Jump to content
Ultimaker Community of 3D Printing Experts


  • Posts

  • Joined

  • Last visited

  • Days Won


Everything posted by geert_2

  1. Do you mean that you edited the nozzle temp, bed temp and fan settings on the UM2 itself, and saved it there? This indeed does save new settings in a name like "custom+number". In that case, on the printer, you can go to: materials > settings > save to SD card (or something similar, I don't know the exact wording by head). This saves a file "material.txt" to your SD-card, which you can edit in Notepad. When editing this file in Notepad, I would suggest you change the material's name "custom4" into something more meaningfull. I use names like "PLA200" for pla printed at 200°C (PLA is always printed with 100% fan). And "NGN225F" for NGEN, printed at 225°C with fan. Or "NGN225NF" for no fan. And so on. In this way I can use only one gcode file with various materials and fan settings, by simply changing the material. There appears to be a maximum length for the material name (8 characters?), and it seems that it does not allow spaces and underscores. So I use capitals and numbers only. This is just from experience, I don't know the official specs. Maybe one of the developers can point towards that info?
  2. On one of our UM2 I also had this a few times, but it was always in dry winter weather, when I touched the printer frame and got an electrostatic discharge with spark. It was on those cold dry days when I would get a shock and spark when touching almost everything. Sometimes the sparks were up to 10mm long, so that must be several 1000 volts. This discharges seems to interrupt something in the printer; it could be a shielding or grounding problem. After a power off and on, everything was fine again. Your problem could be totally unrelated of course. Does it stop when you are not around at all (then probably unrelated), or when or shortly after you touched it (then maybe related)? And in what weather or environmental conditions?
  3. Normally you should not need rafts to print. I even never use brims or "mickey mouse ears". If required for overhangs, I design my own custom supports into the model. So I would suggest you just try leaving out the raft. I don't know the gcode commands, but would it be possible to manually add a "pause" command in the gcode, at the correct spot in the model? Then it would always be at the right height, regardless of any raft or other support stuff used. Maybe one of the UM-software specialists could help you out here?
  4. I don't know the MP Mini either, so this is a *very wild guess*. But to me this looks a bit like you are using a 1.75mm filament in a 2.85mm printer? However, if you would be using the correct filament, I would indeed look for a temperature/cooling problem somewhere in the system, like Nicolinux suggested? Or some piece of tubing or coupler that is missing or mounted incorrectly (similar to the white teflon coupler in Ultimaker printers), which is supposed to contain the molten filament?
  5. I try to simulate that in the design by adding a small rounding. This works somewhat, but is not optimal. The problem is that this approach is not compatible with desired roundings at the bottom. And it gives a bit of deformation of the first two or three layers. So indeed a solution in the slicer to reduce only the outer edge (but not the infill, so we still get good bonding) of the first layer with a user-selectable distance would be very nice. Good idea.
  6. If it is in the gcode, I guess you should be able to see it in Cura when - in layer view - zooming in quite a lot. The "jumps" should be visible. Concerning the "salt method" for bonding, see the full manual (PDF-file) with pics at:https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/
  7. Instead of using water to glue them, what about a thin layer of wood glue, or PVA-slurry? First try manually with a few centimeters if that works. Also, you could design your own tool to align and glue these filament ends, print that, and clamp it in a vise or so. Just two simple plates with some 2.9mm diameter indentations, and some alignment features, will do. It would be an interesting experiment.
  8. I have no idea about the shifting, as I have never seen that. I guess you already took a look at the "layer view" in Cura, or in any other slicer, so it is not a defect in the gcode file? Apart from that, you also seem to have "warping" in the corners of the models, so your bonding to the build plate isn't optimal. This could cause the models to come loose during printing. Depending on your preferences, I would suggest you try any of the bonding methods commonly discussed in this forum: dilluted wood glue (1 part glue in 10 parts water, as promoted by user gr5), hair spray (spray it on a tissue, then wipe the glass plate, as promoted by neotko), 3D LAC spray bottle, or for PLA-filament my "salt method": wipe the glass plate with a tissue moistened with salt water, and gently keep wiping while it dries into a thin almost invisible mist of salt stuck to the plate. Or if you use the standard glue stick, smooth it out with water after applying.
  9. I use those left over ends to do atomic pulls at regular intervals. If you would still have more left over, what about heating them with a heat gun and making toys with it for the kids? Arm bands, rings,...
  10. Hey cloakfiend, A good starting point for searching primers might be car paints and primers for plastic parst, like for wheels covers and bumpers: they need to withstand a lot of abuse, sunlight, rain, stones,..., so they need to adhere well to various sorts of plastics (but mostly to ABS, PU, and polyester). For optimal adhesion, they should chemically bite into the plastic and chemically bond, not just cover it.
  11. Maybe you need 6 slots... User neotko has good results with hairspray (spray it on a tissue and wipe the plate). And user gr5 has good results with dilluted wood glue (1 part Elmer's glue in 10 parts water, if I remember well). Still other people use dissolved old PVA-support material from the UM3, thus recycling what would otherwise be waste. My own "salt method" does *not* work for ABS, so no need to try it. However, for PLA and some other materials it can be used. (=Wipe the build plate with a tissue moistened with salt water, until it dries in a thin, almost invisible mist of salt stuck to the plate.)
  12. I would suggest you have a look at DesignSpark Mechanical. This is a limited version of SpaceClaim, and distributed for free by RS Components (big electronics supplier). It only requires registration to activate it. Its user-interface is similar to that of SketchUp, with pulling and pushing on sides or edges of models, to modify them. But it generates good STL-files for 3D-printing. Main disadvantage: import/export module for STEP and IGES standard 3D-formats is optional and not free (actually quite expensive). On Youtube you find a lot of demo- and instruction films. Also on the RS-website and SpaceClaim website, you find tutorials. Have a look if this is something for you. Concerning Onshape: I tried that, but I can not get around in their user-interface: I find it all mixed-up and confusing. However, other people have made wonderful things in it, so it works. Technically, it is incredible what they can do in a standard browser, but I do not like the online-only philosopy: it reacts too slow, and no internet (or server down) means no editing. I already had an account before the recent limitations, and last time I checked, I still seemed to have the ability to store 9 private models, invisible to others. Yes, 9, not 10 as in the beginning: for some unknown reason one dropped off.
  13. When pushing filament manually through the whole feeding traject, I found the most friction occurs in the teflon coupler and nozzle, when near the end of the spool. Friction in the bowden tube can also be much higher than expected. Also the "unwinding force" is very high: the spool acts as a very strong spring, trying to wind-up the filament again. It is probably the sum of these that causes the problems, when using the original feeder. With the spool normally mounted on the printer, and the filament being fed into the machine, I usually unwind 20cm of PLA, wind it in the opposite direction around a 7cm skater wheel, and release it, after which it is almost straight. Then I unwind the next 20cm and wind it around the skater wheel in the opposite direction, etc., until I have straightened a few meters. This only takes a minute or so. And it can be done easily while the printer is printing. Then I loosely wind the filament up on the spool again (although this is not necessary, you could leave it hanging around), so now it is sitting very very loose on the spool, with a bending radius of about 50cm, and it has very little friction in the whole feeding traject. Then it does not cause underextrusion anymore. I know, I should try another feeder. Actually, I wanted to design one myself, but I haven't come to that yet.
  14. To me this looks like underextrusion. If it is only in the left back corner, I would guess it is because the friction of the filament in the bowden tube, teflon coupler and nozzle is higher when bent more in this corner. This especially when printing very hard filament like PLA, and when it is near the end of the spool. Maybe it could also be uneven build plate calibration (too much distance at the left back). This you can visually check when starting a print: is the first line equally thick everywhere? If it was the teflon coupler worn out, or a dirty nozzle, I would guess the problem would occur everywhere? If the cause would be too much friction, try to unwind and straighten a bit of filament manually prior to printing, and see if that helps. I usually unwind 20cm of PLA, wind it in the opposite direction around a 7cm skater wheel, release it (after which it is almost straight), then wind the next 20cm around the skater wheel, etc., until I have straightened a few meters. This only takes a minute or so. Then I loosely wind it up on the spool again, so now it is sitting very very loose, with a bending radius of about 50cm, and it has very little friction in the whole feeding traject. Then it does not cause underextrusion anymore. Of course there might be other causes, but these are my first guesses.
  15. I agree with kmanstudios: heat may be a bigger problem than water (to test water-resistance, I have made a filter for the sink in PLA a year ago: it hasn't rotten away yet). Just don't use water-soluble materials like PVA. In addition to this, the flexibility or lack of it may also be a problem. Heavy clay or sand will deform a flexible mould due to its weight. And a non-flexible mould of course is difficult to remove after casting. But you know that for sure. What about printing the original model (not the mould)? And then make a silicone mould around it, and a plaster or epoxy shell to stabilise the silicone? Or use a high temp rigid material to print the shell. In this way both water and temperature are no problem. Most silicones can handle up to 200°C. On Youtube there are lots of good videos on mould making and casting.
  16. If you have a highly directional microphone, or a modified stethoscope, that might help in locating the exact source. I have a standard medical stethoscope (like those the doctor puts on your breast and back to listen to heart-beat and lungs), but that is not accurate enough. It does not differentiate enough between sounds from various parts. However, car technicians use the same basic stethoscope setup, but they replace the pick-up element with a thin long tube. So in a car engine, they point the tube towards various parts (valves, waterpump, oilpump, alternator,...) and they can locate which parts or bearing makes weird noises. If you would have access to something like this, or if you could build one (print the connectors and ear pieces, and connect some standard PVC tubing to it), this might help locating the source?
  17. I highly disagree. Slow travel path for combin moves will cause the drip to increase and unless someone makes a math formula to actually know how much drip/mm/heat/viscosity is lost, you will get a bit of underextrusion when the extruder resumes to print. Hi neotko, From a theoretical viewpoint I do understand your concern. But in my tests, I haven't noticed this problem yet. But I will look out for it in the next prints. Practically, for PLA I usually use the default speeds of 150mm/s travel, and 50mm/s printing, which gives good results. Here, a low traveling speed would have little benefits. But for transparant PET materials this high traveling speed sometimes causes a dull line, thus an ugly opaque line in the otherwise "frosted glass" transparant appearance. So, here a slow traveling speed gave me a much better result, since the oozing-line was now also transparant, no longer dull opaque. PET is more sticky and "rubbery" than PLA, and tends to build up on the nozzle (at least the ones I have). So, when moving fast, the "morse-code" that is sometimes left behind, accumulates on the nozzle, burns, leaks, and causes these ugly brown spots. Here too, a low traveling speed outweights the disadvantages. At least for me, but other people have the right to see things differently, of course. So, for PET and co-polyesters, I now use the slow traveling speed, identical to the printing speed. For PLA, I use a high traveling speed. So, in any case, this should be a user-selectable option, and not be enforced.
  18. Concerning the new warranty: Yes, I think the wording is now reasonable and much better: first, it is more clear, and second it reflects your "open philosophy" better. Tanks. (Edit: clarified the context of this reply)
  19. If the nozzle uses roller bearings (I don't know for sure), could a broken ball, or some debris that was picked up, cause this sound? In my case, the sound came from this belt rubbing against the flange in the photo, but then on the other side of the printer, left back side, where its home-position is.
  20. On one of my UM2 I had a weird sound coming from the same corner, left back. But it was rather a squeeking sound, like rubber tyres, or like chalk on a blackboard. It was very hard to locate, as it seemed to come from everywhere around that corner. Eventually I found it came from the rubber belt (the one from the stepper motor) sliding agains the flanges of its wheels. Rubbing the side of the belt with silicon grease solved the problem. I used the thick, inert white grease that is also used in microscopes and binoculars, *not* silicon oil, and definitely not petrochemical oils (may damage the rubber). I found it by manually pushing the belt sideways while printing a small piece, or while moving the head by hand. Your sound is different, but maybe it could come from the teeth of the belt rubbing against the flanges?
  21. This is what I would rather expect from an atomic pull: see the nice orange one at the bottom. It has still a little bit of white, since I was changing colors from orange to white and back. This is from an UM2 (non-plus) with standard 0.4mm nozzle, 2.85mm filament. On an UM2+ with removable nozzle you might see some additional lines, but the general shape should be similar:
  22. If your model and function allow it (we can't see the inside here), another option might be to cut it in parts, and glue them back together again after printing. Or design some holes for hex nuts and screws, and mount it with screws. This might go a lot faster than other post processing, and might be good enough.
  23. I was about to say that neotko *does* promote hair spray and has a video of it. But he was faster. I use the "salt method": after cleaning the glass plate with water, gently wipe it with a tissue moistened with salt water. Gently keep wiping while it dries into a thin, almost invisible mist of salt stuck to the plate. For me, this gives excellent bonding of colorFabb and Ultimaker PLA, and still good but not perfect bonding of ICE PLA. When hot, it bonds like cyanoacrylate. But when cold, the models come off by themselves and are sitting totally loose on the plate. For the full manual with photos, see the PDF file at: https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/ For ABS this does not work at all: it starts foaming and splattering, I don't know why (maybe because of the water in the salt?) For PET it works a bit: it reduces bonding strenght a little bit, but makes it much easier to remove the models after cooling down: they don't get stuck to the plate anymore. I also tried the dilluted wood glue method for PET: it gave a super bonding, but once tore a piece of glass out of the bed. So I am a bit reluctant to use that, if not absolutely necessary.
  24. For some unknown reason, my reply was eaten up by the system and didn't show up on the page. So, let's try again. I use silicon mould release oil: spray it on a tissue and wipe the nozzle. This reduces build-up of molten filament on the outside, but does *not* eliminate it. It depends on the material: it works less or more for PLA, but does not work well for PET, which is much more sticky and "rubbery" when molten. I also tried PTFE oil, which also helps a bit, but less than silicon oil. I haven't tried any other oils. Some time ago, there was a thread in which this was discussed, if I remember well it was about putting silicon socks around the nozzle, for the same purpose.
  25. I think it is a good idea to restart from fresh, in case of hard to solve problems. To me, this looks like underextrusion. If the filament is rather hard, as in the case of PLA, and it is bent into a tight corner on the end of the spool, try to manually straighten it. This will give far less friction in the nozzle and coupler. I usually wind it in the opposite direction around a skater wheel of 7cm diameter. Also, I would try the following things: First remove filament and do an atomic pull to clean the path. Then manually insert a piece of filament (with cold nozzle) to see if it slides smooth through the nozzle and teflon coupler? Then insert the bowden tube, and manually slide a longer piece of filament from the back of the printer through the whole path bowden-tube, coupler, nozzle (with cold nozzle). Is this going smoothly without friction? Then heat the nozzle, and manually feed a bit of filament through. Does it come out smoothly without too much friction? Then start a small print (e.g. a cube of 10mm x 10mm x 10mm), and manually feed the filament while printing. If you would have a similar printer available, you could try the exact same things there, to compare. If all this goes well, then at least you have an indication that that part of the printer is reasonably okay.
  • Create New...