Jump to content

Whom

Dormant
  • Posts

    22
  • Joined

  • Last visited

Everything posted by Whom

  1. Hi everyone, Our S5 died recently, and it's taken a while, but I figured out why. Save yourself the hassle now, check two things. Check the heat shrink on the print head capacitive sensor. Replace if worn. Check your nozzle heater wires aren't sticking down. And if they are, check if the insulation is burnt/worn. Bend the wires up out the way. The issue itself if best shown with Ultimaker 3 CAD files, see below. Obviously the tab is bent on the machine, but if the wire sticks down, it hits the tab. The wire is hot, the heat shrink wears away, 24V from the heater fries the PCB on your print head. Auto levelling fails, no way to turn it off, machine is out of action. I suspect this is a known issue, you can see below the tab direction and the inclusion of heat shrink is a design modification. Picture below taken from UM own website on diagnosing levelling sensor issues. Our 18 month old, lightly used S5 now needs a new print head PCB, I had to diagnose this myself as UM support just wanted me to send in the whole print head. Of course, this is out of warranty, so an unwelcome time and monetary cost. Save yourself the hassle, stop this before it eventually wears out the janky insulation. Please let me know if you've suffered the same failure. On our machine it was the U8 chip that let out the magic smoke.
  2. It really wasn't obvious until I got my macro lens out that U8 was blown. I was tracing the sensor pins around and stumbled upon a tiny darker area next to U8. If it is just a protection diode, I wonder if removing U8 and running without it might get us up and running until a new board arrives. I'm hoping Ultimaker foots the bill, this feels like a design issue to me.
  3. I did some more digging. The insulation on those wires is just a braid I think, so anything sharp can worm it's way into the metal. I also loaded the CAD and it is pretty obvious the tab lines up with the heater wires... I also found one of our cores has marks on the heater wire and the wire sticks down into that tab. Pretty disappointing on a 18 month old machine. I will make sure the wires on all our cores are bent away from the tab, and keep a vigilant eye on that tab insulation heat shrink, the jankiest of "fixes" I have seen on something this expensive.
  4. I have been trouble shooting the same problem and found that the U8 chip has clearly let out the magic smoke on our machine. See attached photo. I presume it was caused by the same issue, but haven't noticed worn through wiring on any of our core heaters. I'll check again. Seems like, from the inclusion of that little bit of heat shrink, this was a design error on the machine.
  5. No I modelled the PLA and PETG parts manually, this is my own design (because I just had to fiddle!) 1mm is probably more than you need though?
  6. I've changed it a bit now as we're up to 3 printers running. A UM S5, UM3 and then the Mark2. Hilariously my Mark2 is doing better than the others hahaha. We're churning out over 100 a day. There are 2 layers that have PLA in them between each headband. But the second is just the outer edge so that the PETG gets laid down inside this. See image. It is super sensitive to the PLA you use, some stick too well, some not at all, some just peeerrffeeccttt.
  7. Overhead Projector Sheets. I spaced nubs on the shield for a standard North American 3 hole hole punch
  8. I can never go back to single extrusion example number 4568 Printing face shields for local healthcare workers here in Canada 40 minute print, can't print at night, high manpower demand flipping prints. Realised that PLA and PETG don't stick very well. Modified my design to stack with a single PLA layer inbetween. Bingo! Pumping out faceshields in easily separable stacks. Mark2 FTW
  9. My personal solution to all of this was to set a tiny tiny piece to be printed with nozzle 1 which takes 2 seconds, then the main part to be printed with nozzle 2 for the rest of the print
  10. Of course, I attached it but added PDF to the end else it would not upload. I also have a custom start wipe procedure that copies the UM3 and poops a blob directly onto the glass. You have to replace the Mark2 file in definitions so remove the 4.5 from the start of the file name. Let me know how you get on. It's pretty finely tuned to my machine so if you're a few mm different in your home position you'll have clashes. 4.5 Mark2_for_Ultimaker2.def.json.pdf
  11. Amazing, thank you. I've got some homework to do! I think I'm going to find some higher temp ones, perhaps a different shape and remodel the heads for my "heavy duty" needs. Yeah I was worried about it no longer being a 'tripod'. But since I stiffened up the coupler with side bars it's basically two flat planes and since I used the head to set the position of the coupler magnets there doesn't seem to be any misalignment gaps (yet). But I'll keep an eye on it. I aligned the slots with the center of rotation so it drops into place nicely.
  12. Ahh that makes sense, I really should have realised this. I miss the absolute value but only because as I switch back and forth between pla or abs if I forget to update the material settings in cura I can fix it when the print starts. But with the "temperature magic" I think you called it I understand why you switched to an offset. In a perfect world the offset limit could be more than 25 for the heated bed. Or I could just learn to check by slicing settings haha 80! Of course, that would have done it then. I was assuming all could manage 100+ like an RC motor but I'm pretty sure I hit 80 at least once in the enclosure. The enclosure I build was too well insulated, I've ordered some fans and control boards to make it regulated. I'll go teach myself about magnets and order some more. Thank you! I hope nobody is offended by my ugly modifications haha
  13. I thought I'd post another update as I've worked through a few issues specific to my use cases. I've been doing lots of ABS prints in a hot enclosure (70C+), and it seems like my magnets are getting weaker, losing alignment or I don't know what. I was never able to reach the furthest corners without the head popping off but it got worse and worse. I modified the head and the coupling to include an additional 4 magnets and printed everything over 11 hours with a 0.25mm nozzle in ABS (fusion+ from Innofil/BASF, it's pretty good!). I also beefed up the coupling to also act as a spacer for the head assembly. What I was finding was that the original design Z height of the coupling was dependant on the top height of the plastic part of the head. And this varied depending on how tight the screws were etc. My modification (which is pig ugly) is captured in the metal part of the head assembly and so is more accurately set over time. It is rock solid now, I think my magnets were crappy but now I've got 7 of them I think there is a nice safety factor. Also just found a great trick using "terrain to STL" and switching the shell extruder to a different colour for contour lines. One quick question for tinkergnome, why is it that you've set the maximum temp adjust to be + or - 25 in the Mark2 firmware but not in your normal UM2 firmware? Thanks again for the epic work everyone, I could never go back to a single extruder now.
  14. I don't know why I keep writing here but it feels like the right thing to do incase it helps anyone. I did switch nozzle 1 over to the other side, which has increased my dual build area. The offset in CURA doesn't apply to the gcode is creates, just to the disallowed areas (I think there is a setting in the json file to make this true or false). I think I will keep tuning the build area, there's a few mm all round to gain. The only thing I can't solve is to get the area in CURA to be shown on the correct part of the glass in the software. But it's the right shape, and it prints fine, just shown too far to the left. It was trivial to calibrate, I created my own gcode as tinkergnome suggested for the calibration, and set offset in X to 0 on the printer and it calibrated easily. There is no clashing between bowdens or wiring. And I have removed the fans from the second extruder, for my purposes I don't mind if the cooling is substandard on nozzle 2. I also found that the default start Gcode was retracting extruder 1 back twice, but pushing the filament back forwards only once resulting in the first 20 seconds of printing giving out no plastic. And the problems I was having with the initial priming poop catching on future nozzle priming I solved by moving the initial priming poop. I also copied the UM3 style and made it dump plastic right up close to the glass, now I don't have to monitor the first 20 minutes of every print as it doesn't have loose poops flying all over the place. Thanks again @tinkergnome and @yyh1002 If anyone has any questions, please feel free to ask!
  15. Ok, I don't know how to quote properly so I'll just respond as best I can! I've been playing with the disallowed areas, and found a few interesting things. - it is possible to crash into Head 2 when using Head 1 using the default .json file from github - under my circumstances, I can have an offset applied in CURA and on the printer and CURA is NOT applying the offset to the gcode. I don't know why your finding was different tinkergnome? - using the default ultimaker profile, with the second head removed and parked outside, could still result in a clash with the internal parking thingy. I made myself a new .json file, and I upped the bed size to it's physical size and I created an excel sheet to allow me to quickly edit the areas then copy them into the json file. I also have the nozzle offset in CURA, but I can only apply the same shape disallowed area to both nozzles, but it's pretty close so not too much lost. I'll compare everything below. Original Mark2 .json file My new .json file, using only extruder 1 My new file using extruder 1 and 2. And below is the actual prints. Blue is the original file disallowed areas, white is my new using only extruder 1, and black is using extruder 1 and 2 on my new file. You can see everything shapes up nicely between cura and real life in terms of the glass usage. And you can see the room I gain when I don't use extruder 2. I think I'll make another .json file for when I have the Mark2 nozzle removed and parked on the outside. I am also still considering if I should move nozzle 1 over to the other side, making the X offset 0
  16. Thank you for such a speedy and helpful response. It may seeeeem like I know what I am doing.... butttt..... no. I realised that if the extruder 2 offset is set in cura under machine settings, then it automatically shifts and applies a second disallowed area when you have a dual extrude print set up. If that doesn't make sense, I'll add pictures at a later stage. Unfortunately, it doesn't seem smart enough to differentiate between disallowed area for extruder 1, and extruder 2, but rather applies the whole area as disallowed for both regardless of what extruder is being used on each part. I think I might try shifting nozzle 1 over to the other side, as I have the ability to do that with my parts. My thoughts are to try and align the disallowed areas as closely as possible. I guess then I can simply adjust the X offset to 0 in the settings, then print the amazing calibration gcode? Then I can map out the disallowed areas using the "expert" move axis feature (which is awesome by the way). And then edit the JSON file. I am also considering removing the side fans from the extra head to gain some room and simplicity. Hopefully enough air makes it over from head. I'll update with my findings once I've maximised my print area, if that's useful and appropriate for this thread?
  17. Firstly, wow you guys! I sent you some beer money, but it wasn't so easy to find the donate button. Make it bigger and flashing? Love the work, love the thought and care and awesome solutions to all the problems. Like angling the head so the nozzle lifts etc etc. I do have a couple of questions, and you are surely the best to help me if you have a minute. 1 - "machine_disallowed_areas". I was going to ask how to change these areas, but then I dug around inside the .JSON and spotted them. I am trying to maximise the print area, so I ran the heads around while "moving material" to mark out the limit of their ability. You can see my traced results in the attached image. It looks like my end stops might be a little away from normal, but it also shows that I could end up printing in areas I shouldn't be. Can I set extruder specific disallowed areas? e.g. a big print, but only one corner requires the second extruder. I was to position it such that part of the print is outside of the area for head 2, but that area doesn't use head 2 so it's fine anyway. Can I adjust these areas inside CURA without editing the JSON file? Black - Original Nozzle Red - Mark2 2 - extruder 1 position Kind of related to the first question, what stops me from putting Nozzle 1 on the other side of the main head in the spare hole, so that it is on the same X position as the second nozzle? The benefit of this would be matching up the X limits, so that a bigger dual colour print would be possible? I realise that the bowdens miigghhttt clash, is that the reason why this wasn't done? 3 - gCode flavour I did search for this too, and understand that the gcode flavour being NOT UltiGcode is related to the temperature profiles? You get more control using cura settings than the machine settings? That popup about overriding machine settings sure is annoying though haha! 4 - Priming, wiping etc I can't seem to get a consistent and reliable prime when switching nozzles, if it primes too much I get blobs that build up and almost cause the head to fall off the magnets, and if I prime too little then I get missing areas on the print. Am I missing a secret trick or do I just have to keep fiddling until I can dial it in for each specific nozzle head/plastic combo? Any info greatly appreciated, and thanks again for all the amazing work.
  18. I fixed it! I tried measuring voltage across the sensor and found that it was 1.3V when the nozzle was in the mV. So I disconnected the power to my (also NEW) heated bed, and the temp worked fine! So I then saw lots of surface solder splatter around the connector on the bed which was shorting the two middle pins. After cleaning that off, it worked. So the sensor had been confused by the short, but not enough to throw any errors.
  19. It got weirder! So I plugged in a spare nozzle temp sensor into temp3 and it then read a temp correctly and all was good, it would even heat up ok. But if the sensor was unplugged an error was flagged right away. So it must be something to do with the sensor, even though I can read the correct resistance from it. Maybe the machine reads the resistance in a different way than my multimeter?
  20. I really hope this isn't simple as I did my best searching and came up with nothing other than the usual heatbed P100/ solder connector issues. So here goes (actually I hope it IS simple) I needed a new motherboard so I bought one from Aliexpress and updated it with the latest firmware. It looks pretty convincing if it is just a knockoff? I was hoping it was from the same production line... The motherboard "temp3" input pins are showing 109 ohms, the same as the sensor being unplugged, which is spot on for the temperature in the room. When I go to "set buildplate temp" it reports 0/0 and it never climbs when I set it higher (e.g. 0/60), and the build plate is not heating. When I go to print, it sits there preparing the print waiting for a buildplate to heat up that has no intention of heating up, and remains at 0 degrees. So what could cause the motherboard to have the correct input resistance, but report 0 degrees? An outright broken board or something salvageable? Thanks for any insight
  21. So.... what you're saying is... there is literally no reason NOT to go Mark2! Thank you.
  22. Hey guys, I've tried to piece together the answer to my question but can't get myself confident so here goes: If I buy an Ultimaker 2 to convert to Mark2, how easy is it to switch between using a single extruder with the full build volume, and the dual extrusion with reduced build volume? Do I simply load the standard Um2 profile in Cura and slice like that? Do I also have to reflash the printer firmware back to standard? And I loaded the profiles into Cura, shown below is what looks to be the reduction in build volume. Have I done this correctly, and if so, why we we lose volume at the left side of the print bed and what's with the little strip in the middle at the front? Thanks everyone and anyone!
×
×
  • Create New...