GregValiant

I think this is a question for a Prusa support site but I'll give it a try. When I plug your code into Cura in a StartUp G-Code it works and the keywords are properly replaced by Cura during the Gcode generation. You cannot add keywords with a post-processor because the Gcode generation is completed prior to any post-processing and I believe it's during the Gcode generation that the keyword replacements are made. ;StartUp Gcode M140 S{material_bed_temperature_layer_0} ;Start heating bed M104 S{material_print_temperature_layer_0} T0 ;Start heating extruder #1 M104 S{material_print_temperature_layer_0} T1 ;Start heating extruder #2 ;Tx ; take this out and see how it works. (This may be Prusa specific and might require your input on the LCD?) M190 S{material_bed_temperature_layer_0} ;Wait for bed to reach temp before proceeding M109 S{material_print_temperature_layer_0} T0 ;Wait for extruder #1 M109 S{material_print_temperature_layer_0} T1 ;Wait for extruder #2 ;Blah, Blah, Blah, more code ;End of Start Gcode By explicitly stating the Extruder numbers both would start heating. I don't know what effect that will have when you have an extruder disabled - it may heat anyway because of the above commands. You would need to check that. Also - the above code heats the bed and both extruders at the same time. That requires a fairly robust power supply and you might find it necessary to reorder the commands. I'm not sure why Cura always adds the M82. If you are using relative extrusion then an M83 will be added after the start gcode. A lot of printers have purge lines in their start gcode and they are pretty much exclusively Absolute extrusions. Maybe that has something to do with it.

The definition file has to include the extruder train. This is from the "Tronxy_x" definition and it looks right to me. "machine_extruder_trains": { "0": "tronxy_base_extruder_0", "1": "tronxy_base_extruder_1" }, Your "Extruder 1" file has two errors. "metadata": { "machine": "tronxy_x", "position": "0" ---------------this should be "1" }, "overrides": { "extruder_nr": { "default_value": 0 },-----------------and this should be "1" When Cura loads the two extruders it looks like they are fighting for Position 0. I'm guessing that's why they work separately but not together.

I was thinking (scary, I know)...if you install a Geeetech A20T in Cura it's a three extruder printer. You can set it up with the shared nozzle and shared heater (in Printer Settings) and set all the material standby temperatures to your printing temperature, and then when you slice, the Gcode will reflect where you need to hand code. T0, T1 and T2 will be located in the code where you need to make color changes. Many changes will be mid-layer. You would need to sync the XYZ and E locations every time so it will still be tedious but searching through the gcode for every change location would be a lot mor tedious and it would be easier to make mistakes. I checked a simple 3 color print and there were 260 tool changes (in 253) layers and each one would need hand coding for the color change. They would all follow the same form but none would be identical and a typo could ruin your whole day. I know yours won't be that complicated but still... Cyano-Acrylic Super-Glue. It works well. It's simple and easy to understand. It would only require 1 color change during 1 print and none in the other.

I don't see a definition file for your "-2E" printer. Did you get a definition file from TronXY? I ask because I could load the model from the 3mf file but Cura didn't like your printer. Using my own dual extruder definition the preview looks right and the gcode looks good also.

Cura works layer-by-layer so... With a single extruder printer, this would require much hand coding of the Gcode. With a Dual extruder printer there would still be a manual color change from light blue to dark blue. With a 3-in-1-out hot end this could be printed as you describe. If you were to print the gray parts ahead of time, and then print the light blue and dark blue pieces as 1 part with a color change, then the pocket would be there to glue in the gray pieces. When I print things like this I just allow the gray part to extend all the way out to the outside. That way it's just another color change (light blue to gray to dark blue). Yes, the outside would show a gray stripe - but for most pieces it doesn't matter.

Embed a small magnet? The arms would just be asked to hold it level.

In Cura - click on your printer, go to Manage Printers and then Machine Settings. In the left lower part of the dialog is "StartUp G-Code". The Ender 3 usually starts with the following lines: 92 E0 ; Reset Extruder G28 ; Home all axes G1 Z2.0 F3000 ; Move Z Axis up little to prevent scratching of Heat Bed Add a line M106 S0 after the G28 line so your start Gcode looks like this: 92 E0 ; Reset Extruder G28 ; Home all axes M106 S0 ;shut off the cooling fan G1 Z2.0 F3000 ; Move Z Axis up little to prevent scratching of Heat Bed There should be an M106 S0 (or M107 - same thing) in your Ending Gcode as well. The M107 is being deprecated in some firmware so M106 S0 is a better choice. Doing it this way will mean that all your prints start with the cooling fan off. Your Cura cooling settings will take effect at the start of the actual print.

This is a bug in the 4.11 software and has been noted on Github. If you change either the material thickness so it's thick enough to slice but not exactly equal to your Initial Layer Height it will slice. So if you have a model that is .200 thick then you can scale the model in just the Z to .201, or change the Initial Layer Height to .199. Either one should work until this is fixed in the next Cura version.

Hi @geert_2, I used 1mm because it was PETG and it would pull harder than if it was in PLA. Even at 1mm the ears on that test piece were pulling up in the center from the part yanking on them. I usually go with a minimum of 25mm diameter for just that reason. If I design in the elephant ears into the part (they aren't always round of course) and I think they need to be thicker, then I usually add a chamfer at the part line to make trimming easier. I add the .4 thick ear and then copy the edges and extrude them .6 at a -45 draft angle. So the added thickness helps to keep them from deforming but the trim line is only .4 thick so the exacto blade goes through it easier.

Again, What printer? Some of the Creality machines (and maybe others) don't like file names over 16 characters. Change one of the file names to something like "A.gcode" and see if it shows up on the LCD.

"I'd had enough of the printer for one day." I know that one. "2) Re "the hot end goes to 170 for bed leveling": Why? What does that do that my "M104 S{material_print_temperature_layer_0}" doesn't?" I don't have ABL but even manually leveling, if the hot end is at printing temperature it oozes. Leaving it at 170 for leveling allows the printer to move the head around without oozing onto the build area (but it's heated enough for most of the thermal expansion to have occurred). Then it heats up to printing temperature when it's sitting over in the corner. Each line of Gcode stands alone so an extrusion does not carry on to the next line. The printer-processor knows where the head is, a line of Gcode tells it where it's going, and it calculates getting all 4 motors there at the same time. When an axis isn't mentioned in a line then that motor just doesn't move. You've got that fancy DD on there so TPU should be no problem. Could it be a spring-tension thing on your extruder arm? Maybe it just isn't putting enough pressure on the filament (although that's hard to believe). Have you had any problems with the extruder skipping steps with harder material like PLA? If you have a 0.6 nozzle handy maybe you should give that a try.

When the nozzle moves around a hole it always pulls the molten plastic towards the center and makes the diameter smaller. It's typically what Hole Horizontal Expansion is used to fix. Over about 30mm it isn't an issue, but on small holes the diameter can change a lot from the design diameter even though Cura's path is theoretically correct. Does it matter if the sides of the batteries are in contact with each other? You could go back to the design software and cut all those thin walls out. Another option is to download and install the Arachne version of Cura (I'm sure there is a thread with a link here somewhere). It's still being worked on by the Cura team but the Beta version is stable and it slices thin walled objects better than previous versions of Cura as it uses variable line width.

It's one of those models that is just tough. You have those really thin walls between the holes. Consider - if you were machining that part out of metal there would be no way. The walls would be like razor blades IF the cutter didn't tear them up when the chips hit those thin spots. If you aren't totally locked in on the final inside diameters then you could try setting "Hole Horizontal Expansion" to a small negative number like -0.25 and see how it slices. That will decrease the diameter of the hole by .5mm but it would slice and print much better. Remember that the Line Width is also the index distance between two adjacent extrusions. I don't know if this was your design intent, the distance between the holes is not consistent and so the lands between the holes is not a consistent thickness. Looking at the area within the circle you can see it is much thinner (it doesn't look like it sliced) than the area bracketed by the arrows.

Because TPU is quite a bit different than printing PLA I think you should consider adding another Ender 3 in Cura with the StartUp Gcode I'm including below (name it "Ender 3 TPU" or something). When you want to print TPU just pick the other printer in Cura as well as your TPU material profile. The main changes that I've made in this StartUp Gcode is that the hot end goes to 170 for bed leveling, then after leveling, it moves to the left front corner of the bed, and then heats to printing temperature before starting the purge lines (your second purge line didn't have any extrusion in it). That keeps it from drooling all over the place during leveling. The second main change is to slow down the purge lines from 25mm/sec to 10mm/sec and also to lower the flow. The purge lines were designed for PLA and are way too fast and heavy for TPU. ;Start GCode M140 S{material_bed_temperature_layer_0} ; set bed temperature to initial layer bed temperature M104 S170 ; pre-heat the hot end to leveling temperature M190 S{material_bed_temperature_layer_0} ; wait for bed to reach temperature M109 S170 ; wait for hot end to reach temperature M220 S100 ; Reset Feedrate M221 S100 ; Reset Flowrate G92 E0 ; Reset Extruder G28 ; Home all axes M420 S1 Z5 ; Enable bed levelling and set Z fade height to 5mm G1 X0 Y0 Z10.0 F3000 ;Move to 0,0,10 M109 S{material_print_temperature_layer_0} ;set hot end to initial layer printing temperature and wait G1 Z2.0 F3000 ; Move Z Axis down G1 X5 Y20 Z0.3 F5000.0 ; Move to start position G1 X5 Y200.0 Z0.3 F600 E12 ; Draw the first line G1 X5.3 Y200.0 Z0.3 F5000.0 ; Move to side a little G1 X5.3 Y20 Z0.3 F600 E24 ; Draw the second line G92 E0 ; Reset Extruder G1 Z12.0 F3000 ; Move Z Axis up to prevent scratching of Heat Bed G1 X5 Y20 Z0.3 F5000.0 ; Move over to prevent blob squish ;End Start Gcode TPU is fussy. Speed is a big thing and 10mm/sec is about right but with a layer height of .12 and a Flow setting at 50% you are hardly moving any material. 0.25mm³/sec is only 0.1mm/sec of filament. When I print TPU on my Ender 3 Pro (stock bowden setup) I turn retraction off but your direct drive should be able to handle some retraction. I don't have real experience with DD but 1.5mm retraction seems like a lot. I do change my Flow for TPU. For PLA and PETG my E-steps are at 97 but calibrating the extruder for TPU they went to 110. Rather than fooling with the E-steps every time - my TPU profile in Cura has all the flows at 113%. I end up pushing more material and so I question your 50% flow number. I'm guessing a single wall calibration cube? I suggest that you go back to 100% in Cura and watch the print and adjust the flow with the Tune menu on the LCD until it looks good on the print. Then you will have a good solid Flow number to enter into Cura. Then we come to layer height. Mostly, I print custom gaskets with TPU. It's pretty resistant to gasoline and if the area where it's mounted sees less than about 140°C it holds up quite well. Consequently I don't care how they look as long as they seal so 0.2 layer height works well for me. My numbers give me a flow of 0.8mm³/sec which is 3.2 times the flow that you are using. I haven't had any trouble pushing the filament at that rate, but with my bowden setup retraction was a definite problem as the TPU just doesn't want to react to the extruder movement.

"I hope you get what I mean" Sometimes it's difficult to describe a problem with just text. A screenshot of your Cura Preview would be good. If you use "File | Save Project" in Cura and then post the 3mf file here, it will contain the model, your settings, and your printer. That makes troubleshooting much easier. If the model is proprietary and you don't wish to post it then if you could use one that is similar and has the same problem then that will work too. The "won't print thin walls" problem can be caused by different things. The more information you can provide the better the chance that someone will determine what the issue is.

@geert_2 I do enjoy your experiments. Here is a Geert Test Piece altered with the addition of two 1mm x 15mm diameter elephant ears on two adjacent legs. This is printed in PETG with the cooling fan at 100%. Basic settings were: Print Temp 235°, Bed 80° with AquaNet Super Hold hair spray, 0.2 Layer Height, .4 Line Width, 8 walls. The two "earless" legs lifted just over 1mm. The elephant ears were being pulled up at the center but the outer edges held.

At 10mm you will be pulling molten plastic into the bowden tube. That is a "bad thing" as the plastic will harden and cause a restriction in the filament path. After a while it will get critical and there will be extrusion problems as the filament finds it harder and harder to push through th restriction. If by "wispy" you mean the cotton candy looking floss - I've never been able to cure it. I have found that some nozzles are worse than others so it might be mechanical. If the nozzle isn't exactly straight up and down then the bottom isn't exactly flat and that might have something to do with it. When I print with white PLA I noticed that my layer cooling blower (a robust 5051 blower) really has an effect and blows the wispy strands all over the place. With black or red PLA the problem doesn't happen and that indicates it might be dependant on whatever the filament manufacturer decided to use for color.

If you scale a calibration cube to 150mm long x 20mm wide x 15mm tall and print it with PETG with a skirt, it's very likely to pick up on the ends. It's a near certainty with cooling running at 100%.

What are you really trying to do? If you just want to print it fast then you need to make the model larger. The print head has to speed up and slow down for each corner and the corners are sharp so essentially the print head has to come to a full stop at each corner. If you scale the model to 750% in just the X and Y it will be 150 x 150 with 3mm walls. To narrow the walls back to .4mm - set the Hole Horizontal Expansion to 2.6. Now you have an object 150mm per side = 600mm periphery (actually 600 - 2 Wall Widths). 600mm / 100mm/sec = 6 seconds to make a lap around. As @Smithy notes you will still need to adjust your minimum layer time to 5 seconds, all your printing speed settings to 100, and the printer should maintain your print speed mid-wall and your Accel and Jerk settings will come into play at the corners. The higher the Acceleration settings then the quicker the starts and stops. Your Ultimaker is much more capable than my Ender 3 Pro and can withstand higher Accel settings. I don't know the Max XY accel on your printer. Mine is 3000 and at that number the stops and starts are fairly violent so I don't set it that high. I suppose you could scale the model, set Remove all Holes, set top bottom layers to 0, and print the model in Spiralize. Again, it depends on what your are really trying to accomplish.

What we have here is a special Ender 3 Pro that has a build volume (in Cura anyway) of 230 x 230 x 1000. I opened my sword blade in MS 3D Builder (any STL editing software will do). I added 2 cylinders 2.2mm diameter x 20mm long and carefully positioned them so they would be centered within the sword blade at 200mm intervals and then subtracted the cylinders from the blade. That leaves holes in the blade. After saving I bring the blade into Cura. Add a support blocker and configure it as a cutting mesh with 0 top and bottom layers, 0 walls, and 0 infill density. Scale it up so it covers the part of the blade I don't want to print (in this example it is 610mm tall) and it is placed so the bottom of the cutting mesh is at Z=200. When I slice the model I am left with the first section of the blade with locating holes on top. Next I sink the blade 200mm into the build surface in Cura and move the Cutting Mesh up 200mm When sliced - I have the center section of the blade. With Horizontal Hole Expansion set to 0.1 my printer will make a hole just about 1.9mm in diameter. I super-glue pieces of 1.75 filament into the holes in one part and then super-glue that part to the next part of the assembly. The pieces of filament act as locating pins and even though the mating surfaces are slip planes, we have positive location. I'm sure there are other ways to do it. The rear luggage compartment of this bike was created using the above method. As long as parts can be positively located in relation to one another - big ones can be made from several little ones. Whether you use locating pins or create lap joints or mortices it will work. You can actually split the model into several STL's by doing "Subtract" operations in your editing software.

Actually it's simpler than that. What I described is how I usually adjust the infill in different areas of a single model. For your multiple models just select one and then select the Per Model Settings and Infill Density.

Within the Travel settings is "Retraction Minimum Travel" and "Max Comb Distance with No Retract". If either is set too high then there isn't any retraction. Set them to something low (like 2mm). For some parts that will be too little but for the dog and the stringing between the model and the muzzle-support it should limit the stringing. When you look at your model in the Cura preview you need to turn on the Travel lines to eyeball some of the distances and then make settings adjustments to insure you get retractions when they are needed (like moving from the neck to the muzzle-support). With my printer I don't use "Retract at Layer Change" and "Retract before Outer Wall" as they tend to leave marks. Z-Hops can also either cause or cure some stringing. Since the muzzle-support is the first thing to print on a layer then maybe you want "Retract at Layer Change" enabled. My experience (so far) is that large models with infrequent retractions are easier to tune than small models where retractions occur pretty much constantly. I found that "Connect Infill Lines" and "Connect Support Lines" can also help to limit the number of retractions. If the machine is constantly pulling molten plastic back into the hot end it's only a matter of time before a partial blockage occurs.

The bug has been reported on GitHub and the Cura team is aware of it. A workaround is to change your "Layer Height" settings in Cura to .199. I think that works.

I'm very surprised by that. Perplexed as well. Time to use the strike through font and my apologies to all.

A picture is worth a thousand words. A screenshot of Cura is worth 1,348 words. You could scale them. You can move some off to the side of the build plate while you slice others. You could export the individual parts as their own separate STL files and load them later. Hard to tell without seeing what you are trying to accomplish.