GregValiant

Welcome to the forum. @Dustin of UM may know of classes for owners of UM machines. For the rest of us it's mostly "jump into the deep end" and sink or swim. A couple of questions... What printer(s) do you have access to? What materials will you use for printing? You will need to adjust the models for 3D printing. Simply scaling them down will cause features to "disappear" because they get too fine to print. You can see them in the model, but the line width of the printer simply can't be adjusted small enough to print them. Things like window mullions, small overhangs, landscape trees and bushes, supports for walkways, etc. don't lend themselves to 3D printing. My guess would be that making those sorts of adjustments was a big part of what your cohort was doing. You must think "fun scale" and not "exactly to scale".

The teeth on the extruder drive gear are sharp. They bite into both materials pretty much the same so I don't think that's it. It might be shrinkage as it cools but I don't have a way to check that. Dimensionally there is no difference in the prints. Visually I can tell (I picked up a microscope somewhere along the way) and that is where the 105% came from.

PETG is so soupy that the first layer over infill always wants to sag into the infill gaps and the sparser the infill, the worse it gets. I use the fan for that layer and the results are much better as the PETG hardens before it can sag much. I have written this post processor (it will be included with Cura 5.7) and you can turn the fan on for a single layer and then turn it off again. PETG generally doesn't respond well to cooling, but a little can help certain problem areas like the first layer over infill. Unzip it and put the AddCoolingProfile.py into the "scripts" folder in your Configuration folder. (You can find it by using the "Help | Show Configuration Folder" command.) I print a lot of PETG. My printer E-steps are calibrated for PLA and I have found that setting all the flow numbers in Cura to 105% helps with PETG. Since the filament diameters measure exactly the same I have no idea why that is, but it works. AddCoolingProfile.zip

I've seen this in my own prints. This is your project. (I've set the line width to .10 and forced compatibility mode for the view. I think it's easier to see.) It's obvious that the behavior extends into the skirt as well as that wall of the support structure. This is snippet of the gcode. It's as if Cura is stepping the line into a series of orthographic moves instead of a single long "point A to point B" move as would be expected. G1 X645.935 Y286.278 E153.93243 G1 X645.941 Y286.843 E153.99585 G1 X645.994 Y287.368 E154.05509 G1 X646.105 Y287.925 E154.11884 G1 X646.266 Y288.469 E154.18252 G1 X646.464 Y288.956 E154.24154 I didn't find a way to get rid of it either. I have not seen a bug report on Github that mentions this. You can go HERE and click on the "New Issue" button to fill out a bug report.

In the material settings all of your flows are at 90% except for the Initial Layer. You are asking Cura to under-extrude by 10% on all layers above the Initial Layer. Your line width is at .35. That's OK but your "Skirt Line Width" of .30 is too small for a 0.4 nozzle. Then you have the "Initial Layer Line Width" at 130%. Here is a generic PLA project for your printer. Only adjust the Print Temperature and Bed Temperature to your favorites for PLA or PETG. The print speed is 50 from the second layer up so it would be suitable for either material. The point here is to see how the printer performs with a bland base profile. (BTW - I didn't care for the stock "Kobra Max" profiles.) Here is a project file. It's a flattened 50 x 50 rectangular shape. The setting profile is a new one I named "GV Test". GV FlatCube.3mf

With Cura set up to slice, use the "File | Save Project" command. That will create a special 3mf file that will have the model, all your settings, and your printer in it. Post that 3mf here. It's really hard to tell what might be going on just from the images.

That's an Ender 3 of some sort, correct? Slashee has a couple and I have an older 3 Pro. If there was some sort of binding in the bowden I would expect it from right to left rather than fore and aft. The bowden flexes with X movement rather with Y movement. I think you are on the right track though. Just because I think X movement binding problem is more likely doesn't mean your printer agrees with me.

I have to revise my comment "I just don't see a point to it. The extra 3 or 4 minutes isn't going to hurt anything.". I only print PETG and some PLA so for me, that comment is OK. Neither material is prone to "cooking off" in the nozzle when there is no movement. On the other hand, there are some materials that don't like to sit there cooking and can start to break down inside the nozzle. In the case of a material like ABS (that needs the bed very hot) waiting until the bed gets up near the set temperature before starting the hot end could be not just useful, but necessary. I really like this new feature of some logic being done in the startup. I always had a second printer installed just for TPU because it needed really slow purge lines. Now a single printer can adjust it's Startup so it can be used regardless of the material. G1 X1 Y200 F{speed_print * 60 * .75} E15

If (for example) you use a custom profile that was based on "Standard" (or "Draft", or whatever) then that last portion of the Cura Gcode file would show all the differences between your custom profile and base profile that was used to create it. If you happened to use "Standard" with no changes, or if your custom profile used for the slice happened to be the same as "Standard" then the final paragraph would be empty. If you go to "Manage Profiles" and then "Import" - down in the File Type box you can select Gcode Files. When you import a Gcode as a profile, Cura will pull what it can from that final paragraph and build a custom profile. You can switch to that profile but you would still need the model so you would have something to slice. The final paragraph cannot be post-processed because it is inserted into the Gcode file after the post-processors run.

Cura adds the temperature prepend lines before the startup gcode if there are no "cura keywords" in the startup for the bed and/or hot end. It's a safety to guard against cold extrusions. Go to Settings, Printers, Manage Printers, and then Machine Settings. The startup gcode is in the left text box. Add the lines I've added in bold type: M221 S100 ;Reset Flowrate M140 S{material_bed_temperature_layer_0} M104 S{material_print_temperature_layer_0} M190 S{material_bed_temperature_layer_0} M109 S{material_print_temperature_layer_0} G28 ;Home That will heat both units simultaneously. Putting in a "dwell" to slow down the faster heating device can be done but it's not very useful. When I print PLA with the bed at 50, the hot end and bed finish heating at about the same time. When I print PETG with the bed at 83 the bed heats a lot slower when it gets over 55 so the bed takes about twice as long to get to the set temperature. I suppose I could stick a logic line in there. Something like: {"G4 S180" if "PETG" in material_type else ""} I just don't see a point to it. The extra 3 or 4 minutes isn't going to hurt anything.

I think the failure on part #2 happened when the support failed. You have the "hoop" of the model near the edge of the bed. Cura noted that and sized the trunk of the tree support to fit in the room it had. Here I've placed one model closer to the edge than the other. You can see that the trunk of the tree on the right is much more robust and less likely to fail. It looks like you had Z-hops enabled at 0.2. Try increasing that to 0.5. If the nozzle hits a skinny support (like the left model above) there is an excellent chance that it will knock it over and cause a failure. In regards to the under-extrusion, I don't know what to make of that. Here it looks like the supports failed because of under-extrusion. That also seems to have been a problem lower near the build plate. As you noted, the good part doesn't show that. If it was me, I think I'd print a couple of scaled up (in the Z) calibration cubes with one in the front of the bed and one in the rear. The point would be to try and duplicate the problem. It isn't likely that the problem is in the gcode but rather with the printer. On another topic - infill. Everybody likes gyroid but at low densities it fails. These extrusions are too far apart so they don't connect to the previous extrusion. More spaghetti. The first skin over this infill will be ugly as it will be mostly printed over air. I understand you are weight sensitive here, but the part has to have some strength.

That's really odd. Set the project up again in Cura and use the "File | Save Project" command and post the 3mf project file here. Somebody will take a look. What material is that?

Sunny Sarasota? It's been raining here in Venice for the past 3 weeks but the sun is trying to break through now. It looks like it will be cool and drizzly for at least the next few days. I guess it could be "El Nino".

That code is more suitable to a printer with ABL, but there are no ABL codes (G29 or M420). You are pre-heating the nozzle. When the print starts the "M104 S140" line is setting the hot end to 140. With an ABL system that would be nice so the nozzle doesn't drool on the bed as the ABL does it's thing. You don't want that if you are manually leveling the bed. After auto-homing (G28), the "M109 S{material_print_temperature_layer_0}" line should set the hot end back to the Initial Layer Print Temperature. "M109 S" means "set the hot end temp and wait for it to be reached". Even though the temperature is dropping to 140 during the startup, that Startup gcode should work. If you are manually leveling I think this would be more suitable. It will work whether you are pre-heating or not. I suggest that you install another instance of your printer in Cura and test this startup gcode. ;-------------StartUp M220 S100 ;Reset feed rate to 100% M221 S100 ;Reset flow rate to 100% G21 ;Metric Units G90 ;Absolute movement M82 ;Absolute extrusion M140 S{material_bed_temperature_layer_0} ;Start the bed heater M104 S{material_print_temperature_layer_0} ;Start the hot end heater G28 ;Auto-Home G92 E0 ;Reset extruder G1 F{speed_z_hop * 60} Z10 ;Move up M190 S{material_bed_temperature_layer_0} ;Wait for the bed temperature M109 S{material_print_temperature_layer_0} ;Wait for the hot end temperature M300 P250 ;Short beep G1 F{speed_travel * 60} X10 Y0 ;Move to purge start G1 F{speed_z_hop * 60} Z0.3 ;Move to purge height G1 F{speed_print * 40} X160 Y0 E30 ;150mm Purge line at 2/3 print speed G1 F{retraction_retract_speed * 60} E28 ;Short retraction G1 F{speed_z_hop * 60} Z0.4 ;Move up a little G1 F{speed_travel * 60} X150 Y0 ;Wipe the nozzle G92 E0 ;Reset the extruder ;-------------End of StartUp Since you are pre-heating (to level I presume) you could do without the M190 and M109 lines. Each of them can cause a 10 second delay as the printer decides if the temperature has been reached. They are there as a safety to insure that the bed and hot end are up to temperature before any extrusions. The Marlin commands are HERE. I suggest you look up the ones that appear in your Startup so you understand what they are telling the printer to do. The Cura keywords are HERE. Cura Keywords must be enclosed in curly brackets. The "speed" words are " * 60" because Cura uses mm/second but gcode uses mm/minute.

What printer definition file did you use to install the printer in Cura? There is a good chance that the problem is in the StartUp gcode. Have you looked into the gcode to see where the command is that is telling the printer to go to 140?

@Phrayzur are you in the correct thread here? This thread is a year old and your comment doesn't seem to jive with it.

I've tried to figure out the print order myself. It's a computer so random doesn't really happen, but I haven't been able to figure out the order. It seems to have something to do with the order the models are created or brought in, but I haven't found a hard and fast rule. Random I think is a good word for what we see with it. A common feature request on Github is to be able to move the models in the list at the bottom-left of the Cura workspace and have that be the print order. It hasn't happened yet though.

Even a blind squirrel finds a nut every now and then. I used to fly RC planes (and U-control) Of course they were balsa then with big ole 2 strokes Enyas, Foxes, etal. I loved the smell of nitro-methane in the morning. I didn't care so much for that fact that often, both planes would fit in one 5 gallon bucket on the way home.

Sorry Slashee. I don't think anything will help make this less confusing. Cura won't build support on a support interface so yes, the "Interface" Horizontal Expansion needs to be less than the regular Support Horizontal Expansion. How much less depends on the Support Wall Line Count (coupled with the XY distance). With the Support Wall Line Count at "1" the Interface Horizontal Expansion can be 1 line width less than the Support HE. If you happen to have Support Wall Line Count at "0" then the Interface HE needs to be 3 line widths less. In that project, the Support HE is 2.0 and the Support Wall Count is 0 so the Interface HE needs to be 0.8mm ("3 * Line Width" less than the Support HE). That will provide supports for the supports.

Use the "File | Save Project" command and post the 3mf project file and someone will take a look. Off the top of my head (which I don't like to do) it appears you might have "Wipe nozzle between layers" enabled (???).

Infill will always extend through any connected volume up into the next volume. You can change that by increasing the "Skin Expand Distance" which you will find near the bottom of the "Top Bottom Settings". These are the same model. I used the "Per Model" settings and the one on the right has the "Skin Expand Distance" set to 8 while the left model is at 1.2.

On the first layer you should see the support brims only on the inside of the tree trunks. Here I have set the supports to everywhere, and brought in a cylindrical support block and set it to "Print as Normal Model" and it is 1 layer height thick. Support Floor is turned off and the Support Bottom distance is set to "0". The tree support grows from it as if it was an external brim. You can see that the other trunks have brims on their insides and that limits the brim size to the trunk diameter. I think that's about all you can do though. Supports always have their brims on the inside.

Yes, you can adjust the flows within the Material Settings of Cura. You should have a "Tune" menu on the LCD. It's hard to believe they didn't put the E motor in there under the other axis movement commands. Skins are the best indicator of flow so in Cura: Scale that calibration cube to 50 x 50 x 1mm tall and set Cura to 3 top and bottom layers (at 0.2 layer height). Set all the flow rates to 100%. As it prints you should be able to adjust the flow from the LCD "Tune" menu. From the start set the LCD flow to 110%. As a skin finishes look closely to see if the individual extrusions are welded together nicely side-to-side. Keep moving the flow up in 5% increments until you see that the skins are going down flat and are welded together. Note where your LCD flow rate ended up. Back in Cura go to the Material Settings and set all the flow rates to that new number. Set the LCD flow rate back to 100%. Print the flattened cube again. A magnifying glass is a good tool for checking the top layer.

I'm old and can't remember stuff anymore. G1, G0, my name, things like that. I have noticed that stand alone "Z" moves are always "G1". Creality does fool around a lot and then doesn't tell anyone what they've done. A case in point was a couple years ago when they ran out of mainboards for the Ender 3's and just stuck in whatever they had laying around. Another instance was changing to TFT LCD screens and not re-working the firmware to talk to them. All of a sudden any command that sent a message to the screen was ignored. That meant M0 and M1 (and M117 and others) didn't work on those machines. I have a couple of ideas to maybe fix the problem If M25 is pausing the print - the printer firmware may still "read ahead" and finish any moves that are in the buffer and that can have some odd side effects. Something you can try would be to add a string of M105 lines into the "Gcode after Pause" box. "M105,M105,M105,M105,M105" will pad the buffer with do-nothing commands instead of movement commands. Another thing you could try would be: Using M25 - when you are ready to resume the print, first select "Pause" on the LCD and then select "Resume". That might fix the "return to Z" problem. The stock Pause at Height doesn't allow for an entry to define the Z lift so maybe that problem just can't be fixed. M600 would be the other filament change option. Once again, the firmware has to support it. If M0 works then it is the best option and it's as Slashee says and a button click on the LCD should cause the print to resume. In every example that was posted, there is a "return to Z" line. If the machine is going someplace different, then the firmware is causing it. It would be a rare thing if firmware allowed a G4 dwell to be dismissed while it was active. With the exception of Repetier firmware, you should keep the steppers enabled. If they lose their position you are basically screwed. I'm at a disadvantage here for a couple of reasons. One is that M0 works on my older E3Pro. Another is that when I changed Pause at Height, not all the changes were implemented and so it remains confusing. That leads to the other reason I'm at a disadvantage. It was confusing enough that I use my own script "Pause at Layer". Functionally, it works the similar to "Pause at Height. The Pause at Height sequence is: Retract (if selected) - Park - Pause - Move back - Resume. My Pause at Layer sequence is: Retract (if necessary) - Park - Unload - Pause - Reload - Purge - Retract - Move back - Unretract (if necessary). There is a bit more there. I doubt that Pause at Layer would fix the problem but if you would like to try it, let me know and I'll post it. (Don't let Slashee see it though. She'll pick on me about my horrible coding style.)

If your gcode is in "Absolute Extrusion" mode (and it looks like it is) then there won't be an E-7. There will be an extrusion line like "G1 X111 Y111 E50.5678" and then a "G1 F2100 E43.5678" which would be 7mm less than the previous E. Search the Gcode for "00 E" and you should come across the retraction and print lines. The retractions will be "retraction distance" less than a previous E value. They are always in the form "G1 Fxxxx Eyyy.yyyyy"