gr5

These are metric by the way - I think 2mm. So if you have the wrong hex wrench (in inches) don't do anything until you get the right one. Oh and if you want to prove me wrong you can mark the shaft of the motor and the pulley with a tiny dot and see if it moves when the problem comes back (do all 6 pulleys while you are at it).

This is very common. It's always one of the 12 pulleys. Since you say "to the right" it's the X axis so now you only have to worry about 6 pulleys and 2 of them might be connected - I forget. Anyway the most likely one is the one on the motor. I know you said you tightened almost to stripping them but: 1) Did you get the one on the x motor? Back right corner. The cover comes off the motor if necessary with just 2 screws - very easy. 2) You want that one tight as hell - the allen screwdriver should be twisting quite a bit. Then make sure you get the other ones. The second most likely problem pulley is the second pulley on the short belt which is the one just above the motor. The other 4 pulleys are rarely a problem as they have half as much force plus two of them have to fail to get a shift.

Print that upside-down on the glass, then flip glass over. Now it will be in every photo of things sitting on the print bed!

I usually print ABS at 245C but I don't get great layer adhesion - even with the fan off. It's strong but not nearly as strong as PLA. It breaks along layer lines if you push hard enough. I should probably print closer to 260C to get strong parts but I don't want to wreck my isolator so I'm waiting on a nice solution from Anders Olsson. We'll see. Until then I print at 245C. It's a good safe starting temp for ABS newbies like me (only printed about 20 parts in ABS ever).

Images of almost zero slippage (< 5%):

Is the filament going through the nozzle while doing this? If so then 6% (188 versus 200) underextrusion is fine and won't be visible in the print. What happens is the splined (or knurled) wheel that pushes on the filament digs tiny holes into the filament. The filament gives a bit such that you loose a little bit of forward momentum. The more you are losing the longer the holes get until they connect and you get slippage.

There's not much you can do. If you set infill to 0 and top/bottom thickness to 0 and shell to .4mm and skip spiralize you will get almost what you get now - a .4mm shell with no top or bottom. But if you set top thickness to even just .1mm you will get "shelves on many of the sloped layers (top 2/3 of your part) to support the layer above - this will show through to the outside and change the look. Really the only thing you can do is a gradual curve to close the top very very gradually combined possibly with thinner layers so the steps aren't too obvious at the top.

Just to clarify - from some other things I've heard lately, the latest Marlin firmware can have crashes such that Marlin is over-writing memory and doing very bizarre things. It may be a few people have defective arduino or memory on the PCB but I'm very suspicious of the latest firmware (both 15.01 and 15.02). This other guy also tried 14.09 and it fixed his issues.

I looked at my extruder during "move material" and in that mode it takes 14 "clicks" to move 1/4 turn or about 57 to go all the way around. Watching your video again I can see that it was working properly *after* power cycle. This blows away my theory about the wrong number of microsteps because this controller doesn't do anything more than 16 microsteps yet you seem to be doing very little movement. Now I'm wondering if it's a firmware issue. I had another report of very very bizarre firmware crashes in one of the newer firmwares. I recommend you go back to one of the "14" firmwares (from 2014): http://software.ultimaker.com/old/ Try 14.09 - you have to install Cura 14.09 (Cura versions install side-by-side so you can have many installed at the same time) then use the USB to load the 14.09 firmware into your UM2, then use a newer Cura for slicing (or whatever one you want).

I'm having trouble telling if you are too close or too far from the bed on that first layer. After much indecision I guess it's probably about right - but really not certain. There are gaps as it printed which indicates the extruder skipped backwards - but I suppose it's possible the nozzle was too far from the bed and the pla never stuck at all. I really can't tell. If you heard the extruder skip backwards then I would raise the temp to 225C. If you didn't hear it skip backwards then I would move the bed closer to the nozzle by about 1/8 turn on the back screw (problem seems worse towards the back). 210C is a reasonable print temp but you have to print a bit slower. Your printing speeds (20mm/sec with .3mm layer (or is it .6??) and 50mm/sec with .1mm layer) all seem reasonable for 210C but your filament might be slightly more viscous than typical and might require a bit more heat. I almost always print at 220C and sometimes up to 240C for PLA. Have you ever printed ABS? If so that could be part of the problem - that stuff pretty much never completely leaves the nozzle. You could try some cold pulls. If changing to .3mm bottom layer doesn't help and changing to 225C doesn't help you should print this test at 230C (the test is meant for 230C so never print it at any other temperature or no one can help you): http://umforum.ultimaker.com/index.php?/topic/4586-can-your-um2-printer-achieve-10mm3s-test-it-here/ This test is only valid if your extruder skips backwards - it isn't valid for UMO or UM2's that are adjust such that the extruder doesn't skip backwards under high pressure. If you can get the cylinder up to 5mm^3/sec you are probably okay. If not then we can talk about fixes.

Change the bottom layer to .3mm. Realize that printing .6mm bottom layer at 20mm/sec is 2.4X more volume extruded per second than 50mm/sec at .1mm layers. Or maybe you had a typo and you already are at .3mm for bottom layer? In that case you are only printing 1.2X more volume - not a huge difference.

Also when you post the support ticket include a link to this forum post! Oh - what I failed to mention is that when all 3 pins are high (MS1,MS2,MS3) then you get 16 microsteps. But when MS3 is low you get 8 microsteps. That's what I think is happening. It's arbitrarily switching between modes.

Okay I looked at the schematic. The microstepping is controlled through hardware - not software. The 3 pins that set this are MS1,MS2,MS3. They are all supposed to be pulled high through a resistor R88 (circled in green below). A 4.7K resistor. Something is wrong there - either MS3 has a bad contact (maybe it didn't get any solder paste) Or maybe R88 is missing on the board completely and the other pins just barely float high most of the time. Or U12 (the driver) is defective. I recommend you go through support.ultimaker.com. Make sure you tell them you are in USA and you will probably get a new board shipped out today or first thing Monday. When you take out the old board inspect to see if R88 is missing! Or if something looks wrong with the MS3 pin (pin 11 on U12 as shown with blue arrow).

I watched the video again. I must have had the sound off for the first 15 seconds or so because I didn't hear your voice the first time and didn't see the first part. Yes - there appears to be something wrong. I think the steps/mm is off. I'll look up the "pololu" part specs but I suspect you need a new PCB. Fortunately for you, you reside in USA where you can get "local" support.

If dust and stuff gets on the filament while still spooled it can slowly make it's way through the bowden to the nozzle. That's one possibility. Leaving the nozzle too hot but *not* printing can slowly caramelize the plastic. That's another possibility. You should know that you can get an "Olsson block" replacement heater nozzle which then allows you to swap e3d nozzles very very easily. The e3d nozzles are only $10 each and you might consider them expendable. The other advantage of the Olsson block is you can use larger nozzles for much faster but lower resolution prints or a smaller nozzle for very small fine parts. post #329: http://umforum.ultimaker.com/index.php?/topic/7689-custom-heater-block-to-fit-e3d-nozzle-on-ultimaker-2/?p=95991

That's what I thought but I printed something that *did* hit the block and only tried to move 1mm farther and the belt stretched a bit but the motor didn't skip. Fortunately for me it was printing the skirt and so it only tried once on the whole print. If you go a little more (maybe 2mm instead of 1mm) it will eventually skip as you say and the rest of that layer and all layers above will have moved a bit.

I'm sure it *did* hit the right side - the print head hit the right side block. I'm sure of this because you said your right wall was 15% as thin. I've had the same thing happen - you can press the head against the right block and move a little farther with the stepper and the belt will stretch a bit but it won't move the head any farther. This happened to you because you lied about the width (232mm) and so it put the center of the part at 116mm (instead of 115mm) and printed the part 1mm to the right of where it should have and went just a tiny bit to far on the right edge when it printed to approximately X position 231mm. So telling Cura the machine width was 232 may have fixed it so it printed but it printed 1mm to the right of desired position. YOu could subtract 1mm from *every* X coordinate in the gcode file with a new plugin. Or fix it with the G92 command.

I did a quick experiment with Cura. I loaded an stl file that was 10mm on a side and pushed that cube around on the bed until it turned gray. Cura shows a gray/blue checkboard pattern with 10mm squares to simulate the glass bed (the actual glass is wider so it simulates the printable area). Then it also shows 4 gray areas which are "keep out" areas or zones where you can't print. Whenever my object to print got too close to one of those keep-out gray ares the part went gray. Those of course represent the clips that hold the glass. Then I tried something else - I tried to scale the box up in only X and used the "to max" feature after making it long in the x direction and the max print space I could do in X was 228 so cura wants 1mm on either side for some reason (even though the nozzle width of .4mm should buy you an extra .2mm on each end). So I would think I can go to 230.4 but I could only go up to 228mm. The fix is obviously to go to 232mm in machine settings. But then it prints 1mm to the right - to fix that you can do this: G1 X0 (move to the left most position) G92 X-1 (tell the printer that this is at postion -1 Make sure you do that after all the moves to X=0 in the gcode - if any other moves in the code go to X<1mm then they should all be fixed but I don't think that's an issue - there should only be one or two moves to X=0 and only at the very start. I'm not sure if G92 can even work with negative numbers. It will probably work. It seems like Cura shouldn't steal those last 2mm from me. In fact it is stealing 2.4mm from me. I shouldn't have to hack the gcode to get it back. Maybe I'm missing something.

The print head will hit an endblock - that's usually what stops the head from moving farther. The print head is made from black plastic and has a very flat surface. The endblock that is attached to the belt and holds the rod that the head travels on - that is *also* very flat. Those two surfaces touch when you are at the extremes. You can get a little bit more than 230mm out of the X axis if you play with the limit switch - you can loosen the screws and move it slightly or you can just bend the end of it with some pliers. If you push it too far then the head can't reach the limit switch.

I'm not sure about skirt length 0 versus brim length 0. I don't print this large more than twice and I forget what I did. If it didn't print the skirt then cura probably didn't take that into account so not sure why 225 model couldn't fit in 230. It should. Just note that when you lie and tell it you have 232mm bed it's going to move the center of the part over 1mm and you are more likely to hit the right side. By 1mm.

Excellent. Now you should know that at 240C your teflon isolator may die sooner - maybe after 500 hours of printing? I'm not sure. And if you can keep it at 230C that would help extend it's lifetime. But the part is cheap (free since you can get one from UM for free) and there are isolator washer's coming out soon I think from third parties (they isolate the isolator).

If it is skipping on the bottom layer it is almost ALWAYS because you are too close. Rotate the screw clockwise (tighten) such that you have more space on that side of your print bed by 1/8 turn. If it is skipping on HIGHER layers it has nothing to do with leveling and the most common issue is printing too fast or too cold (unlike the UP the Ultimaker has extremely high acceleration such that if you ask for 50mm/sec you actually get 50mm/sec whereas you are lucky to get 25mm/sec out of the up).

There's subtle differences in the clicky sounds. While printing the stepper can get noisier or quieter as the pressure is higher. When you are in "move filament" mode I believe it is doing large steps (not the minimum) so it is noisier then also. So I think these changes you are noticing have more to do with pressure than with turning the printer off/on or how hot it is. These steppers can handle very high temperature - they don't overheat easily. In fact some people turn their heated bed up to 110C and cover the top and front of the machine and print for hours with the steppers in that 40C environment (about 100F) and they are fine. When you are printing and it clicks much louder - if you look at it it actually skips back 1/4 turn - and it underextrudes (stops extruding actually) for a second or so afterwards - that's a problem. That happens when the pressure builds up extremely high in the nozzle. Your from kansas so I will use pounds - about 10 pounds force on the filament at the feeder (enough if you try that you will lift the printer off the table) and over 100 psi in the nozzle (if there isn't friction elsewhere e.g. in the bowden or in the teflon isolator). You may indeed have underextrusion issues but your video isn't showing this so I assume your printer is fine.

By the way you really should print with glue. You need very little so you can mix in a little water and spread it around with a tissue. It should be so thin it's invisible once it dries again but you can test it by peeling/scratching some with your fingernail - it's still there.

The bed looks pretty close. Stop doing the bed level procedure and simply adjust the 3 screws as necessary. You want the filament squished flat but not too flat, lol. Most of the print looks properly levelled - not sure about the failing area. Now on to your problem - I'm not certain (video would help) but I think those failure spots are the filament feeder skipping backwards. This happens when there is too much pressure for it to push on - typically 10 pounds (that's for you imperial unit british people). You didn't answer key questions earlier - what is your print speed and layer height - I assume bottom layer is .3mm but maybe you did thinner? You need even better leveling for thinner layers. I'm guessing your only problem is printing too cold and too fast. Here are max speeds versus temperature I recommend for quality printing on UM2 or UMO. This is valid only for .3MM LAYERS. For .2mm layers go 50% faster for .1mm layers max speed is 3X these values: 14mm/sec at 200C 20mm/sec at 210C 27mm/sec at 225C 33mm/sec at 240C Note that Cura does a special speed for the bottom layer only. I think it defaults to 40mm/sec? Or maybe 20mm/sec? Another thing that can cause skipping on the bottom layer is leveling - if the nozzle is too close. But usually you will see that the layer is so thin it's transparent before that's a problem. But still - it can happen. The bottom layer is very tricky and if you want it to look perfect you need extra perfect leveling but don't use the procedure - that just gets you close (within about .1mm). Instead do final adjustments by looking at the bottom layer and turning only 1/4 turn at a time or even a little less (1/8 turn).