GregValiant

Please do a "Save Project" and post the 3mf file here. The Y dimension looks OK. How was the Z?

Yes. That F120 is mm/min. Very slow. Just out of curiosity, what is your z-hop speed setting?

I wrote a little app for calibrating Accel and Jerk. After some trial and error the max speed I allow is 400mm/sec. At 400mm/sec, 3000 Accel and 30 Jerk the Ender3Pro will average 340mm/sec in a round trip from 0,0 to 230,230 and back. At speeds above 400, the stops and starts are so abrupt that the belts skip and the machine can definitely hurt itself. It's ugly. A larger build surface would allow higher top speeds but in my particular case, the hardware can't take it. The heavier Y axis is especially problematic. Prints might self-release. I will not be taking the printer out to Bonneville any time soon. Here. It is an unsigned Windows (only) app in VB.net and uses Marlin commands. There are a few different race courses. The circles use G2 and G3 commands. Sending commands for the Nürburgring course was a problem. That's out of the way. So @gr5, did you know (or did you have to look it up) that the indicated speed was the speed of light? (actually 1/100 the speed of light (and I did have to look it up) but that's close enough for me).

I use other peoples questions and problems as tools. I do some research, play with the software, and see if there is a way around, through, over or under a problem. It helps me to develop my own skills as a designer and printer of plastic craptastic. It also gives me something to do while waiting for a color change or other pause event. So I had played with your situation for a while and really couldn't come up with anything except "slice twice & splice". Several hours later I decided to try again, and there it was. A big blocker on top resting on the little blocker in the hole. I think it was the height of the hole blocker being set to 3.84 that made the difference. Then the big blocker bottom is at 3.84 so it takes over from there and doesn't block the walls. Or something like that. Anyway, it worked. The 3mf file has the settings for the blockers as well as the rest of the Cura settings. Have a good holiday.

And then I looked at it again... REV_CE3_LampaCreponata.3mf

The good news is that the cutting mesh works fine. The bad news is that all it is doing is creating a pocket in a solid model. Those walls and roof that appear around the cutting mesh are not part of the cutting mesh, they are walls and floor of the main part around the pocket. Extending the cutting mesh through the roof of the part eliminates it's top because it's a floor. The walls remain. You have 3 options. Shell the model and then thicken the floor and put a hole in it. Your settings for walls and floor may remain the same. Move to a .8 nozzle. It would allow you to push the line width to 1.0 or even 1.2. Then you could use spiralize and get at least somewhat thicker walls in 1 pass. In that case the cutting mesh would work since Cura is throwing out the inside volume of the model. It would require the cutting mesh to be 3.84 tall (12 layers at .32) and resting on the build plate or it will still grow walls. Another alternative is to slice the model twice. One would be the skirt and bottom 12 layers with a cutting mesh to make the hole, and a huge cutting mesh covering the rest of the part. The second slice would be the rest of the model except for the first 12 layers which you would expand your 1st cutting mesh to cover, and delete the huge mesh, and set bottom layers to 0. Then you could copy and paste file #2 starting at layer:0 into file #1 at the end of extrusions for the top layer (delete the end of file stuff like shutting off the heaters). There would likely be a G92 to sync the extruder at the splice. The Z would work out. You aren't moving the base model so the X's and Y's would work out. There is no additional homing required. Nice.

There are several bug reports and a lot of discussion of this on Github. The Cura team knows about the bug. It appears it is not an easy fix. I think "at height" works better if adaptive layers is off and there are no z-hops changing the height all the time.

There are two supremely useful tools on the left side of the screen. Support Blockers and Per Model settings. A support blocker is a special mesh (part) you can bring in. It can be sized, moved, rotated into position to block support. It can also have it's own set of special settings that are different from the rest of the things on the build plate. It can be used to add, or to subtract things (a cutting mesh). There is a plugin available in the Marketplace that provides round support blockers if the rectangular ones aren't the right shape. Any regular model can have special settings as well. Not all settings are available in Per Model settings, but a lot are. Click on the model you want to be different and select "per model" settings. Then add the settings you want to be different and change the values for that setting.

That setting can have un-intended side effects in the X and Y. Setting it to Middle is the correct thing to do. Regarding aHoeben's suggestion, you can print two calibration cubes side-by-side and set the Per Model Settings for one of them to the Horizontal Expansion of your choice. That way you can tell if Middle or Horizontal Expansion made a difference.

It would appear that Tiny Machines is a re-seller of various manufacturers printers and that they upload their own firmware. To make sure, you need to go back to Tiny Machines and find out what the gcode pause command should be, and how to restart. Layer numbers in a Gcode file are "Base 0" and in Cura they are "Base 1". When you look at a preview in Cura and decide that you want to pause at layer 9 then in the Pause at Height dialog you would enter an 8 (the Cura layer -1).

What is "Experimental | Slicing Tolerance" set to?

Aaah Haaa. And that is why discarding the settings worked. The PETG went back to default and "Make Over Hang Printable" was turned off. I had played with settings for an hour and missed it. I tell ya, if I knew what I was doing I'd be dangerous. Once again I fumble around, come up with a fix that works but it's for the wrong reason. Thank you @ahoeben. My knowledge of things Cura goes up another notch.

It could be. I just use the program. The thing is that the printer settings are submitted by other printer manufacturers or (as in the case of Creality) by users of the printers. I would think that Ultimaker checks them a bit but they don't go out and buy other manufacturers printers and build other peoples software for them. "Open Source" does not imply "Non-Profit". Ultimaker sells their own printers. Regarding the material settings I would think that they come from a manufacturers recommendations. Does Ultimaker have a laboratory where they test other manufacturers products? It would seem unlikely. Cura is not bug free but from all of the posts here and there on the web in the year I've been hanging around these sites, there aren't a lot of bugs. This problem may indeed be a bug in Cura but it involves a third party description of their printers capabilities and a fourth party description of what their filament requires.

I don't get involved in Ultimaker machine discussions because I don't know them, but gcode is gcode. If the instructions were in the file then the printer should have read them and followed orders. Post that gcode file. I don't use the plugin either but I do find the topic interesting. I also find the noises interesting. The sound of words and concepts flying by about 3 meters over my head gets my blood flowing. (geez...I gotta get a life.)

Using Cura 4.8.0 With the part vertical and the Anycubic Mega X the active printer I got the same results as you - no slots. So I switched to my Ender3Pro and the slots showed up. All features are there. Nozzle was .4 and the layer height was .2. So it isn't Cura. That leaves your printer definition and/or the material profile. I reloaded the 3mf file and with the Anycubic active selected "Generic PLA" and when prompted I hit "Discard Settings". The slots and all other features showed up. I then re-selected Generic PETG and re-sliced and everything was good. All the model features were there. There is something in your printer settings that is interfering with the material settings or vice-versa. I made some changes to the Cura print settings and saved this 3mf file. Take a look. You will probably want to make further changes. If the slots disappear again you will have found the setting that is causing the interference. Rev_AI3M_HS_B.3mf

This is likely the best way https://youtu.be/lnH9Nze5cHM?t=43 . It's certainly the most complicated way. The thread is HERE but it doesn't go into any real detail on the motor mount or wiring. I won't be doing it, but I like it.

I think you have multiple problems there. The support appears to be extremely under-extruded. There is a lot of stringing and oozing (usual a problem with the temperature and retract settings.) What printer is that? If you could do a "File | Save Project" and post the .3mf file here it would really help. The gcode file would be less helpful but maybe put that up here too.

When you customize a machine you make it yours, and what little support there was goes away. Good job chasing this down. I like the "Present Print" code that slides the table all the way forward after a print. It's a personal preference, but I would copy that "Old" End Gcode back into Cura. The BLT code is done once the initial G29 is finished doing it's thing. Being the nasty person I am I'll take a moment to remind you of a line from your opening post: "Now before you say bed leveling, temperature heating or bed adhesive. ..."

What exactly is your wrong result? Support Density = 1% and Support Line Distance = 120mm. That is very, very sparse support. Almost none. Are features missing as well? Since they don't appear in the preview they are either too fine to print (doubtful with a 0.3 nozzle) or is it possible there is a flaw in the model?

It isn't likely that is related to the No Slice issue. The build plate size is in the Manage Printers | Machine Settings dialog. The default numbers are in your machine definition *.def.json file. Something has changed to change your build plate size in Cura. When you open a 3mf file it can change your active printer. Did something like that happen?

And if you would post one of the "bad gcode" files here we can peruse it. Sometimes the clue to where to look in Cura is within the gcode.

Every once in a while I print a calibration cube. There will be a pause at height in there and I'll throw in a pause from the LCD, from my own software, and practice recovering from an aborted print. There are certain steps you can take to save a print, but it certainly isn't possible for every mess we make.

Thank You. Not many people say that to me. Cura makes pretty good time estimates based on accel, jerk, and printing speed but every printer is different. I like the "fudge factor" of that plugin because it allows a user to account for their particular printer. Play with it. Who knows, you might find it useful. It has to be worth at least what you paid for it.

I would guess it depends on how your firmware is determining the time. The print time is expressed differently in various slicers. Cura: ;TIME:62183 seconds Simplify3D: ;Build time: 3 hours 29 minutes Craftware: ; Time:13.9775 min KISS: ; Estimated Build Time: 40.67 minutes In Cura and Craftware the time estimates are at the front of the file. In the cases of Simplify3d and KISS, the lines are at the very end of the file so it appears unlikely that the printer is reading time estimates from the file but rather extrapolating the time from the file size, and the bytes/second it's processing. It's in real time and so is correct especially at the end of a print when it just needs to display "here's how long I've been running". The ShowProgress plugin adds this to every layer: ;LAYER:0 M117 1/254 | ETA 5H13M The M117 line updates the projected ETA on the bottom line of the LCD at the beginning of every layer. (It has no effect on the internal time that the printer is keeping.) With a close fudge factor the ETA would be something approaching reality. If you change the feed rate by making an adjustment on the LCD then of course the gcode would not compensate. On my Creality printer the estimated time as expressed by that M117 line with a fudge factor of "1" is very close to how long a print takes to complete at 100% feed rate. The response to M31 on my Marlin firmware is "echo:Print time: 5245s" which at least implies that it's simply reporting the time since a print started. How about putting an M300 and an M240 at the start of the last layer. The M240 could trigger an alarm rather than a camera. Or at the end of a print have the table move to MAX Y and hit a switch that triggers a relay and rings a bell? Or maybe triggers a blue-tooth device (or something) to send a message to your phone? That's what I have sir. I'm going to stop now before I start bouncing laser signals off the moon.

@gr5 that is a nice video. I'll have to watch it again tomorrow when I'm more awake.