Jump to content

gr5

Moderator
  • Posts

    17,521
  • Joined

  • Last visited

  • Days Won

    373

Everything posted by gr5

  1. So if you could provide the project file that would be much better. Do "file" "save project" and post that file. It contains your settings which highlight your exact issue. how you positioned the part, your machine profile and much more.
  2. I know him personally. We texted today. He likely posted the files on youmagine.com. Oh yeah - here it is: https://www.youmagine.com/designs/let-s-go-for-a-swim
  3. Oh. So the original poster only uploaded photos. They didn't choose to upload an STL.
  4. Some people only notice that it goes *outwards* at the bottom but you noticed the symptom more correctly than most. The quick answer is "lower bed temp" or "change filament brand" but... Why it happens: PLA (it happens with some other plastics as well) when molten sticks to itself strongly - like snot or mucus. Like a liquid rubber band. It also cools quite a bit in the first few milliseconds while still liquid which makes it shrink. So as the printer is laying down a trace and going around the corner it gets pulled inwards while still liquid. Like a liquid rubber band. On the bottom most layer it is held nicely in place by the bed. But as you go upwards, if the already "solid" filament is still above it's softening temp (it should be! See below) then the solid layers are like a soft clay and get pulled inwards slightly. Eventually you get far enough from the heated bed (1/3 of the way up the indent) and the printer starts to recover and by the 10th or so layer is recovered. Solutions: Lower bed temp. 60C is recommended for PLA. Some people do 70C and that makes the problem worse. PLA softens right about 52C and it's important to keep the bed well above that temp! If you don't then the corners will likely warp up off the bed and the part can even come loose and that can be a disaster - even destroying the print head if you have an S3 or S5. You could try 58C. It doesn't soujnd like a big difference but there is a world of difference between 52C and 60C so going to 58C is lowering it by 25% which is significant. Filament. If you manually extrude some filament and it stays in a thin line - similar width as the nozzle width then that's a good filament. Some brands of PLA thicken just below the nozzle. I'm not certain but I suspect the liquid rubber band effect is especially strong in these brands and you can get that "elephants foot" much worse. CAD - you can model the reverse shape in CAD. In fact the only part of this that bothers me is the very bottom which can be sharp. So I sit "initial horizontal expansion" typically to -0.3 (yes you can do negative values). This makes the bottom of my prints much better. You might say - that's crazy - I shouldn't have to make my shape "wrong" in cad to make it come out right when printed! Well that's what they do for injection molding - the people who create the molds (often low paid engineers in asia) do all kinds of tricks like this (e.g. no corner is 90 degrees such that when it comes out of the mold it then *is* 90 degrees).
  5. Some of these older printers will only read up to 32GB SD cards. Is it possible you got a new SD card? Do you still have the old SD card? I'd try a 4GB card if you have one. This looks like an old Makerbot Replicator which only takes gcode files in x3g format. But it looks like a clone which I would then assume takes gcode but I'd try renaming a file to have the extension x3g just to see if the printer sees the files. The contacts for the SD card can get hair/dust in them. I'd try to vacuum it out with a vacuum cleaner for 2 seconds although it's possible that will create static electricity and destroy everything? I've done it - vacuumed out my SD card - it didn't destroy the printer. But I did it in the summer when humidity was high. Re-seat the 2 ribbon cables that I expect run to the board with the SD card reader (or maybe that was just an ultimaker thing - not sure).
  6. It is possible to print 1.75mm filament without any modifications but... you have to print extra slow - probably 25mm/sec top speed. Because the print head has a 3mm diameter in the hot end and the 1.75 - if pushed too hard - will just start coming back upwards and can get clogged in the upper (cooler) reaches of the hot end. There are 1.75mm conversion kits but really you only need the 1.75mm TFT/TFN part. 3dsolex.com used to sell that - I'm not sure if they still do. aliexpress.com might have some from china although they might not fit very well (those chinese companies aren't great about testing their own product on a printer they don't have). But really - get some 2.85mm filament. The experience is 10X better.
  7. Please post a project file showing what you mean where you set it at 91%. do "File" "Save project as" and then post here.
  8. Me too!!! but if you read the actual Marlin reprap documentation for M84 it says that using "0" disables the timeout feature which is what you want. It's like infinite timeout. Ultimaker uses Marlin but it's a fork where they might have messed with M84 but very unlikely since as far as I can tell M84 is never used normally and probably outside the radar of the developers.
  9. There is also a feature called "lift head" in cura. Make sure that is not checked. If it was then that is probably what is happening on those last few layers. Hover over that feature and read the description.
  10. Another option is to set the Z value of the part in cura so it is mostly below the glass, slice it, print it and glue the resulting top to your existing print. This will be kind of ugly of course. There is a setting in cura preferences (not in the normal settings) that when checked drops all parts flat on the glass so you have to uncheck that before messing with the move feature and changing the Z value.
  11. By the way - before doing any of that - I'd load the gcode from the SD card back into cura to see if the weird movements are in the gcode. Cura has a feature to display gcode files graphically. It could be a bad block on the SD card or a bad slice result from cura.
  12. So to continue a print you have to edit the gcode. The first two (two!!! not one!) posts on this topic are excellent. The G92 is critical. You want to set the E value to whatever is the last E command that you delete such that the new E commands only move the extruder a little bit and so the UM2 doesn't try to extrude many meters of filament. HOWEVER, there is a feature in Marlin that if you do an unreasonable extrude (I think it's over 1 meter?) it will be ignored so maybe you can skip the G92 step occasionally but I'd definitely do that step. That's why the first post doesn't mention G92 because he just got lucky. The E values actually reset every 10 meters or some such limit so it's possible that the E value will be less than the ignore value and then you have a disaster if you skip the G92 command.
  13. Yes but it may take you an hour or so. I guess if you really want to do that then make sure the heated bed stays hot because as soon as it cools down the part will pop off the bed. continued...
  14. In general the closer to the bed, the better filament sticks. If it gets so close that you can barely see the "trace" then it sticks a bit too well and you might break the glass getting it off. Every filament is different in how well it sticks but more squish always means it will stick better. You can verify this while it is printing the first layer. Push up a bit on the bed if it's not sticking at all. I don't know XF-CF20 - I assume CF20 means 20% carbon fill. There are about 5 key things to get filament to stick. One is to clean any oils off the glass (e.g. from fingers). One is to add certain types of glue. One is to get the right temperature. I describe how to get the right temperature, why it matters, and other tricks in this video: https://www.youtube.com/watch?v=t58-WTxDy-k
  15. Ah. So removing the gunk on the *outside* of the nozzle tip helped?
  16. You just need a version *older* than 5.6.8 and if it starts working then you can upgrade the "normal" way using the internet or USB flash drive. Trying to get exactly version 5.6.8 is tricky. The only people who know that stuff are Ultimaker support or resellers. I can help you with that but I think I posted a link to 5.7.3 above. Isn't that close enough? So to get 5.6.8 or older I guess the best way is to click the 9 dots in the upper right corner of this window, then click "support", then near the top do "submit a case". That should hopefully go straight to UM in Netherlands (sometimes it might go to your reseller) where they can point you to where to get the img file you want. Assure them that you will be careful not to electrocute yourself and assure them that you have the R1 with the "robot" on the side and the 9051 id. You might want to include your serial number as well but then they can look up your reseller and might move your request on to them. Link to my website is just a few posts above but here it is again: gr5.org/unbricking/ There are many helpful posts I believe earlier in this thread.
  17. To clarify - ultimaker is written to support "generic" printers and then some of those printer users customize cura a bit for extra bonus options like "printing directly" (whatever that means). Rarely does a chinese company actually contribute anything such as print profiles or machine profiles for their own printers. Sad really since so many of their users use Cura and since Cura welcomes contributions that help Ultimaker's competitors.
  18. interested. I have 4 UM2's with olsson block that could use this. Not sure how it fits with the fan shroud almost touching. And the fan shroud keeps the "cold" air away from the block already.
  19. Oh yeah. Sort of. But it's usually the pulley on the motor that needs it's set screw tightened because that one experiences the most torque (close second is the other pulley on the short belt).
  20. I believe the plugin mentions this: But I believe setting it to zero means the servos never turn off. Did you try zero? The max value you can use is probably either 32767 or 16383. 16383 gives you 4.5 hours so you don't have to sit next to the machine the whole time.
  21. For an S5 it uses the ufp format. Rename it to zip and open that - there is a file in one of the folders that ends in "gcode". Edit that file. It's pretty easy to read. You can search by "layer" although in the gcode file it starts with layer 0 I think so you may be off by one. You can also search for the letter Z which after the first few should only appear once per layer (unless you have z hop enabled). Even better if you used the plugin search for "M84" to find the inserted code. Look at the code that was inserted by "pause at". Get rid of the M84. I don't think it's needed. I'd do a test print first. Check to make sure the servos are enabled by pushing the bed up and down and it shouldn't move. When done, rename zip back to ufp. In windows you probably have to extract the gcode file, then edit it on your hard drive, then put it back into the archive.
  22. I don't know a lot about this issue but it was discussed recently. I would get rid of the "disarm steppers" timeout. Or set it to zero. If it's zero then it should be disabled according to documentation. I have never tried this. Even better, just remove that gcode from the "pause at" plugin. I think this disarm feature was added recently by someone without an Ultimaker printer. More info here: https://github.com/Ultimaker/Cura/issues/15004#issuecomment-1508792641 I don't know but I suspect the issue for you appeared with an update to the pause at plugin and is unrelated to the version of firmware or the version of Cura.
  23. I don't see anything about how to tighten the set screws (aka grub screws) in that list. Probably because it's usually not an issue but... this symptom where you are losing Y position is most often caused by the grub screws thing (but check friction first).
  24. loose belts? No. Loose pulleys? Maybe. First if this is an old printer try oiling the rods. Just one drop on each of the 6 (not 2, not 4, 6) rods in the gantry. Push the head around to see if resistance is much higher in one axis versus the other. They should be the same. Your issue is missed steps or slipping pulley. The first is usually caused by unusually high friction. The second is when a pulley came loose. So check that next. It's almost always the pulley on the stepper. So if you have the long hex driver that came with the printer (black and green handle) then you can do it without removing the stepper covers. Push the head around until you see the setscrew deep in there between the 2 sides of the short belt. Tighten the hell out of that set screw? It only takes two screws to remove that metal cover that hides the stepper motors. That's your backup plan. Tighten the other 5 set screws for the slipping (Y) axis as well but it's usually the one on the stepper and you have to tighten it so much the hex tool twists a bit. Scary tight.
×
×
  • Create New...