Jump to content

aroth

Dormant
  • Posts

    61
  • Joined

  • Last visited

aroth's Achievements

1

Reputation

  1. Time for one more round of necromancy. I just updated my UM2 firmware for the first time ever, after installing Cura 2.5.0 and having a print fail 3 times in a row at exactly the same spot. I thought that maybe there was some issue between the new Cura and the old firware version, so went ahead and updated the UM2 (and seems I may be right, as the print is just finishing up now). Like others in this thread, I'm using Venkel's fan mount, and had the printer set to run the fans at 100% after the initial layer. The first print I tried after updating the firmware failed almost immediately after the fans switched on, with the "Heater Error (ER03)" message. After seeing this thread I retried the same print with the same settings, manually dropping the fan speed down to 50% as soon as they switched on. That did the trick, with no other fiddling (or a factory reset) required. I also noticed that at 100% fan speed, the printhead temperature can drop by at least a solid 15 degrees. At 50% it holds steady on its target temperature. Anything above 60% on the fans and it starts dropping. Guess now I just have to remember to dial the 50% fan speed into Cura every time.
  2. In my experience even the new ones will deform under normal usage. Though judging from Dreamworker's photo I'd say his coupler is barely deformed at all. Mine end up looking far worse: Though in any case, I think a deformed teflon part is a reparable issue. And probably also the most common problem to occur with long-term usage. I was recently able to get a nearly-crippled (mostly due to problems with the teflon bit) printer back to like-new better-than-new performance by doing: Remove the filament from the printer. Disassemble the print-head and the entire hot-end except for the temperature sensor. Turn the nozzle upside-down and lightly sand the top surface of the threaded portion (which is now at the bottom) with some high-grit sandpaper to loosen/remove any accumulated dirt/filament. You want to have a clean, flat, smooth surface here. Clean the threaded portion of the nozzle using acetone. If necessary, remove any lingering bits of debris from the threads on the nozzle using a dissection probe/pick or similar tool. Repeat the cleaning process on the interior of the metal isolator part. Verify that both parts are properly clean by screwing the metal isolator back onto the nozzle; you should be able to easily hand-tighten the isolator until it's sitting flush against the heater block. If you can't then one part or the other is not clean yet. Drill out just the scorched bit of your teflon part. I did this using a spade bit like this one: I placed the teflon part vertically on a level surface, with the scorched part at the top, and then inserted the spade bit and gently rotated it by hand until the lip that had formed in the scorched area was gone. Get a bamboo cooking skewer, and bring along your sandpaper, level surface, and teflon part. Put the sandpaper down on your level surface, then stand the teflon part on it vertically, scorched side down. Insert the skewer (blunt side down) into the teflon part, and then gently press down and rotate the teflon part while holding the skewer to keep it vertical and in one place. As with step #2, your goal is to get back to having a clean, flat, smooth surface here. Reassemble your hot-end and print-head. Try to get the metal isolator part as tight as you possibly can (you won't get it all the way down to the heater block when the retaining screw is in place) to ensure that you'll have adequate tension to create (and more importantly, retain while printing/during retraction) good contact between the teflon part and the nozzle. Bad things happen if you allow even the slightest gap there. Relevel your buildplate, clean the nozzle (atomic method), load some filament, and print. I'm not sure how much (if any) of that is officially recommended, but it worked for me. The part I was most concerned about was sanding around the nozzle, as I didn't want any dust to get inside and clog it (hence turning it upside-down to encourage the dust to fall away from the nozzle instead of into it). Doesn't seem like that caused any issues, however. And having a clean, flat surface there seems fairly important. Took a bit of time to work through all of that, but still faster than waiting for a new teflon part to be shipped to me every couple of months each time one starts getting scorched/deformed. I wonder if it would be viable to fabricate a comparable part out of glass/metal and then just hit it with teflon spray every once in awhile to keep the friction down?
  3. To adjust the left and right corners, turn the screws underneath the build plate, not the button on the front of the Ultimaker2. I'll try to say that in French: ----- Pour faire l'ajustement des coins gauche et droit, il faut tourner les vis au dessous du plateau, pas le bouton rotatif en face de la Ultimaker2.
  4. Yes, it's very nuanced. In general 20% infill will give you a part that feels quite solid. I usually print around 10-15% (when I can't get away with 0%). The amount of infill you specify will generally have a linear effect on your print duration; 50% infill should take half as long as 100%, and 20% should take one fifth as long, etc.. You can also have the printer do infill at a different/faster rate than the rest of the part. I usually set a 75mm/sec infill speed, with the overall print speed set to 50mm/sec or less. Layer height also has a linear effect on print duration, such that a print using a 0.2mm layer height will take half as long as one that uses a 0.1mm layer height. So you should be able to bring that 90 hours down quite a bit by adjusting some of the print settings. 90 hours may be doable, but it's stretching it in my opinion. The problem's not that the printer can't handle 90 hours of constant operation, but you only need one small thing (such as having a piece of dust/hair clog the nozzle, or the filament spool getting snagged) to go wrong and the entire print will fail. My longest successful print to date has clocked in at 40 hours, for a part that maybe used about 50% of the available build volume (if not less).
  5. Have you tried decreasing your layer thickness or increasing your top/bottom thickness? You're only giving it 3 layers of material to work with there. If you've got good internal supports/lots of infill that might be okay, but I generally find I need at least 5-6 layers (and preferably more) to get a good looking top surface when printing over a hollow portion of a part. What usually happens is that the first layer will have some gaps/issues/broken bridges, and then those gradually get corrected on subsequent layers.
  6. It's difficult to take good pictures of the printer's surroundings when the printer is still on.
  7. Yes, I think you're right. I gave it a second try, and this time I watched as it printed the skirt and added some extra glue in each spot that it placed a support. Worked better: Though I do think that Cura should be more clever when it issues a 'G0'/travel command and either 1) route the printhead path around parts of the object that are in the current layer as opposed to straight through them, or 2) lower the buildplate slightly before issuing the G0 and then raise it back after the travel completes.
  8. If you're already leveled, you should be able to just click through everything except step 1 (or 4) of the leveling wizard (where you move the entire buildplate vertically). Just move the plate to the right height, don't touch any screws, and you should be set. You probably have to do this regardless since I assume that the Z-offset from your previous level would have been lost when you reset the printer anyways. Not sure about the filament loading, but I'd think probably you can just cancel out of that step when you get there? If not it seems like anyone who happens to reset the printer settings with filament still loaded would be kind of screwed. Though you could probably just click through the filament insertion too. The worst that should happen is that the extruder will either tick back a lot or grind the filament when it tries to forward material that's already been forwarded. I've had a similar thing happen a few times (as far as I can tell because there's some sort of firmware bug that triggers if you move/swap filament often without printing or restarting in between that causes the printer to start forwarding new filament too far when it loads; so now I always turn the printer off and on before I swap filament), and although the sounds the printer makes when it tries to forward material that's already at the printhead aren't pleasant it doesn't seem like any damage occurs. Or if damage does occur, it's to the filament rather than to the feeder.
  9. Successful 40 hour print (0.12mm layer height, 35mm/sec shell speed, 75mm/sec infill speed, 10% infill because I needed something to support the bottom of the eye sockets otherwise would have done completely hollow): Worked on the second attempt, after I installed my super high-tech solution to the 'dust on the filament can clog the nozzle' problem:
  10. Probably better tools already exist for this, but here's a web-based tool for inspecting the GCode generated by Cura: http://jsfiddle.net/h3wbf9g4/8/ It'll let you view individual layers (extrusion is green, travel is blue, and retractions are red) and also choose a specific layer/set of layers to export (for instance, if you want to http://umforum.ultimaker.com/index.php?/topic/6802-2-ways-to-resume-print-from-last-layer/).
  11. I get comparable results with a damp paper towel. Clean glass gives more adhesion than you need, most of the time: Though I still put a bit of glue under any areas that I'm particularly concerned about (such as corners, or support structures or similar detached 'islands' that exist at the start of a print).
  12. Anytime in the near future? No. Eventually? Yes. I see it as similar to the difference between color laser printers and traditional inkjet printers 15-20 years ago. One offered significantly better performance and precision than the other, but also tended to cost substantially more to purchase and operate (thousands of dollars versus hundreds). Fast forward to today, and you can get a good color laser printer for not significantly more than the price of a good inkjet printer. I expect 3d printing to go the same route. From a purely theoretical standpoint, I think there's no question that SLA is a better approach than FDM in terms of performance, and it's main drawbacks are simply cost and process issues (SLA components are expensive, resin is expensive, resin is messy, resin requires extra curing/finishing steps, resin is harder to clean etc.). Given enough time, I think technology will solve those problems and we'll see SLA start to take over. To be completely honest, if it weren't for those problems I'd have gotten a Form 1+ instead of an Ultimaker 2. But SLA just isn't quite ready for consumer adoption yet. In another 10-20 years, however, I expect that to change. I'd be surprised if I didn't end up with an SLA printer sitting on my desk at some point within that timeframe. As much as I love my Ultimaker 2, I have to concede that I'm pretty much on the equivalent of an old-fashioned dot-matrix printer at the start of the 1990's, and that after a few generations of 3d printer technology its capabilities are probably going to start looking pretty sad.
  13. I second that recommendation. Though I'd say fan shroud is fine, what's under-designed is the nozzle placement. What I think Ultimaker should have done is this: ...instead of this: Then there'd be symmetrical cooling with the stock fan shroud in both single- and dual-extruder setups. I'm sure that change would cascade through all of the printhead parts, however, and probably also have some implications for Cura.
  14. I'd also suggest slowing the printer down when it resumes, to make it more likely that the first new layer will adhere to the existing part. After the first few layers are down, it should be okay to turn the speed back up. And is there any way to compensate if the new layers are shifted out of position when the print resumes? I was having a nice and successful 40 hour print: ...up until this thing (I think it's a piece of fur from the cat) managed to make it all the way to the nozzle and then decide that it wasn't going to go any further: I managed to track down the right layer to resume from and edit the Gcode, but the new layers are all shifted back a few mm from where they should be: Is there some way to correct that sort of thing when editing the Gcode?
  15. Yes, the layers looked okay in Cura.
×
×
  • Create New...