GregValiant

That is further proof that computers are not an exact science.

That shouldn't happen. I have the same printer (although mine is a couple of years old). Auto-home the print head and measure the distance from the nozzle to the edge of the glass. Post the X distance and the Y distance. Neither should be more than a couple of mm. I have set my home offset to the right front corner when I needed to mirror a gcode file one time. Your newer firmware is likely configured differently, but it shouldn't be that different. The printer looks at a gcode file and puts the origin at the Home Offset position. Right now since you haven't defined the Home Offset position the origin is at the Auto-Home position and so the prints aren't centered. Another thing to look at is to: Auto-Home and then move the print head to X=235 and Y=235. Can it get there? At that point the nozzle should be just inside the right rear corner as the Auto Home position is outside the left front corner.

I wonder if that works because adding the parameter changes how the firmware reacts. My older Marlin always sends "Click to Resume" to the LCD. When adding the S parameter, and not noting any text for the LCD, might change something else. As an experiment - if you would try "M0 S300 Click to Resume" it might provide a clue as to why a plain Jane M0 doesn't work. My theory is that the Creality firmware on the 4.2.x boards doesn't play nice with the TFT style LCD's. An experiment might prove or disprove that.

Some settings (like Layer Height) are not enabled "Per Model". A few are noted within the FDMprinter.def.json file. I don't know if that is a complete listing. This one is noted... "retraction_amount": {blah, blah, blah... "settable_per_mesh": false, "settable_per_extruder": true I'm sure there is a reason for that.

There was just another post here that said using M0 with the S parameter worked. Something like "M0 S300". That would give you 5 minutes to change the filament, insert nuts, whatever. "S" is in seconds. If you manage to get the filament changed in 1 minute then clicking on the LCD knob is alleged to resume immediately (instead of waiting for the timer to run out as when using G4). I think Cura is always going to add M0 without the S parameter. You would need to manually change it in the gcode, use Search and Replace AFTER the pause at height command, or maybe enter "M0 Sssss" into the "Gcode after pause" box.

I know nothing about firmware. Can I assume that the printer worked "as delivered" with the stock firmware? If you would load a multi-color model and prepare Cura to slice it, and then use the "File | Save Project" command and post the 3mf file here someone will take a look.

I'm just a user but I have written a little code (poorly) for serial port communication with everything from a 60 ton Baldwin Universal Test Machine to robotic puppets. I tried to build an app to be a print server that would transmit a gcode file to a 3D printer. After numerous hours trying to emulate the printer in my code (so the transmission rate would match the rate of usage by the printer) I gave up and threw it all away. That may have been what happened at Ultimaker as well and they chose to go a different way. I don't think anyone has updated the "USB Printing" plugin for several years. Since it is open source - you are welcome to try. If you really want to print over the USB then you will likely have to invest in an Octoprint/RaspberryPI setup. That seems to work pretty well and avoids the problems like (in my case) Microsoft pushing updates that cause a re-boot, or buffer under-runs and over-runs that cause stuttering and missed commands, or M115 being truncated to M1 by the printer causing an "emergency stop". All of those are critical errors and they absolutely do happen.

If Cura doesn't see any temperatures in the StartUp G-Code (and you currently don't have any) then it will add temperatures before the StartUp GCode to insure there is no Cold Extrusion. I'm not sure how it picked T0 to heat first. It may be the Initial Extruder (or Cura simply defaulted to T0). Since it picked T0 to be first then the other three extruders are only brought up to Standby Temperature for the material each has loaded. Although T0 does heat first the M109 S215 line insures it is up to printing temperature prior to the StartUp Gcode being executed. The other three are not...they are brought up to Standby Temperature. Which extruder do you need active for"...I have code to run after tool selection."? Do you need each extruder in turn up to print temperature? Are there prints you run with (for example) T1 and T3 active but with T0 and T2 disabled? (That combination would be tricky.) Is the "...Code-After-Tool-Selection": Within the gcode file or are you calling a separate sub-file with M32 or something similar? You could post a gcode file that you like and that contains the Code after Tool Selection. You might have to hand code parts of it. Without knowing exactly what you want to happen (and when) I'm shooting in the dark.

OK @gr5, You made me look up gnomon and I'll buy that. But your claim that kids can stand in one place long enough to cast a shadow is a bit hard to chew. What goes around comes around though. Both of my boys are chasing kids of their own. The wife and I laugh and laugh and pass out more candy. Then we go home while we can still see the sundial.

"my test print came out somewhat forward instead of centered." This is a popular one. Get out a thin metric ruler. Auto-Home the print head. Raise the Z 5mm or so. Move the X until the nozzle is 2.5mm in from the left edge of the glass (not the bevel, the edge). Move the Y until the nozzle is 2.5mm in from the front edge of the glass. Lower the Z to 0. On the LCD select Set Home Offset and then Save Settings. In Cura, go to Manage Printers and then Machine Settings. Set the build plate to X(width) = 230 and Y(depth) = 230. That's it. Now both the printer and Cura know where to start printing from. The midpoint of your glass will be X=115 and Y=115.

There aren't many printers that need the X, Y, or Z calibrated. What I usually do is clean up the surface by giving the cube a couple of strokes with a 10" mill file. That is enough to knock off any high points that might be getting in the way of the calipers. A setting you could try is "Coasting" in Experimental. It can cause problems with the initial prime after a retract and a long combing move, but on that cube it won't be a problem.

Remember - the printer doesn't know where the bed is and neither does Cura. The bed needs an initial leveling. The ABL system doesn't move the bed up and down, it simply creates a virtual grid showing where the bed is high or low. The firmware then adjusts the first layer to the grid. It won't make an adjustment if the bed is off by 10mm. Once you have the initial leveling close then the ABL should work. You might have to do (manually level) again once in a while, but not every print.

You have Cura in "Relative Extrusion" mode. These are the first few lines around Layer:0: G92 E0 ...................................................Reset the Extruder location M83 ;relative extrusion mode ;LAYER_COUNT:86 ;LAYER:0 M107 ;MESH:300 box.stl G0 F3000 X153.5 Y153.5 Z3 ;TYPE:WALL-OUTER G1 F900 X446.5 Y153.5 E964.51129 First extrusion is 964mm. G1 X446.5 Y446.5 E964.51129 Succeeding extrusions of layer:0 are the same amount (Relative Mode) G1 X153.5 Y446.5 E964.51129 G1 X153.5 Y153.5 E964.51129 The first layer is 3.0mm thick and succeeding layers are 3.5mm thick. That is why the numbers are different for succeeding layers. Those numbers work for 2.85 filament. When I first looked I thought they were for a paste extruder (7mm nozzle?) but the volume numbers work out to a standard filament diameter(?). If I might ask - what do you have going on there?

You should switch to the Arachne Engine Beta II version of Cura for this model. It is capable of filling in the middle "third" wall. With the current version set to two walls consider a capital "I". The vertical leg of the letter has two walls on the left and two walls on the right for a total of 4 walls. If the vertical leg is 1.2mm wide then 4 walls won't fit and the center remains empty unless you use "Fill gaps between walls = Everywhere". Essentially, the program only does an even number of walls across an object.

You are going to have to fool Cura. How you do it can be tricky. The slices Cura creates are horizontal (parallel to the build plate). Since that rib (or whatever) is only "near horizontal" the wall thickness is being compensated for. It's a trig function of the angle to the build plate. This is a 6 x 6 x 75 rectangular solid tipped up 10° with 1.2mm wall thickness (3 walls / loops). Here is the math. The top view is at 10° tip and the bottom view is at a 5° tip from horizontal. The infill is generated to maintain the 1.2mm wall thickness "normal" to the outer surfaces of the model.

Here is one of your photos that I zoomed in on. You can see the 2nd layer below through the top layer. That is some serious under-extrusion. Have you calibrated the E-Steps on the printer? All those extrusions should be welded together. With black filament you should not be able to see through a layer.

Cura will typically color Outer Walls red. The outer walls are yellow in the Cura preview and that can indicate that something is going on with the Line Width or maybe another setting. What version of Cura is that? If you can create a project file (File - Save Project) and post it here someone will take a look.

I'll ignore it after I make a suggestion... Change the beginning of the StartUp Gcode to this: ; start code T0 ; select tool 1 M140 S{material_bed_temperature_layer_0} M109 S150 ;heat the hot end to 150 G28 ; home all G29 ; do abl M109 S{material_print_temperature_layer_0} ; set extruder temp and wait M190 S{material_bed_temperature_layer_0} ; wait for bed temp ; G29 S1 ; use stored bed mesh etc, etc, Setting the hot end to a temperature below the "ooze" temperature can keep the bed clean during the ABL stuff. If you decide to go to different "Print Temperature Initial Layer" or "Bed Temperature Initial Layer" then adding the "_layer_0" to the replacement words will get those "Initial Temperatures" to turn up in the gcode. I added "M220 S100" and "M221 S100" to my end gcode in case I tune the flow or feedrate during a print. It will reset them to 100%. If I roll right into another print then changes I might have made won't carry over.

It's always humid here. Today it's 97%. I've found that if I leave the spool on the machine instead of properly storing it in a Ziploc with anti-desiccant packets, it will get wet. It prints poorly when it's wet and dryng helps, but causes the filament to get brittle. It prints at the same temperatures, just really poorly. PETG is really bad. If you are getting good prints at 240 then print at 240. When my printer has sat overnight and I turn it on, the bed and hot end are 3 degrees different. That speaks to the quality of components used in the machine. That can't even agree on what the room temperature is. The only way to check is to use a handheld infrared thermometer and verify what the hot end temp really is. In the end it's just a number displayed on the screen. I have a feeling that the LCD might say 240, but that the hot end is really closer to 210. What's important are the results. The business end of the thermistor must be in the hole in the side of the heat block. The little screw next to it should be just tight enough to keep it from falling out, but not so tight as to damage the insulation of the wires or they will short and cause glitchy results. And take some time to do your nails while you're waiting for the next print to finish. Geez.

Cura doesn't make things up. It displays and slices the model you opened. No matter what software I've opened your STL file in, there two cylinders, the big one has a partition inside with two rectangular holes, and there is a keyway feature. The little cylinder is a tube with no other features. This is from PrusaSlicer. It sees that file exactly the same as MS 3D Builder sees the file, and as Cura sees the file. The problem is in the export from the CAD software, or the way the pieces were modeled in the CAD software. If there are Unions or Subtractions in the models maybe they just aren't making it into the STL file. Here is an air cleaner and velocity stack I designed in Mechanical Desktop. This is how they were designed, and this is how the STL files represent them. There are tapers, revolutions, and lofts. Cura shows the features because the features are in the STL files.

The nose piece is separate (you can just see the seam). The big piece s 229 long as that is what will fit on my bed. It just isn't a fast process. There are some things you can do. On those pieces I made the "Infill Layer Height" 0.4 and the rest of the model is 0.2 layer height. You can do the same with Support Infill. The effect is that those areas only get printed every other layer. You may be able to kick up the print speed but there is a point where artifacts from shaky mechanicals start to show up in the print. I could have gone to a .6 nozzle to speed things up, but then I have to print slower. The fin was at 75 with 50 for the outer walls. There are a lot of long straight runs on the print. When a model has a lot of curves and circles I usually print them at 50/35.

Congratulations. That is very nice. Now you get to figure out how to get fancy with the two extruders. That's where the fun really starts. A few years ago the wife fell in love with the vintage trailer thing. We have a 1972 Shasta Compact (boy do I have a lot of hours in that little thing) and Shasta's have wings. I printed this wing for the roof of the tow vehicle and since we are in Florida, I thought a shark fin would be nice. Two pauses for color changes, one pause to add the 10-24 nuts, one pause to slide in the 3/16 x 9" steel rod for a backbone, and two Search and Replaces to adjust the accel and speed as it got taller. It set my personal record with 8 post-processors active (there are two that I always leave on). The mount brackets are PETG but when I tried to make the big part of the fin with PETG it pulled off the bed and warped badly. I had significant elephant ears on that one but they weren't enough to keep it from warping. We'll see how the PLA holds up in the sun with that piece of steel running down through it. The star is a separate print 1mm think and is glued into a pocket. Total print time was 14 hours (if you were wondering) and I still have to do one for the right side of the vehicle. The spring trip of the Tin Can Tourists is coming up and I think a lot of people will be jealous when they see us pulling into the septic dump station in such fine style.

That's very good for 100mm/sec. On a model that small it probably didn't hit unless your accel was dialed up as well. It probably hit 75 to 80 though. I use Z-Hops when there are features like large horizontal holes that come to a feather edge at the top before the gap gets bridged. Those feather edges always curl up, and without Z-hopping over them the nozzle can catch on the print when it crosses the gap, and that can cause the print to break free of the bed.

I'd say that looks pretty good. What was the print speed? Did you have Z-hops enabled? I was thinking that any problems might show up in a print with Z-Hops. That's what that test kind of emulates.

I put together this little gcode file. It bounces the Z up and down four times in the left front corner from 0.1 to 0.0, 0.2 to 0.0 etc and then repeats. Then it moves up 10.0 to 0.0 four times and repeats. Get a piece of paper ready. Auto-Home and level the bed paying attention to the left front corner. Keep the paper ready and print the file. After the Z drops to 0.0 there is a 3 second pause before doing the next move. My thought is that you may see what is going on. You did not mention Auto-Homing when returning to the left front corner. The end stop switches on Creality printers are notoriously inconsistent. I ended up changing the G28 in my start Gcode to G28 X Y so it wouldn't re-home the Z after I leveled (piece of parchment paper) because the Z=0 location kept changing. Z Lag Test.gcode