don't know.

Sounds like a Marlin bug for both issues. Marlin should set flow back to 100% when changing filament (incidentally I don't use this anymore - I just pull it out by hand and shove in the new one - Robert taught me that trick). Bug #2 is the G codes for retraction (G11, G10 or something like that) should take into account the current flow.

I'm a little confused - when you change filament it likes to go to the front so that the bowden isn't bent so tight. This is fine and no one else complained about this. Where did you *want* the head to go during filament change?

You can change these settings easily with gcodes. But if an axis is off by 10%? What the hell? You should double check this: 1) Get pronterface: http://koti.kapsi.fi/~kliment/printrun/ 2) Connect to UM2 with USB cable and connect with pronterface. Move the bad axis around with gcodes: G0 Y0 -- this goes to location Y=0 G0 Y200 -- this goes to location Y=200 Now measure the actual distance with ruler. Is it off by 20mm? Really? That's extreme! Maybe you have the wrong model stepper model - make sure both steppers are the same part number (x and y) or one of your pulleys is 10% too big or too small. There are no other sources for error. Anyway you can adjust your steps/mm with gcodes: 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/ Pay particular attention to M500.

Are you printing PLA or ABS?

I think you should print this test print to see what your cubic print speed is at 230C. 230C is strongly recommended for this test print. At 230C you should be able to get to 7mm^3. At 200C maybe 2mm^3 (not sure - I have a graph somewhere on this forum): http://umforum.ultimaker.com/index.php?/topic/4586-can-your-um2-printer-achieve-10mm3s-test-it-here/ Do the above test so we can see if you have a fully functional UM2 or if instead you have some issue.

In cura, the first layer defaults to .3mm. This is good because... When you extrude the first layer if you want it to not over or underextrude by more than 20% then you need to level to an accuracy of the first layer thickness times 20% or .06mm accuracy. This is about half the width of paper and difficult to do. If now you make the first layer .1mm it's extremely hard to set the first layer leveling that perfect. However, many people make the first layer .2mm or even .1mm and spend an hour leveling by trial and error until it is perfect. Anyway - the point is that is a Cura setting. The slower first layer is also a Cura setting. This is a good feature because it allows the PLA to flow better and stick better to the glass so you get a nice smooth first layer. Alternatively you can print the first layer hotter. But for quality, nothing beats slow. For an iphone case, personally, I would print the whole thing slow - it's not so big - will only take a few hours at 20mm/sec. So why not?

Do everything Illuminarti says but be aware that the most likely failure is inside the nozzle - the wiring gets stressed the most there. But if you have a multimeter, you can keep the meter hooked up and while watching it push and prod the wiring in different places until the resistance jumps suddenly to an extreme (e.g. 1,000,000 ohms or more, or less than 10 ohms).

Not related. The print bed can be at any temperature and it won't affect clogging or underextrusion issues. We should tackle that on another thread. Many people have underextrusion issues and there are many possible causes.

This is not related to the print head. It's the print *bed*. There is a 4 pin connector soldered to the bed. Usually the problem is in one of 3 places: 1) On the 4 pin connector one of the screws is loose or a wire isn't in properly. 2) The 4 pin connector has a bad solder joint - this is impossible to see so I recommend reheating all 4. 3) Underneath the UM2, remove the larger cover and check the wires on that end also. The 108 ohms is the PT100 part which you can google (Pt stands for platinum). 100 implies 100 ohms at 25C. 108 ohms implies typical room temperature. It looks like a resistor but it's the temperature sensor. The problem is with the wiring between the PCB under the UM2 and that part. Like I said 90% of the time it's at one of those 3 places above. I'm guessing #2.

It's probably all in the feeder. Take the feeder apart and look inside.

The existing fan is a 12V fan and it seems to work fine. I don't understand. With computers you can install thousands of programs. They normally don't interact. I have probably 10 different CAD programs for example. I have 4 browsers. I have hundreds of programs installed. Thousands maybe. They don't interfere with each other. I plan to convert to PNG format some day (not pdf though). It's not simple since it's a large schematic - I think it needs to be converted to several different images. I've been meaning to do this for a while but haven't gotten to it. I think I have the relevant portion already on this forum - let me check... this is all I've got - doesn't explain that VCC2 is 19V.

It's not you - it's the shipping (and packaging). Many people have warped UM2s and yet they print beautifully. Can't you just put a piece of paper under there? Regarding the micro switch, it might be easier for you to just fix the broken part yourself. It's actually possible to print with a broken X or Y microswitch (but not Z). You get the "rattle of death" but it prints fine after that is over. This is something worth fixing of course but if it were me I would ask UM for the part only - or more likely it might simply be you need to fix the metal part of the switch. Or something moved during shipping. I think the guy who delivers to your house hates you. You will probably never get a working UM2 unless you just fix one of these yourself and again - it's probably pretty easy to fix.

I have wacked the bed hard enough to slide my UM2 50mm across the table with a hard smack using the heel of one hand against a putty knife in the other hand where the other end of the putty knife is pushing against a difficult-to-remove part. I keep thinking this will mess up my leveling but I can go weeks without re-leveling even with this abuse.

Eagle is free: http://www.cadsoftusa.com/download-eagle/ Why not make a pdf for everyone else who don't want to install eagle? And post it somewhere that it will last for decades (I can do that step if you want). I would personally appreciate that effort.

I don't know the answer but you should know that the voltage is actually 19V. Some 12V fans are fine with that and some die within a few seconds.

That's getting pretty good actually. I don't think you can do better on the overhangs - this robot has nasty overhangs and you did very well on those - I think that's around the best you can do. The easiest next step is "cool head lift" to fix those antennas although it works even better if you print 2 robots side by side. Cool head lift might create some little strings that need cleanup with a razor but it will be 10X better than what you have now.

Papier sollte 0,1 mm dick sein, wenn der Referenzfahrt. Ich entschuldige mich für Google-Übersetzung.

Rereading that "raided edges" thread I see once more that fans are critical. I'm wondering if something is not right with your fans. For me the fans helped a lot. Maybe your fans are hooked up backwards and they are sucking instead of blowing. Check the airflow.

I really don't think that's a good idea. It's probably mostly unrelated but 70C is only useful if you have curling/lifting/sticking issues. And you dont. So 50C is plenty. Unless the part is lifting off the bed and all other methods fail, or if you have other shrinkage issues. This is not a shrinkage issue. It's the "raised edges" issue discussed here: umforum.ultimaker.com/index.php?/topic/4094-raised-edges/ Especially look at the second video which is on page 2 of videos which shows the problem in slow motion. I'm really not sure what the fix is. But I have printed parts like this with no problems. So I am very confused. I guess maybe it's time to try a different filament? Here are settings I used to print a perfect 45 degree all-around overhang very similar to your rock: .2mm layers .8mm shell infill 20% 50mm/sec fan 100% There should be zero bumps! Surface should be beautiful! Lower your part more into the bed so you don't have to wait so long for your test. No need to print the whole thing. I really have a strong desire to do more "raised edges" testing but I am so busy with other stuff right now! I'm not sure what you are describing as this could be 2 different things: 1) It could be the raised edge issue described in my link above where the outer edge is pulled inward while it prints. It's like it sticks to the head as it goes around corners and is pulled inward, then it cools and the lip is too high in the air. 2) It could be that other layers haven't cooled enough before the head comes back and prints the next layer or pass. Infill might help with this to give it time to cool.

I don't think this is underextrusion. It could be issues related to movement due to printing infill. The walls may be .8mm thick in 3 dimensions but when you slice the walls, as they tilt more and more the thickness increases until the top is probably something like 6mm "thick" (meaning wall to wall distance of one of the slices a few below from the top). So as they get thicker, suddenly infill is required. So this is more of a model/slicing issue. I don't know what the solution is though. Maybe reduce infill overlap to 0% or even negative 10% (if it lets you do that). There are bugs with infill overlap where it does more than you want sometimes depending on certain angles.

The very top can be improved with cool head lift, yes. That tends to add it's own problems but it will be a big improvement. Even better is to print a tower over to the side where the printer goes over there to print something thus giving the very top of the rock time to cool. (is it a rock?)

This should be easy! I've done 45 degrees on a UM1 and it had zero bumps. Quality was excellent. UM2 also. Usually the bumps on overhangs are temperature related but in this case... I think not. The 70C will only affect the bottom 10mm or so and this is an issue higher up. My first thought was: temperature and speed - keep it cool and slow. But again - I don't think this is the problem. Unfortunately I think it's your layer height. Just for the hell of it, print this at .1mm and .2mm and I think you will see HUGE improvements. STUNNING improvements. Please do it as a test. For me. Thanks.

I posted 4 pictures deep inside wikipedia on several boston marathon related pages. But someone took one of them and put it on the ITN (in the news) page. I posted one picture of each winner of each of the 4 main classes (male/female/wheelchair). All my pics are "CCSA" - "creative commons share-alike attribution". Same licenses used for my models on thingiverse or youmagine.