gr5

I'm sure it's fine. Do you have access to the schematic? There's one here: http://reprap.org/wiki/Ultimaker%27s_v1.5.4_PCB and you need to download "eagle" to view it. Anyway I beleive this is the relevant area. Q4 is the part that you replaced, right? If no fan is connected, PWM2 should be at VCC2 whatever that is (19V? 12V?). The arduino puts out a digital signal which should be around 0V or 5V leading into R20. The other side of R20 goes into the base of the transistor. Current should only flow downwards. When the arduino is "on" aka in a "high" state aka at 5V, the highest the base can get to is typically .7V for a NPN resistor but this is probably a darlington and might not be silicon so I'm not surprised if it only gets up to .2V but that's probably voltage when in the "on" state whether the fan is connected or not. The transistor acts like a switch that switches from the collector (top connection) to the emitter (bottom connection) and not much current goes in or out the base. The base is the control of the switch. The arduino doesn't have the power to switch the fan on and off directly so it uses this transistor. If you are familiar with relays it is the same concept. transistor off With a fan connected, and if the transistor is off you expect 19V at PWM1 and PWM2, very close to 0V (less than 0.1V) at the base and 0V at ground. With no fan and transistor off you expect PWM2 at 19V, and the 3 transistor pins close to 0V. PWM1 might be floating and could be anywhere. transitor on With transistor on you expect .2 to .7V at the base of the transistor, and a pretty low value at the collector - maybe 1V or less. 0V at the emitter (it's connected to ground so it will always be 0V by definition). With transistor on and no fan you expect base probably around .7V (somewhere .2V to .8V I suppose) and the other 2 pins around 0V. Actually the collector might be floating around randomly. In PWM mode the transistor switches between these 2 states many times per second. There is no flyback diode so this is kind of a bad design as you can get quite a bit of voltage and power from the fan each time the transistor is switched off. Someone should add a diode from pwm1 to pwm2 to let the current "fly back" up into pwm when q4 is switched off suddenly as fans are inductive. The diode needs to be able to handle 19V reverse voltage without conducting but otherwise can be very wimpy. PWM mode is probably a bit of a strain on this part. I can see why it fails a lot. In 0% and 100% modes it should be fine. It's a shame because a 1 cent diode could fix this. Maybe there is already a diode built into the fan. But there probably isn't because then if you hooked up the fan backwards it would short out and destroy that diode in a second instead of simply running backwards like others have reported.

You don't need cygwin for Cura. Just install the windows version of cura on windows. Click on the very large "C" on this page: http://software.ultimaker.com/ oh - do you want to be able to modify Cura and compile it yourself?

The x,y,z axes of the UM should be accurate to at least .1mm but the printed model won't be. If you print a cube you should expect some shrinkage. With PLA 20x20x20 mm cube I would expect no more than .5mm X or Y error in the final product (except maybe the bottom few layers). But the height should be correct within .2mm. So getting 12mm when expecting 15mm is pretty far off. You can calibrate any axis without actually extruding plastic. For the Z axis you can tape some paper or a ruler near one of the back corners of the machine, put something flat like a pencil or ruler on the bed so that it points to the ruler or paper and then move the bed up and down with pronterface or cura. But you shouldn't have to calibrate x,y, or z axes as the UM always should have the same values for those axes. The extruder is the only axes typically that could use some minor calibration: #define DEFAULT_AXIS_STEPS_PER_UNIT {78.7402,78.7402,200*8/3,760*1.1} // default steps per unit for ultimaker I got the above line from the Marlin code in configuration.h. These are the steps/mm for x,y,z,e (extruder) axes.

Some filaments are better than others. Some filaments you can just lower the temperature and they won't string at all. Other filmanets - no matter how low the temperature it will make strings - but the lower the temp - the thinner the strings - until they are as thin as a spider web silk string and they are easy to remove afterwards. Can you repost your picture in the ultimaker forums? USually I can click on your pictures to see larger version but not this one. On the top left corner click on "gallery" then upload your picture to a gallery. Then make a posting and click on "my media" to insert the picture into your posting.

Go to "file" "preferences". Set the 5 variables in the bottom left to the recommended settings. If gantry is 0 (or smaller than your object) it will print them all at the same time (parallel) so that is the most critical dimension. Just hover over each one and type in the default values for ultimaker.

I'm glad you fixed it. This is a common problem. You really have to tighten the hell out of those little tiny set screws. You should have gotten spares in your UM kit. I have heard that the spares are much better quality and if you get slips in the future throw away the old set screws and put in the newer pointier ones.

The "banding" is occurring right at the starts and ends of letters which makes me think it is a slicing issue. I would look at the gcode very carefully. If you aren't going to be able to tolerate minor bands like that then I don't think you are ever going to be happy with 3d printing. However they might be fixable by changing slicing settings it's just that I don't know what's going on exactly. It could be underextrusion after having slowed down and then sped up (the extruder takes a bit to catch up). It could be infill too close to the edge and having a thicker skin setting would help (.8 instead of .4) it could be a bug in the slicer. It could be a solid infill layer. It could be you turned on "joris" which you definitely don't want to do on this model but has been known to cause similar bugs. If you email your gcode to illuminarti he's been known to look at those issues occasionally. It could be that your model is a bit messed up and somehow you have inner walls or an inner gap when you unioned the F with the cylinder. That might be visible in the "xray" mode and fixed with one of the "fix horrible" settings. It could really be a lot of things and I think all of them are slicer related. You could try kisslicer or cura 13.04 which had a very very different slicing engine. Or just look at it yourself - go to the "layer" view mode and get to that layer at the top of the F and look at it very carefully. Illuminarti would actually check the extrusion amounts against the move distances.

To reduce stringing look at the first picture: http://umforum.ultimaker.com/index.php?/topic/1872-some-calibration-photographs/ Every filament is different so I recommend you do the same test i did at different temperatures. I have retraction amount at 4.5mm which is cura default and is the perfect amount. I also have one of these: http://www.thingiverse.com/thing:46157 If you retract too much you will get air in your nozzle followed by blobbing.

I don't know but if you google ultimaker and thermocouple there are all kinds of failure modes. The most common failure mode is the wiring at the head if you don't use the strain releif properly. Everytime the head moves, the wires move and the strands in the wire break one at a time. When they are all broken you still have intermittent contact but the temp bounces around. Also the amplifier board which is located on the print head can go bad and add some kind of voltage offset to the signal. Also people have reported problems that the temperature changes suddenly when the fan or heater on the hot end are turned on. It will go from 210C to 230C instantly. This I think is caused by signals getting from one wire to the other (through radio waves or magnetic induction or cross talk or whatever you want to call it). Keeping the fan and heater wires farther away from the temperature wires might help but they typically all get stuffed tightly together. There's just too many failure modes. YOu have to measure the voltage before and after the amplifier and again at the UM circuit board to figure out where the problem might be. But a simpler test is just to raise the temp to 240C and see if it helps or not. If not then you can probably eliminate this as the problem.

My screws stick down far enough to keep the block from spinning. I don't know what is different about your hotend. Maybe your screws are shorter? My block is loose enough that it can indeed rotate. I don't think that is a problem. Your white cable is pushing on the fan shroud and may be partly to blame why there is a rotational force. When it rotates enough to stop pushing against the fan shroud it will probably stop rotating. I suspect it's fine. Many hot ends leak a little PLA. Over many hours of printing the PLA gets stickier and harder and blacker and will eventually probably stop future leaks. You could also shove some ABS through your hot end which UM used to sell a few feet of just for this purpose (plugging leaks). I would just leave everything as is until you are forced to take it apart again.

Make sure your filament is clean and won't be dragging any dust or anything into the nozzle as it goes up through the bowden tube causing a new clog.

So many possibilities... Here's a thought: What if the amplifier for the thermocouple is reading high. Many people have been complaining about this. Some people have even modified their firmware temp tables to fix this. People who were printing fine but now their nozzle temp has an offset. So if your nozzle is much colder than you think, it might actually be at 150C even though you think it's at 190C. This would explain how your delrin wheel got damaged (too much pressure). The quickest test is to just raise the temp to 240C (I often print at 240C anyway - I print anywhere between 190C and 240C depending on a few factors). Also I recommend setting the nozzle to 190C, turning the extruder by hand to feel the resistance. Try turning the big gear slowly and then quickly. Then raise to 240C and try it again. Try turning slowly, then try a quick half-way-around rotation. Does it come out differently?

So I was most likely wrong about the temperature - after you fix z steps, bring your normal printing temp back up to 220C.

Aha! There's the whole problem! Your z axis is messed up and it's putting out almost twice as much filament as it should be. Now you need to calibrate your z axis. Don't print anything. Use Cura or pronterface to move the z screw up and down 10mm and measure how far it actually goes. You need to fix the steps per mm in the z axis. You are probably off by a factor of exactly 2. If you have an ulticontroller, make sure you save the settings to the prom or you will lose these when you power cycle the UM. If you don't have an ulticontroller you definitely want to get pronterface. It's free and wonderful. It can make it easy for you to adjust and read back the current steps per mm. Also with pronterface you will need to save to eeprom. In addition to pronterface refer often to this guide of gcodes: http://reprap.org/wiki/G-code Especially M92 and the M50X commands.

Did you feel it? Are you really really sure? The next most likely thing is you need to increase "minimum layer time". I found that 5 seconds is enough but whatever it is for you then you should double it. These 2 things are by far the most likely - I've seen this many times and your basic problem has to do with cooling. Other causes: Are you printing in an environment that is very hot? Say over 90 or 100F? (Over 35C?)? There is some kind of problem with heat - just don't know what it is. Maybe the nozzle is hotter than you think. Even though the UM thinks the nozzle is at 200C it might in fact be at 260C. So lowering as low as you can go might help - for example lowring the nozzle temp to 180C might in fact bring it down to 240C which should be low enough. If you have some kind of method for measuring temperature accurately that would help. Maybe you have a weird filament that is melting at lower temperatures. You could start with a cold nozzle and increase the temp slowly from 50C to 180C and keep some pressure on the extruder wheel with your hand. See at what temp the filament starts coming out. This is unlikely the problem but gives you an idea. There is definitely a heat issue here.

You should not change the steps per mm on X or Y or Z axis as they are dependent on the hardware which shouldn't change from UM to UM. The variation you are seeing is from things like shrinkage, overextrusion, and play (aka backlash). You should print the part you want to print, and then figure out all the dimensions that need correcting - ones that must absolutely be corrected - and fix only those dimensions in the CAD. I don't think you can get more accurate than .1mm so if you are getting .06mm accuracy then that is probably better than anything I've heard of being possible.

I recommend you mark the pulleys with a sharpie (permanent marker) so you can tell if something is slipping. I hope you realize if only one axis is slipping then you only have to fix one axis (X or Y).

Maybe change UM to "Ultimaker" logo? And add the robot? The Sex and Beer part is fine. Beer is one of those words that is recognized in many languages.

Your owl is getting better but there is some serious serious cooling issue. Did you check the fan? What was the deal? Could it be off? Could it be backwards? Please reply about the fan. The fan. The fan. The one near the nozzle. It should be spinning. Need more info. I'd like to suggest other things it could be but most likely it's the fan. It could be backwards. Seriously - check the fan. Let us know. If you are convinced the fan is working, I will tell you my second choice for the problem, but I suspect it's the fan. I mentioned the fan in an earlier post because - well that's what it looks like when the fan is off. There are other possibilities but I'm afraid you won't answer my question about the fan.

I'm not sure what this is but people have described a similar sound caused when the belt rides up onto the pulley edge and then snaps back into the grooves. I'm 90% sure that's not happening with the 2 long belts in that corner so that would leave the short belt. One fixes this by making the 2 pulleys lined up and perpendicular to the stepper axis. Or it could be something totally different. ARe you sure it's not the exturder? Remove the extruder from the back of the UM while printing - if nothing else this will make everything quieter so you can hear the noise better.

Someone recently printed several different objects in cura 13.06.X (not sure which version) and cura printed every other object opposite order. In other words one would print brim outer to inner, then the next inner to outer, and then back again. I think there is a bug but I don't know the details. I suspect it's an intermittent thing that has to do with the shape of your object or something. I think Daid knows about this and will hopefully fix soon.

You can print just fine without the right limit switch. I have one but it has never triggered. You can print without the fan but I recommend fixing this at some point. The darlington transistor that controls it often breaks and it has two common symptoms: always on, or always off. Either way it's often that transistor. Inexpensive and easy to replace if you are good with electronics. Or get a new controller board from UM. Not sure exactly what you are describing but this sounds like a common levelling issue. You probably don't beleive me but this can happen and *does* happen to pretty much everyone at some point if they don't level their bed properly. Here's my tutorial: How to level: Levelling here is defined as setting the Z height and also levelling. It's one procedure that does both at once. Optionally heat up the nozzle to 180C if first layer is .1mm or less because a cold nozzle shrinks and you will be setting the bed to the wrong height. Make sure tip of nozzle doesn't have any plastic on it or you may level to the wrong height. Home the z axis only. If you must home all 3 then you need to disable the steppers once it's done so you can move the print head by hand. Move the head as close as possible to each of the 4 screws in turn. Once at a screw tighten the screw and then slip a piece of paper between the nozzle and the print bed. Make sure the paper slides very freeley. Then loosen the screw until the paper gets slightly stuck. You want the paper to easily be able to slide in and out under the nozzle with one hand pushing the paper. If the paper gets stuck it's probably too tight under there. Repeat this procedure for the other 3 screws. Then go back to the first screw and repeat on all 4 screws again. Then repeat on all 4 screws again. Then again. It may take you 20 minutes to do this the first time but the second time you do this it should take much less time because you are both better at it, faster at it, and because there isn't much to adjust the second time. If you levelled with a cold nozzle you are done. If you levelled with a hot nozzle you should then loosen the 4 screws 1/8 of a turn to compensate for the thickness of the paper. Once done levelling rotate the z screw by hand to keep the nozzle off your bed. This makes it less likely to damage your bed surface and gives the nozzle room to leak. On a new ultimaker repeat this procedure before every print (at least every hour) because the print bed can move/droop like a new guitar string. After many months the droop slows down.

Well something is wrong. Well for one it looks like it might be too hot. Maybe the hot end is hotter than the requested 220C. Can you measure it somehow? 220C or even 240C should be fine for this print but it just looks wrong. Try lowering to 190C just to see if that changes the print much. Make sure the fan is working - the fan could possibly be the entire problem if it is broken. Or if it is spinning backwards and sucking instead of blowing. Also I think your belts are loose - the loose belts will keep the infill from reaching the edge/sides. There's a good video explaining how to test the belts and how to tighten them: www.youtube.com/watch?v=grHmmmSoOfc If the belts are fine (long and short) I still think there might be some backlash from something else but the belts are the most likely so check those first. Don't change too many settings at once. Those other settings suggestions shouldn't be causing the problems you are seeing. And get rid of the "minimum speed" of 5 in the cool settings. Set it to zero. That can also cause things to be too hot. Also maybe disable retraction just to see if that helps - I don't know what retraction settings to use for 1.75mm as I'm not sure if it is "bunching up" in the clear bowden tube or if a smaller tube was installed. The UM really doesn't work well with 1.75mm unless you modify both the hot end and the bowden tube although some get away with modifying the bowden only. For 3mm you want 4.5mm retraction. You have 7mm but maybe that's the best setting - I don't know. So maybe just disable the feature for a test. If you don't retract you can get something called "stringing" but since you don't have gaps in your print it doesn't matter. Maybe you can take a video of the second layer as it goes down? Or clearer pictures of a particularly bad layer?

Good theory. You might want to disable the "skirt" and "brim" settings in Cura for such a large model. Or can you print this smaller? Scale it back by 10%? e.g. scale= 0.9?

I think it is looking for any SD cards. Go to File Preferences and uncheck the "auto detect SD..." checkbox. Let us know if that fixes it.