gr5

This is a common problem for overhangs. I call it "raised edges". There's much discussion. The basic fix to reduce raised edges on overhangs like this is to use more fan. But this leg looks like it should be much stronger! I suspect you have bad layer adhesion and should be printing a bit hotter. If you destroy the part later does it break easily along layer lines? If so you aren't printing hot enough. This part should be unbreakable in Taulman Bridge. It should bend in a complete circle and still not break. So: More fan, more heat.

240C is too cold if you want good layer bonding. I've printed ABS and it had "okay" layer bonding but compared to PLA it was weak. I'm not sure what the correct temp is but 240 or 245C is definitely on the cold end and you can get these layer separation issues. Fan at 25% should be okay but I recommend with ABS that you put a box over the top and cover the front with clear plastic to get it up to about 60C inside (don't go much hotter than that because the steppers get too hot). A photocopy paper box is the perfect size to put on top of the printer with room for the bowden:

I would keep it all one speed. When you change speeds you tend to get over or under extrusion at the change point. Well bottom layer should be slower but the rest all the same. For 210C and .1mm layers keep it under 60mm to avoid underextrusion. But this is all a different problem. Have you used more than 2/3 of the spool? I find I have to print at half speed (30mm/sec in your case) for the last 1/3 of the spool. Have you printed much over 240C? If you've printed say 10 hours at 260C that *might* be enough to deform your isolator - especially if it was under extra pressure. I am going to say the most likely 2 things are: pressure on isolator, and partial nozzle clog. You can check out the isolator by loosening the 4 screws until the bowden is loose, then tightening them slowly until it stops moving up and down and then tighten maybe one more turn. For the nozzle - I recommend getting some hypodermic needles or acupuncture needles (available on the internet in usa - which amazes me) that are smaller than .4mm 28 gauge needles for cleaning nozzle: http://www.healthwarehouse.com/needles-syringes/disposable-syringes/gauge/28-gauge http://en.wikipedia.org/wiki/Needle_gauge_comparison_chart Or you could try a cold pull (atomic method) a few times (I don't take apart the head for this but pull from the back through the feeder). Or you could take the nozzle off and burn everything out of it. Really the easiest if you have one is scraping around the inside of the tip of the nozzle with a hypodermic followed by a cold pull. The other symptom of a partially clogged tip is the filament doesn't come out straight but at an angle so severe that it touches the tip again. Even if you pull maybe 2 inches straight out of the nozzle first and then let go and extrude some more - 2 or 4 inches of hanging filament should be enough force to keep fresh filament from going right back up against the nozzle. If not there's probably a partial clog slowing you down. Consider getting the olsson block - much easier to take apart and clean: post #329: http://umforum.ultimaker.com/index.php?/topic/7689-custom-heater-block-to-fit-e3d-nozzle-on-ultimaker-2/?p=95991

Steel also doesn't conduct heat well but the problem with steel is that PLA sticks to it much too well. That's what is great about teflon - things don't stick to it. I don't know if things stick to titanium or not. The high temp feature is common in all metals The low-conductivity is common in many materials The non-stick is rare

You should be able to get a temp sensor from ultimaker for free - just tell them you had to break it to get it out when there was an issue with filament all over the head. Alternatively you can get them here: http://3dsolex.com/?page_id=579 but only if you also buy the Olsson block. The Olsson block has a hole so that you can PUSH the temp sensor and heater out! Nice design.

Those are for UMO only. Not UM2 or UMO+.

Turn printer off and on, go to CONFIGURATION ADVANCED and set the nozzle temperature to maybe 40C. See if it starts out at room temperature (20C?) and moves to 40C. Try pushing on the wires leading into the test head to see if the temperature changes suddenly. Push the print head to the front of the machine and look from behind - there are 2 cables going into the heater block. One is heater, one is temp probe. Are both still inside the head? Upgrade to the latest firmware. There is a bug/feature that caused lots of temp probe errors. This is fixed in 15.01. Also sometimes the heater block is touching the fan shroud and when the fans come on full (several layers from the bottom) the heater doesn't have enough power to keep the nozzle at goal temperature. Check that you can manually set the nozzle to 230C with the fans at 100% (do this by doing PRINT and then choose any print job and then immediately TUNE and stay in the tune menu so it won't start printing. Then set the nozzle temp manually and the fans manually).

Clean your blue tape with isopropyl (aka rubbing) alcohol! It's found next to bandages in any store that sells bandages (like supermarket or drugstore). There is some wax on the blue tape to keep the glue off the non-sticky side. You need to clean this wax off before printing on the blue tape. This will greatly improve sticking to it. Your bottom layer is critical. Halt any print if it isn't really a good base. Also use the brim feature for prints that need extra stickiness like a tall thin print or prints that are warping/lifting corners off the tape. Also if the tape pulls off the bed buy some wider painters blue tape. On the internet (it costs too much in stores).

spiracy - please validate your email. You can re-validate (have the forum send you a new validation email). Also you should know that UM usually honors the warranty even if you are out of warranty and the problem was your fault. Just be honest and kind and beg a little. You might get lucky.

Friends with the right equipment = major value.

It's very similar on UM2 as it has the same 2 Z bearings and the same Z nut and screw. That sand noise is ball bearings falling down. It's not a problem. The ball bearings are in the 2 side linear bearings (the z nut has no ball bearings). It would really help to loosen the screws on the side bearings to see if maybe they rods are bent slightly and you get a stuck spot. This could move your problem spot by loosening - moving the bed to the problem area and retightening. I'd hate to see you get a new z screw from ultimaker only to find the problem is actually in the 2 side rods.

If your printer can do 8mm^3/sec at 230C then your printer is better than average. I recommend 3mm^3/sec max to have a good safe margin in case your filament tangles or some other temporary friction shows up. So what speed, layer height and temp are you printing at exactly? Feeder clicking is definitely bad on anything other than the first layer just as shown in your cute octo photo (those photos are tiny by the way - mean they are like 20 pixels across or something - .0004 megapixels?). Oil works great - it's okay if it gets in the feeder. One drop will last maybe an hour. Even 10 drops per minute will not affect quality of part. But my recommendation is to move the filament spool onto the floor so that it goes up into the feeder vertically and not at a severe angle. Also try loosening the 4 screws on the print head one full turn - they may be making the bowden push down too hard onto your white teflon isolator (at the base of the bowden).

I'm told all the wear and tear of carbon fiber only happens on the *outside* of the nozzle at the tip. It's like sanding down the tip of the nozzle. You can get a stainless steel e3d nozzle. Plus the nozzles are so damn easy to change and they are only about 10€. Anyway I think you can get them here: http://e3d-online.com/Extra-Nozzles Those nozzles work great with olsson block - just make sure not to get the 1.75mm filament ones. and if you scroll down you will see there are stainless ones. Also note that Swordriff's nozzles are probably slightly better and cheaper.

The tube like box will print mostly fine in this configuration as the printer can "bridge" very well. However there is an overhang lip along the top edge that will need to "print in mid air" in this orientation. One solution would be to add a temperary post to support that corner and clip the post off later. Also inside the tube-box is a "shelf" shape. Because this is a 2D drawing of a 3D object I can't tell if that is a) hanging from the ceiling b ) sitting on the "floor" c) suspendid in mid-air. If B then it should print fine. If A then you could flip the print over. If C you will have to add a support at each end if you print in this orientation. More tips here: http://support.3dverkstan.se/article/38-designing-for-3d-printing

Are these corners rounded? If there are more than 12 or so segments in the curve at the corner then that is your problem - you really want to limit the number of segments that Marlin has to navigate to less than 12 for every 20mm or so as if you don't it has to slow down and be prepared to stop within 12 gcode commands. Also check your acceleration and jerk settings. Why? Well... gcodes move the printer in line segments. For example a circle might be 20 line segments. Before marlin and sprinter we had acceleration only and the head would come to a complete stop at each vertex of the circle. This was incredibly slow. This was because at the vertex the acceleration change is instantaneous so you have infinite jerk (accleration is derivative of velocity, jerk is derivative of acceleration). The fix was to introduce a "sprinter jerk" or "marlin jerk" term that is in mm/sec (not mm/sec/sec/sec) which is the allowed instantaneous velocity change at a given point. For ultimakers with it's extremely light weight print head we use 20 for the jerk setting and typically 5000 for the acceleration setting. For a rep rap printer it's more like 5mm/sec for jerk and 1000 for acceleration. Anyway Marlin runs on a very weak computer that can only look ahead about 12 gcode commands and as it is moving quickly printing a circle it has to be ready to slow down or even stop at the end of the current string of gcodes because it doesn't know what's coming up on the 13th command (may be a stop and retract). So it has to go pretty slow. As long as the next 12 gcodes (line segments) take you far enough Marlin knows it can slow down in time but if there are 12 line segments in the rounded corner of a cube for example it might have to slow to 5mm/sec for those corners which visually appears as a complete stop. Alternatively USB could be saturated. But the fix is the same - reduce the number of triangles in your STL file. Here is a guide to reducing them for very large STL files (meshlab is free): http://www.shapeways.com/tutorials/polygon_reduction_with_meshlab

180 is a fine temperature for colorfabb especially when printing slowly or printing thin layers. For example if you are printing with "normal" ultimaker PLA at 180C and .05 (very thin!) layers I wouldn't want to go over 40mm/sec. But for .2mm layers I would stay under 10mm/sec. But colorfabb I believe is less viscous and can be printed faster than those speeds at 180C. This is from other people's posts - I have never tried colorfabb.

Please post a picture. This is a reasonable speed but I recommend you don't go any colder or faster than this. I can print this fast for the first 2/3 of the spool of filament but not for the last 1/3 as the curvature is too strong. There are many many possible causes of underextrusion on UM2 (about 20 possibilities). Having a picture would help. The first thing I recommend is to move the filament to the floor such that the angle as it enters the feeder is better. Consider printing a bit hotter or slower if it underextrudes as a temporary solution. It could be the Isolator. Or many other things. There's ways to test each possibility but start with posting a picture and using google to search this forum for "underextrusion".

#329: http://umforum.ultimaker.com/index.php?/topic/7689-custom-heater-block-to-fit-e3d-nozzle-on-ultimaker-2/?p=95991

No one can help you until you at least figure out which of the 3 problems above is your particular problem.

Sorry you feel that way Taxez. I have had problems on and off with my printers but over all I love them. They are working great for the last year. It's confusing that two different people are posting similar but possibly completely different issues here. Taxez it would help if you posted every day instead of only once every few weeks as it's hard to remember everyone's issues. on the spool itself. Sometimes you can purposely tangle it - wind the filament on the spool but purposely run the filament *under* a thread of filament in a lower layer. Then try unwinding it - it's tricky! Hopefully you aren't in this condition but its VERY EASY to get in this situation and the only fix is to remove the filament from the machine, unsppool several meters and fix any of these that you find. This is just one theory - there are many things that can cause a gap layer. The filament tangle is by far the easiest to check as you can just unspool some filament and realize it's stuck. Somewhat. Another theory is that the temperature dropped on that layer for some reason - seems unlikely as it appears to only happen once in the entire print - usually changing temperatures only happen on a repetitive bases every few minutes. You can verify this is not the problem by watching the temperature carefuly in the TUNE menu while printing and make sure it stays within 1C (or maybe 2C) of goal temperature while underextruding the bad layer. The third theory I have (these are all things that happen to people all the time on this forum) seems to me the most likely and possibly the most difficult to prove - it's that there is something sticky or wrong with the Z axis where it moves a little bit less than it is supposed to (say .19mm instead of .2mm on each layer) and then suddenly it finally catches up and drops 2mm all at once. Something like this. Because something is sticky maybe in the Z axis. You can test this by pushing the bed up and down and trying to feel for a bad spot. You can visually inspec the Z screw - also look for it warping. But the best test is to print a few different parts - preferably a 10mm square tower that goes the same height as the print you already printed to see if there is a gap layer. Then print another one 15mm square to see if the height is *still* the same. If 3 out of 3 prints hafe the same problem at THE EXACT SAME HEIGHT and the 3 prints are all different sizes (different amounts of material on each layer) then the problem must be related to Z screw or Z nut, or Z bars and you can have Ultimaker send you new ones. It might be as simple as adding some grease to the Z screw (in the green plastic container) and some light oil to the z rods.

Then panels are no problem - it's all posted. The main problem would be getting the circuit board printed for the UM2 even though the schematic and layout is public. I recommend you choose the electronics as your first step. The UM2 has a board that I think you can buy on their website - I think it's around $200 possibly but not sure. You could see if it's for sale on fbrc8.com for pricing only (they ship us only). Just because it's not listed doesn't mean you can still get it. Alternatively you can get a ramps board much cheaper and it's a fine bit of electronics. The quality of the UMO is about the same as the UM2 - they are both quite good quality. Note that the UMO is *not* called the UM1 because it is continuously improved. Now it comes with heated bed as standard equipment.

@small - your printer is kind of typical - it looks like you made it to 5mm^3/sec and I usually tell people if it makes it to 6mm^3/sec it's probably fine. It might be you were near the end of the spool for that color PLA? I found that I have to print at about half speed for the last 1/3 spool because of the strong curvature as it passes through the final (unmelted) section inside the head - both for UMO and UM2. There are many many causes for underextrusion - I would start by putting the filament on the floor instead of on the back of the machine. Anyway here are my recommended max printing speeds - these are about half of what the UM can do so this gives you a safe margin. You are going over these values I think. This is for .2mm layers - for .1mm layers you can print twice as fast: 20mm/sec at 200C 30mm/sec at 210C 40mm/sec at 225C 50mm/sec at 240C

I'm not sure how many meters of filament fit on a typical spool - but I've noticed that after I am down to the last 1/3 of a spool I have to print much slower - about half the speed - due to the curvature as it passes through the print head. Both for UMO and UM2.

50mm .2mm layers, 210C is too cold or too fast. You are getting intermittent underextrusion, probably when the filament is slightly stuck on the spool as it is probably "on the edge" the whole time. I recommend you increase temp to 240C if you are going to print thick layers at 50mm/sec or print thinner layers, or slow down the printer, or some combination. You will also get better quality at slower speeds - I would actually recommend 35mm/sec, .2mm layers, 220C. Here is a table of max recommended printing speed at .2mm layers for PLA (for .1mm layers you can print twice as fast): 20mm/sec at 200C 30mm/sec at 210C 40mm/sec at 225C 50mm/sec at 240C

Actually I think I thought about this wrong. If you put in 128 what really happens is it is at 24V for half the time and 0V for half the time. But the power is V*V/R. In other words it is proportional to the voltage squared. So actually that means it is at 100W half the time and 0W half the time or an average of 50W. That's too much. So better to go for a PWM setting of 64 which is 24V for 1/4 of the time and 0V for 3/4 of the time. The volt meter may claim this is only 6V on average - which is true - but it's still 25W on average. So I think you may need to set it to 64 for a max setting. This non-linearity issue will affect PID values also in a not-clear way. I would probably just cut the PID values by about 1/4 or 1/3 and try those initially.