Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


Everything posted by alexjx

  1. I think this is a good idea. Though I agree that the root cause is the leveling. But for beginners to do a good leveling is difficult. Because it is not really a measurable thing. It requires quite a lot experience. So why not make the tool do more that we could get started. Instead of being frustrated and leave.
  2. Do you do regularly updated build? Or github I could get the latest. I also like the extra control. 🙂
  3. For one, I think you have to choose "marlin" when you use host software like RepetierServer. marlin, reprap and ultimaker 2 here, are something like dialects of a language. They are all gcodes but, there are a few differences. As you have figured out, when you choose ultimaker 2, the start procedure is handled by the firmware. The same behavior could not be achieved when using host software since it's now the host sending gcode one by one to the printer. While there is no "start", so the firmware doesn't know when to do the start procedure. Therefore the start has to be done by the slicing software or host software. When changed to Marlin, Cura has a default start script. You could use that as a base and tweak to the behavior as your liking.
  4. Aha, I see it now. so "within infill" will only consider the start and end point if it's "infill" but not in between. That's why once I was printing a spring, it travels across the whole spring only if I set to "within infill". Thank you again for explaining this. 🙂
  5. @smartavionics Thank you very much for the infill tip. 🙂 I tried "not in skin" too. but there is also a similar issue. I'm not sure if they are related.
  6. @smartavionics Thanks for helping. The gcode is attached. And the issue happens in layer 8 of a corner. mcstnd_base.zip
  7. @smartavionics Here it is. Let me know if I'm doing it right... Should I use "save" or "export"? mcstnd_base.3mf
  8. Right... Then, CuraEngine itself doesn't have a hard limit in the code, so it will take whatever the Cura GUI provided. Setting it to 9999 then the upper limit is 9999. This is from reading the code. And for a practical example, The "bird in the cage" as a lot of small features need retractions. In the first picture, max is set to 99 and the second with 9999, and no other change. Notice the dark blue lines indicating none-retraction travel since retraction count exceeded the limit, while the second doesn't.
  9. I think it only warns you. I've set this to 9999, and didn't observe any slicing problem. But of cause, setting retraction count so high will risking chewing your filament if the window is very small.
  10. Hi all, I'm trying to understand a travel path that causing blob on the surface. I've enabled both "retract at layer change" and "retract before outer wall". The travel is to the skin wall with retraction. Combing is set to "within infill". What makes the print bad is that the path generated seems a bit off the skin before it could start print. (see the picture). Is there any option I could tweak to get rid of this? Thanks. Jia
  11. I think this behavior started at 3.6. I had a hard time understanding it too. But it makes sense after you realize if two material interleaving with each other, the adhesion tendency will not as strong as one. So Cura chooses to use one material always printing a shell for the tower, so it will not break during print. For support materials like PVA, this makes sense, and for dual color prints, for most of the time, each layer will have two colors so it also makes sense. The only problem is where things like a traffic cone, which different colors printed in different layers. I think this is a small price to pay...since the original behavior is to print different colors alternatively in each layer. And in this case, it will make the tower even weaker.
  12. Interesting, I've been using DXU for a month, but didn't notice the issue mentioned here. I recognize mark2 is a great design, and what makes other solutions like DXU possible. However, it doesn't make sense to rule out other also good solutions. The heat breaker that DXU uses also works well enough to handle two nozzles. I could use a stronger fan to cool them. I'm using the 2510 variant and observed the cooldown time is much shorter than the original head. I printed regular PC in one head @ 290C, and loaded PLA in another, with an enclosure. Under this situation, the PLA in another head didn't deform enough to jam it. As the leaking, I think Cura's standby temperature did a great job. We are talking about modifications, there is a trade-off. The thing I believe mark2 made is space. The extra head will take up a much larger space than what DXU takes. I think the downside of DXU is really the source of material... It is not so easy to get as mark2.
  13. Doesn't ChangeZ alter the feedrate factor instead of the "speed" itself? So it affects all the "speed" with a ratio. Could you look at if the output gcode has M220 in it? If there is, please check if your firmware has M220 support.
  14. I'm afraid not... since one of the walls is already much less than 1mm thick.
  15. For what it worth, here is what I measured. The diameter of the hole for the heater is 4.05mm and 16mm deep. So maybe for 4mm heater. The diameter of the PT100 hole is 3.3mm also 16mm deep.
  16. Interesting, what firmware r u using? As far as I know G92 will be "reset" once homed.... so unless you print second one without homing.... the z offest will be gone...
  17. I asked a similar question recently, I guess that's because the red one is your first nozzle. If you also using 3.6 then this is the trade-off made to make sure the prime tower stable. I guess Cura developer consider most people are using the second extruder for support, therefore different material. So Cura will generate a tower with a shell that always printed by the first nozzle even if there is no this color in this layer. In this case, I think it makes sense, but it sacrificed cases that we print different color of the same material. I have started using Cura since 3.6, but I heard the old one that has a different implementation of this. But it has another problem. Unfortunate the developers consider dual color minor and providing an option to toggle the old and new will make them maintain two variants of the implementations. And that is something they choose not to.
  18. I think I used the wrong words... it works as intended. but just not my expectation. It does travel within the gaps. regarding this, the avoid distance doesn't seem to work. Please see the attached image, I tried to set an ... really large distance, but it still travels within the gap. What I expected was following, when I set to "Within Infill" or "Off".
  19. Thank you all. Combing does have an impact on this. It's interesting to find out. "Not in skin" does not work in this case, since the spiral path is "skin" after all. @smartavionics I agree that there should be an upper limit. I thought the "Max Comb Distance With No Retract" was the limit... but it seems it only impact the "retraction"... not the travel...😂
  20. I would suggest trying to change the line width, the yellow is the infill but it has to follow the infill pattern instead of the shell's. But I think what you are trying is to make it the same as shell?
  21. Hi all, Thank you for reading this. I'm trying to understand the travel pattern of Cura. I'm using 3..6, trying to print a spring of a toy bike. What I understood (I hope) is that cura tries to print the inner wall then infill, so it has to move around. But, why it has to move in the pattern as if it's trying to avoid the printed parts??? And even more... it dosesn't matter if I disable the "avoid printed parts when travel" option. Thanks in advance. Jia Additional info: This is a snapshot of the first layer.
  22. Thank you for your fantastic work. I really liked this compact and efficient design. The documentation is thorough and the design is well thought. It's not difficult to print and assemble. I printed it with transparent PC, so it has its style. I got pretty good results with it. And most of all, the DXU itself look really good. 😎
  23. I once hit this issue before, while my UM2+E was set to run in MARLIN instead of ulticode. The plugin inserted the recovery instruction as "M220 S-1". However, the UM2.1 firmware doesn't work correctly to handle the negative value so it turned out to be extremely slow.
  • Create New...

Important Information

Welcome to the Ultimaker Community of 3D printing experts. Visit the following links to read more about our Terms of Use or our Privacy Policy. Thank you!