Jump to content

GregValiant

Expert
  • Posts

    5,183
  • Joined

  • Last visited

  • Days Won

    211

Everything posted by GregValiant

  1. Cura really needs a solid body to slice. It can be thin, but it needs an internal volume. I think that's what @MariMakes was suggesting when she mentioned that the walls should be 0.8mm thick. The walls of your box are surfaces and infinitely thin and so it can only be sliced in Surface mode as the model doesn't qualify as a 3D Solid and in fact could not exist as a physical thing in the real world.
  2. Within the Special Modes settings is "Surface Mode". Have you tried setting it to "Both"? The way the model is designed (infinitely thin surfaces) it really can't be done any other way. The thicknesses everywhere (except the bottom) will be dependent on your Line Width. The bottom thickness will be dependent on "Bottom Layers". I can see where travel movement could be a problem as all the extrusions are essentially one-way streets.
  3. I think what's happening is that Cura is adding a regular retraction and prime around the combing moves (;MESH:NONMESH) according to the settings "Retraction Minimum Travel", "Combing Mode", and "Max Comb Distance with No Retract" and then it is explicitly adding a retraction when the layer changes. So it appears that there is no check to see if there was a retraction that was based on other settings, it's simply adding the "Retract at Layer Change" regardless of what else might be going on. Turning of "Combing" eliminates one of the retractions and leaves the "Retract at Layer Change". Now the question is "Is this a bug?". I don't think so as it appears to be working the way it was designed. But that raises the question "Is this annoying?". I could see where it could be annoying as blobs may be involved. You could write up a bug report over on GitHub. I did a quick search and didn't see any other mentions of this behavior. I'm often the greeter over there and it would get past me but at some point it might be changed to a feature request. (BTW I get the behavior you describe in 4.8 in 4.10 as well. 4.13 seems to be like the 5.x behavior. I will assume that a change was made for 4.11 or 4.12.)
  4. Alright. A contributor on Github (SmartAvionics) suggested changing the Mesh Tools "Maximum Deviation" from the default 0.025 to 0.100. That will filter out some of those very short line segments. I have a print running now and it looks very good. And it finished very nicely. I don't see any flaws.
  5. Cura will stick a model out at X200 or so if the model doesn't fit within the printable area of the build plate. If that dog's tail is sticking straight out the back then maybe the model is too large in the Y? If the dog is actually a pig then I don't know.
  6. I can't duplicate that so it may have something to do with the way you have Cura set up and your combination of settings. Use the File | Save Project command and post the 3mf file here. Someone will take a look.
  7. This is at 75% scale with scrap "natural PLA" at 0.8 line width. My previous print was with red PLA and I missed this. Light refraction makes this a tough shot. Loading the Gcode into AutoCad results in this view of the area with the bulges. Those jig-jogs are a maximum of 0.1478mm. I missed this on previous print as it isn't really visible when using an opaque filament. I think I'll post this as a bug report on Github.
  8. I printed this myself with some leftover PLA at 0.8 line width and didn't get any issues. Your model itself is fine. If you turn off the "Shell" line type in the Preview and turn on "Travels" then if it was spiralized correctly there should not be any travels above the solid bottom portion. This is my slice of that project file I created. You can see the travel lines within the base, and that there are none above the base. This indicates a single long extrusion up to the top. This leaves me wondering if: Your extruder is working correctly, the E-step calibration is correct, the condition of your filament, and whether you may have developed a partial clog just above the nozzle. So I went back to the images you posted of a print. In your first image you point out a hole in the side. No matter how it sliced that should not be there and I'm sure it isn't in the gcode. Looking at another image I see a spot with more holes (in red) and something wrong with the extrusion (possibly under-extrusion) that I've marked in green. There should be no holes and no ripples. When printing with PETG it can collect around the nozzle and form a blob that will occasionally come loose and get stuck to the print. This is a known issue and it's why people invest in nozzle cleaning stations. It's something else that could be going on, but a hole? Nope. Something else is happening there and it has nothing to do with the slice. It may have something to do with the printer calibration and that includes the mechanical aspect.
  9. Search and Replace can add multiple lines of code by using the "\n" newline character, but it cannot take wild card characters. There would be no simple way to maintain whatever temperature is in the line. That means a separate instance of Search and Replace for each temperature change and you would need to know the temperature in the line. If you knew that the line you want to replace is M104 S200 then this would work: Search = M104 S200 Replace = G91\nG1 Z3\nM109 S200\nG1 Z-3\nG90 Assuming the temperature is changing at 5° intervals the next Search and Replace would be: Search = M104 S205 Replace = G91\nG1 Z3\nM109 S205\nG1 Z-3\nG90 It's a bit clumsy and you could get a blob (as the hot end oozes) but you could add a retraction and a prime using M83 and M82. It starts to get clumsy pretty fast though. Replace = G91\nM83\nG1 F2100 E-6\nG1 Z3\nM109 S200\nG1 Z-3\nG1 F2100 E6\nM82\nG90 It would look like this in the gcode file: ;LAYER:5 ;TYPE:CUSTOM LAYER G91 M83 G1 F2100 E-6 G1 Z3 M109 S200 G1 Z-3 G1 F2100 E6 M82 G90 ;TYPE:WALL-INNER
  10. If they are both running Marlin firmware and both use the same filament size then that would be good. Other than that I think the two main differences would be in the build plate size and possibly the StartUp Gcode. Place the model in a position so it will fit the printer with the smaller build plate, slice it and save the gcode. You can copy the gcode file all you want. If all the printers run the same flavor of firmware it will work. If all the printers use the same size filament it will work. If there is a mix of 2.85 and 1.75 machines and they are running Marlin then you can print by switching the Gcode flavor in Cura to "Marlin (volumetric)" and add "M200 S1" as the last line in the StartUp gcode and "M200 S0" as the first line in the Ending Gcode. That will turn on Volumetric in the printer after any purge cycle and then turn Volumetric off at the end of the print. When printing in "Volumetric" the diameter of the filament is known by the printer and the printer calculates how much filament to push according to the diameter. You can bring a calibration cube into Cura and scale it to 1mm tall in the Z and slice it as a test print for your machines.
  11. Rip Van Winkle was on watch that day but he was taking a nap and missed it. @Cuq may be the author of that plugin and will certainly know more about it. If the post processor is adding M104 lines then you could add a "Search and Replace" post processor AFTER it and have Search = M104 and Replace = M109. You need to think about it though. The print head will stay in contact with the print and you will get a divot because it will sit there while it is heating up. For a calibration part it shouldn't matter but you need to be aware of it.
  12. Hi @Tom2. My printer is also an Ender 3 Pro so I thought I'd stick my nose in. That's an interesting model. It's almost like a tolerance calibration test. A couple of things stick out in your project file, You have most Flows set to 131%. Have you calibrated the E-steps on your printer? I kick the flow up for PETG but only to 105%. It looks like you have a DD extruder. If you changed the motor did you re-calibrate the E-steps? (You didn't happen to use a single wall calibration cube did you? (I'm not a fan of those.) Your printing temperature is 245°. I found that PETG gets pretty soupy and stringy when printing that hot. I print it at 230°. If you enable "Outer Wall Wipe Distance" at about 0.1mm it will help to hide the Z seam. Enabling "Outer Wall Inset" at about .08 will increase the gaps between the pieces. I would suggest that you set your "Retraction Minimum Travel" and "Max Comb Distance with no Retract" to very short distances like 0.8mm. That should force retractions between the segments and avoid gluing them together with strings. When you have "Retract before Outer Wall" enabled it also means "Prime before Outer Wall". That prime move is at the Zseam location and can make the seam more noticeable if the prime leaves a blob. (That is another reason to print this at a lower temperature.) When you figure out where you want the Zseam on a model there is an option to enable it as "Zseam Relative". With Z seam relative enabled and the location at 0,0 and the Layer Start and Layer End also set to 0,0 you will get the Z seam down the middle and limit the amount of travel. With "Avoid parts when Traveling" enabled the nozzle moves off the part and then comes back in. That can leave a mark at the exit point and at the re-entry point. Those marks can line up and appear to be another Zseam. You have combing set to "Not in Skin" but there is almost no skin on that model. I would just shut combing off for this. The faster the nozzle gets from point-to-point the less likely it will be that oozing glues the segments together. Models that inter-leave like that one are very intolerant of over-extrusion and stringing as either one can cause the segments to get glued together. The "appearance" of the Zseam is also tied to those two factors.
  13. Congrats on the upcoming special occasion. All I can recall is being a nervous wreck. In the "Per Model" settings are your custom supports set to "Print as Support"? An option would be to send me a PM and include a Project File ("File | Save Project"). I'm well past baby showers and to paraphrase - a project file is worth a thousand words.
  14. It's fine. In the Prepare view red means that an area needs support. In the Preview red indicates "Outer Wall" extrusions. I think it will come out much cleaner if you print it vertically. It will be stronger if you print it laying down. Use a pair of needlenose pliers to crush the shell of the support. Just keep working at it. An option I've been using lately is to lower the Support Flow and Support Interface Flow to 90% (the settings are in the Material section). Solidworks, Inventor, and Fusion360 are very good at creating good STL's from parametric models. Blender, TinkerCad, and some others need to be watched or the STL files will have errors. Here is a Cura Project File. My Ender 3 Pro is the printer. I don't know your printer but when you open a Project File you will be prompted to Install a new printer, or if you happen to have an Ender 3 Pro the prompt will be to "Update" the printer. If you "Update" then your printer would end up with my StartUp and Ending Gcode so "Install New". Then you can slice it and see how I've butchered it what I've done. GV_Toolcallibrator.3mf
  15. Every model gets it's own love. You need to stare at any model for a bit to figure out what the best setting configuration will be. Should the model use normal supports or tree supports? Touching build plate or Everywhere? Can I print it at 125mm/sec because there are big flat areas, or do I want to do it at 35mm/sec because there will be lots of islands? .1 layer height or .3 layer height or something in between? If there is a big horizontal hole that's going to curl up as it closes then maybe I want z-hops enabled. Every model is at least a little different. Rather than have a lot of profiles (that I can't remember why I made them), I tend to save project files. If I alter a design I can open the project file of the previous design, throw out the model and bring in the new one. Then start tweaking.
  16. I don't know how they wrote the code but I think it should show the defaults and NOT the changes you made in the main settings boxes. The profile should remain unchanged until you ask for it to be updated. The changes are shown in the "Current" column and when you "Update Profile" those overwrite the previous settings and become the new defaults. No matter what changes you make in the main settings - if you re-select your profile all the defaults will show up again. I use that to reset the Cura settings before I slice a new model. It sets everything back to square one and I can start making changes specific to the model I'm going to slice.
  17. The star takes you to the dialog. Instead of clicking on the star - click on the profile name. That will bring up the drop down box and at the bottom are three tools. One of them is "Update Profile with Current Settings/Overrides".
  18. If you have "Support Placement" set to "Everywhere" then check settings like "Minimum Support Area" and "Support XY Distance". There are a lot of support settings (73) and some can change whether or not supports are applied to certain areas. If you use the "File | Save Project" command and post the 3mf file here some one will take a look.
  19. In that top view they look like they might be set to "Print as Support". Select one and click on the "Per Model" settings button on the left toolbar. Set them all to "Print as Part". It's something like that (but my house is far away and I don't see so well anymore).
  20. Adaptive Layers ON. Adaptive Layers OFF. You an see that the "steps" on the roof are closer together with Adaptive Layers enabled. That's the improved resolution that Nallath mentioned. That also means more layers (355 vs 201 with the settings I used) and that directly affects the print time. Using Adaptive Layer Height will generally be slower unless you are slicing something with a regular shape (like a cube) where it will stay at the Max layer height and so there will be less layers to print.
  21. I moved to the 5.x versions. There have been glitches in dialog boxes and 5.1 in particular had slicing problems on Windows installations and with particular high resolution models. My prints have been fine. I thought 4.13.1 was excellent so you might consider installing it. You can keep your 4.5 installation as well. If you still have 5.x installed please load a Benchy or calibration cube and slice it. Use the "File | Save Project" command and post the 3mf file here. Sometimes things stick out and either way, the developers may want to take a look at what's going on. I have a couple of macros in AutoCad and Excel that I use for gcode analysis. The "variable" line width of the 5.x versions make the flow analysis difficult as there isn't any way to tell how far one extrusion is from the one next to it. Without knowing the line width the flow rate can look off but actually be correct. If a user sets up with different line widths for different things (infill at one line width, support at another line width, outer walls at a third line width and inner walls at a fourth line width) then my analysis quickly becomes useless. The same problem occurs with any on-line analysis software.
  22. I used the "repaired" model in this project file. The bottom and top portions were "unioned" so it is a single model. It is at my preferred speed of 35 but as I say, once it gets above the base it will move at 1 layer/10seconds unless the minimum layer time is changed. GV_BangiukasBottom.3mf Those artifacts look like what I call an XY seam. They can happen when "Spiralize" doesn't really work on a model for whatever reason. When a model is successfully spiralized the "spiral" portion of the gcode will look like this with the Z constantly moving up. (This is from the gcode file of the above project file). G1 F2100 X112.667 Y162.741 E2709.01761 ;Last extrusion line of of the non-spiral base G1 X90.701 Y152.887 Z2.216 E2710.67544 G1 X90.248 Y152.693 Z2.217 E2710.70938 G1 X88.811 Y152.04 Z2.218 E2710.81807 G1 X88.359 Y151.845 E2710.85197 ;the Z doesn't move up on every line. G1 X86.922 Y151.192 Z2.219 E2710.96066
  23. I exported the model as an STL and uploaded to https://formware.co/OnlineStlRepair. Here is the report on your model. --> 0 Naked edges (?) --> 0 Planar holes (?) --> 0 Non-planar holes (?) --> 269 Non-manifold edges (?) --> 383 Inverted faces (?) --> 0 Degenerate faces (?) --> 9 Duplicate faces (?) --> 0 Disjoint shells (?) -> Repairing: 100.00% ----- Repair completed in 48804ms ------ -> Vertex count changed from 185215 to 185571 (+356) -> Triangle count changed from 370980 to 371138 (+158) When a model is non-manifold you should get the warning that "the model may not print correctly." Cura may have made mistakes in the slice because of the errors in the model. You have "Remove all Holes" enabled. Is that because you don't need the holes in the bottom plate? Your settings look reasonable. I'd print it at 35 but that is my personal preference for PETG. I just think it looks better when printed slower. The upper part probably doesn't matter as it will bump up against "Minimum Layer Time" either way.
  24. Load the model and set Cura up to slice and then use the "File | Save Project" command. Post the 3mf file here. That model is solid through-and-through?
  25. Or get a new mainboard that has an SD card slot that works. I control my printer over the USB and I can ask the printer to list the files currently on the SD and then pick the one I want to print on my laptop and send M23 and M24 and the file will print from the SD. My point is that there are options besides putting it out on the front porch and hoping that some dummy will steal it.
×
×
  • Create New...