GregValiant

From HERE... "PLA falls in the category of non-conduct filaments because it has a resistivity in the order of 1016 Ωm, similar to most other plastic types, making it a great electrical insulation material as long as it is in a solid state (cold). Once heated, PLA will soften and become more conductive and no longer safe to use as an insulation material." I would think much depends on the thickness of the part, the distance between the conductors you are insulating each other from, and the amount of current that is flowing. The above snippet would seem to imply that if the conductors heat up (due to high current flow or simple ambient temperature) then the PLA can deform and allow the conductors to get closer to each other. Any current passing through a plastic insulator can (in time) cause a carbon trace to develop. When that happens the resitivity drops dramatically as the carbon is conductive.

It is nearly infinitely thin there. This image is with a .15 nozzle at .1 line width. They still don't connect. You need to add some thickness there.

If you change the view of that folder with the Cura file to "Large Icons" you should get the thumbnail images back. They can take a while to load though. Windows 10 has the Preview sidebar on the right and there are two buttons at the bottom that toggle the view. When you select a file the preview should come up.

You have looked at your Initial Layer Flow %? The flow of layer 0 is just about exactly 5.00% of the flow shown in Layer 1.

Most Ender's and the CR-10's have the same hot end. They are prone to clogging at the bottom of the bowden tube where it is supposed to butt up against the nozzle. After printing for an hour or two the clog gets enough warm material built up that it impedes the flow of the filament to the nozzle. It usually doesn't skip steps, but extrusion gets iffy. Zooming in on your first image - the extrusion looks iffy as a couple of layers hardly extruded at all. The problem can be caused by a combination of things. The cheap Creality hot end, insufficient cooling from the hot end fan, too long retractions, a lot of retractions in a short distance, bad coupler on the top of the hot end, the bowden tube is damaged where it meets the nozzle, the bowden wasn't cut off exactly square, heat creep, and on and on. Those parts probably don't have many retractions so you should focus on the hot end. Heat it up to 180 or 190 and then pull out the bowden tube. Remove the nozzle. Find a piece of wire (like a coathanger) that is about 1.75mm and shove it down through the warm hot end. See if a plug of plastic comes out. Cut 6 or 7mm off the bowden tube and make sure it's a nice 90° cut. Put the nozzle in and snug it up, then back it off 1/2 turn. Shove the bowden tube back in so it is hard against the back of the nozzle and put the lock clip back on the coupler. Then snug the nozzle the rest of the way so it's tight. That should provide a good seal to the bowden. Now, I did all that several times. I replaced the couplers with Parker-Hannifin models, new bowden tube and a little jig to make the cut exactly square. My Ender 3 Pro continued to have problems an hour or two into a print. I changed the hot end fan from the stock 4010 to a 4015 ball bearing model with about twice the flow. Added a baffle inside the hot end housing to direct the fan discharge right at the heat exchanger. I still had problems. At that point I bought a new hot end that was a clone of the stock hot end. That lead to a new saying "When you clone a POS you shouldn't be surprised that the clone is also a POS!". Now I have an all metal hot end and it's all good. It still needs the bowden trimmed back once in a while as it does rotate in the coupler and get chewed up, but the printer prints very well.

What printer is it?

It's all about practice. I think that would be true with a BLT system as well. The BLT will confirm that the midpoint is higher and it will build a profile to raise the Z a bit in that area. I don't know how many layers that Z adjustment lasts for, but eventually it will be 0 and the gcode will just be read and adhered to. There is a gcode command (M32) that doesn't get used much that will call another gcode file and then return to the first one. The downside is you have to know the DOS 8.3 name of the file you are calling. All files have a DOS 8.3 file name. It can be a pain to tease it out of Windows. M32 P !SAWWED~1.GCO# If that line was in my StartUp gcode it would call the SAWWED~.GCO (actually named "Saw Wedges.gcode") every time I print a file. There is more info on the MarlinFW.org site. I've played with it and it works on my E3Pro. I level before every print. It takes maybe 30 seconds using the buttons in my app and most of that is waiting for the head to move. So my prints always start perfect...and then I remember that I needed to move the Z seam and I have to abort anyway. A BLT wouldn't help that at all.

Printers are essentially simple machines. They are four axis robots that have a hot end instead of an "end-of-arm" tool. Adding a BLT or CRT auto-level system is fine but it requires certain hardware changes and firmware changes to work correctly. When they are working I guess they are OK. They can be difficult to integrate into some printers. (The word "Nightmare" has been used a lot when referring to BLT installation.) I would think Creality has a kit for your E3V2 and it would likely require a firmware change to accommodate the G29 (or M420) command that calls the Auto-Level routine in the firmware. A kit may not remove the word "Nightmare" from your experience. The probe is usually a spring-loaded switch. The print head moves down and the switch trips. The first problem I see is that the point in space that the switch trips at is slightly variable. My Z stop switch is so inconsistent as to be useless and I don't see the same people who made that switch making a perfectly repeatable Z probe switch. My piece of parchment paper is ALWAYS .05mm thick. By moving around and checking a bunch of points with the probe it builds a map of where the build surface is in space. The Z point that the probe trips at is below the tip of the nozzle. The distance between the "trip point" and the physical tip of the nozzle is called the Z-offset and you have to dial it in so the printer firmware knows where the nozzle is in relation to the probe trip point. The problem that most people have with leveling by hand is using the LCD knob to move around. You can build a simple gcode file that will do all of that movement and pause at the end of each move to allow you to adjust the bed to the nozzle. I went another way and wrote an app that includes printing from the SD card and a manual leveling routine. Just click a button and the print head moves to the next point. You just need to get your piece of paper there before the print head drops down. There are two attachments here. One is a gcode file that would be copied to the SD card. Print it before starting your real print and level with a piece of paper (CHEP has much the same sort of thing available on his webinar/youtube videos). The second is a zip file containing Greg's SD Print Tool which is my interface for printing and includes a leveling system. It is for Windows only as I code in VB which doesn't port to Linux or MAC. Here is the GUI for my app. The text box contains my Ender3Pro's response to M503. The app is unsigned and you would likely need to fool your anti-virus to get it to install. Unzip the file and run the setup.exe file to install. Gregs SD Print Tool.zip LevelWithTemps.gcode

When the printer mainboard sees the 5volt power at the USB port it will reboot. That is hardwired as an external reset. It doesn't mean that the printer can communicate with the printer. Some USB cables that come with phones are for charging only. There is no data capability. In that case you would see the printer re-boot because it sees the 5v coming in, but no data can transfer. If you have USB printing enabled there is a box on the Monitor page that allows you to send a command to the printer. With the printer turned on you can try to send G28 and see if the printer properly Auto-Homes. You can try installing Pronterface/Printrun and see if you can communicate using it's interface. Most of the time it's a driver problem though. With the printer plugged into the computer Windows Device Manager should see the printer and note that it is a USB-SERIAL device. If it doesn't then it is definitely a driver issue. Greg's Rule #2: NEVER upgrade an operating system.

If you use a support blocker as a cutting mesh and configure the per model settings to modify overlaps and set the top/bottom layers to 0 then I think you can eliminate those lines you don't like. That will impact the wall thickness of the ribs though. The Cura slices are horizontal while the wall thickness is normal to the surface and since the angle looks to be about 45° then the actual wall thickness will not be .8 but rather .56 on those horizontal areas (because the yellow lines are added to make the wall thickness correct). The vertical walls will still be .8.

It's hard to tell from the screenshot but I think they are there because that area is at an angle and Cura is attempting to maintain the wall thickness. If you would use the "File | Save Project" command and post the 3mf file someone will take a look.

Look around here for Arachne Engine Beta and the Beta II release. It is slated to become Cura 5.0 and is much better at models like this. It uses variable line width and those thin spokes will slice. As Cura progresses I think it is getting more accurate and consequently fussier about the model geometry. Notice that the spokes are yellow. That is the color for infill and you really want those spokes to be defined as having "Outer Walls" which would make them straight single passes which would show as red in the preview. Depending on your dimensional requirements you could make the Outer Wall Inset -0.1 and the model will slice. Another problem is that the features aren't very tall. At a .28 layer height the nozzle is moving up past the features after only 1 layer. Moving down to a 0.2 nozzle size may be an option as well. There is a lot of bridging going on however you decide to print this. I think your 160mm/sec print speed for a model like this with such fine features is iffy at best. You should try putting together a profile for small delicate models. Viewing the settings in the profile (in that 3mf) it seems they may work well for larger prints but on something small and essentially round, going with a less aggressive profile I think would help. That has nothing to do with the slice of course, just sayin'.

I'm not real fond of that profile. For one thing, the 200% line width on the first layer can cause adhesion problems and jst looks crappy as well. If you look at the bottom of that Benchy you will see that the iindividual extrusions aren't welded together. Here is another way to print that. I've moved the Zseam to the back left and set the temp to 205 as that photo does make it look like the filament didn't like 210 (210 is my normal print temp for PLA). I'm also sticking in a calibration shape I made. It isn't better (or worse), just different. GV_3DBenchy.3mf GV_CalibrationShape.stl

Hello. I was ready to hit the send button so I did anyway. As nallath says (and as the popup indicates) you can't alter or modify the Gcode in Cura. But there are ways. If your earlier Gcode files were created with "Origin At Center" ticked, and if your printer is an "Origin At Left Front Corner" machine, then the print will indeed show up at the left front corner. There is an indicator in the gcode file. In the beginning header information of a Cura gcode file are the Min and Max movements of all three axes. If MinX and/or MinY are negative, then the gcode was created with Origin At Center ticked. The way to print that gcode on an "Origin At Left Front Corner" printer is to change your home offsets. You do that by adding an M206 line just before Layer 0, and then another line at the end of the gcode file that re-sets your home offset. If you have never set the Home Offsets on the machine, then they are 0,0,0 and the location is the Auto-Home position. If you were to add M206 X-100 Y-100 Z0 to the front of one of those files you want to print... Example: ; ;End of StartUp Gcode ; G92 E0 G92 E0 G92 E-3 ;LAYER_COUNT:135 M206 X-100 Y-100 ;LAYER:0 At the end of the file: M82 ;absolute extrusion mode M104 S0 M206 X0 Y0 Z0 ;End of Gcode

BTW - This is the report from the STL repair site (https://formware.co/OnlineStlRepair). Not good. When attempting to repair the file, all the roof holes disappear. --> 104 Naked edges (?) --> 10 Planar holes (?) --> 0 Non-planar holes (?) --> 16018 Non-manifold edges (?) --> 18179 Inverted faces (?) --> 0 Degenerate faces (?) --> 7181 Duplicate faces (?) --> 0 Disjoint shells (?) -> Repairing: 100.00% ----- Repair completed in 44801ms ------ -> Vertex count changed from 9688 to 22220 (+12532) -> Triangle count changed from 29348 to 44622 (+15274)

Here is a close up of one of the roof holes (using MS 3D Builder). The facets of the hole that are highlighted in black look OK. The red arrows indicate facets that are wrong. Those "wrong" facets are the problem as they are what is making the model "non-watertight". There are a lot of those problem areas in that model.

+1. Before going any further - that model is not repairable. "Non-Watertight" means there are open seams between surfaces. In the case of the holes, the cylinders that comprise the periphery of the holes (in the roof and sides) don't meet the walls. Consider a wall of your house that you fill full of water. It will leak out around any electric outlet cover because the cover is not "watertight" to the wall. Tall skinny vertical features (like your walls) are tough to print. The nozzle itself can knock them around while printing. If I get stuck HAVING to print something like that then I would build in removeable supports to keep the walls from wobbling as the nozzle travels and as the bed moves back and forth. My first choice would be to break it down into parts and make it an assembly. The three separate parts could be printed flat and both PLA and PETG are good with cyano-acrylic Super-Glue. The model still needs to be "watertight" though. Back to the CAD program for that.

"...but when i slice it, it put's a wall behind the holes." I think what we have here is "expected behavior". Typically, a model is a solid through-and-through. When you put a bore in it's like drilling a flat bottom hole into a piece of wood. If that hole doesn't go all the way through then there will be a bottom. Setting Infill Density to "0" doesn't change that. If you want a piece like you show (a cup with two holes in the wall and open into the interior) then it needs to be designed that way. The smaller model here is your cube and the larger one is my design. In CAD - I added a "shell" feature (emptied the interior) and ignored the top (so it has a top opening) and then extruded a 24.4mm tube 20mm long through the side wall. Then I cut a 20mm diameter hole through the tube. That left the walls of the tube 1.2mm thick. The other two holes are just through the wall. Since there is no "interior" because of the shell feature, there are no bottoms in the holes. The 3mf file I've attached has both models. My model would require support on the inside to hold the tube up while it is printing. GV_CE5P_testholes.3mf

Search around here for Flashforge without the model number. There have been posts regarding getting one to work with Cura and they should come up.

When @gr5 refers to a Cad model tree I believe he is talking about the list of operations that have been performed. There would be the initial extrusion, and you have two holes. The tree for that model would have those three items in it. You have a 63mm cube with two 20mm diameter holes. One hole is 6mm deep to a flat bottom and the other hole is 19.83mm deep to a flat bottom. I didn't find any errors in the model and it sliced just as it looks like it should.

Check the fan numbering in the Machine Settings. I think Extruder#1 should be fan 0 and Extruder #2 should be fan 1. If the gcode is trying to operate fans 1 & 2 then there will likely be a problem when internally the printer has them numbered 0 & 1.

If the USB sees 5v it will cause the printer to reboot. It's the emergency external reset. Never connect during a print.

Close, but no cigar. I don't think it will work. It looks like the Jerk only changes when "TYPE:FILL" is there. For the outer and inner walls the Jerk doesn't appear to be inserted.

No travel settings and no cooling settings on modifier meshes. Try again.

Somebody mentioned changing the cooling based on the feature. That isn't all that hard to do as the features are in comments throughout the gcode. Unfortunately within a gcode file you won't find anything that says "overhang starts here". It would be really tough to code for. If a support blocker could be configured...hold on... I feel a workaround coming on. You must have "Enable Jerk Control" turned on for this to work and all the Jerk values need to be the same. For this example lets say we have Print Jerk and all the little jerks at 10. The fan speed for the overhangs will be 50% and the rest of the time the fan will be off. So you take a support blocker and size it and place it covering the overhang. You select Per Model Settings / Modify Settings for Overlaps and make "Print Jerk" one of the settings. Set the Jerk within the support blocker to "1". Every time Cura slices through the blocker it will add M205 X1 Y1. When the nozzle leaves the blocker Cura will add M205 X10 Y10 (our default Print Jerk). You add two Search and Replace post processors. The first one is: Search = M205 X1 Y1 Replace = M106 S128 (fan at 50% PWM) and the second one is Search = M205 X10 Y10 Replace = M205 X10 Y10\nM106 S0 The second replacement re-inserts the Jerk line so you don't have to go back and change the first line manually. I like it. Now it's up to you to prove why I shouldn't.