GregValiant

Play with "Infill | Gradual Infill Steps". You can start with your regular infill at 10% and that Gradual setting at 4 and the Infill percentage will increase as it approaches the roof giving good support to the first layer above the infill while leaving the lower area with sparser infill.

The author configured the firmware that way. There is no Gcode command for hot end fan speed. This is from the configuration.adv.h you linked. * Extruder cooling fans * * Extruder auto fans automatically turn on when their extruders' * temperatures go above EXTRUDER_AUTO_FAN_TEMPERATURE. * * Your board's pins file specifies the recommended pins. Override those here * or set to -1 to disable completely. * * Multiple extruders can be assigned to the same pin in which case * the fan will turn on when any selected extruder is above the threshold. */ #define E0_AUTO_FAN_PIN FAN2_PIN #define E1_AUTO_FAN_PIN -1 #define E2_AUTO_FAN_PIN -1 #define E3_AUTO_FAN_PIN -1 #define E4_AUTO_FAN_PIN -1 #define E5_AUTO_FAN_PIN -1 #define CHAMBER_AUTO_FAN_PIN -1 #define EXTRUDER_AUTO_FAN_TEMPERATURE 50 #define EXTRUDER_AUTO_FAN_SPEED 255 // 255 == full speed

What printer and firmware is that?

One is your canoe and one is mine. Can you tell which is which? Attached is the repaired model. I like Service.netfabb.com for model repair. They can even float boats and it's free. Canoo_fixed.stl

That 3mf file doesn't want to open correctly for me. Wrong printer. Wrong number of extruders. Something went on in the translation because instead of the parts having different settings the different extruders were given those settings.

That's a tough print. When I opened it the first time, there were no walls. I opened it again and made a couple changes to your settings and walls appeared in the over-lap areas. I started making changes and couldn't get rid of the walls in the overlaps until I discovered that the scaling of the separate parts was different. The main part is scaled to 70% in X Y, the two front parts are scaled to 75% X Y and the instep is 72% X Y. I changed them so they were all scaled to 70% in the X Y and that caused the overlaps go away and so did the walls at the overlaps. Unfortunately, I was not able to duplicate any of that. If there are no walls in the preview there shouldn't be walls in the print. I used Cura 4.9 and tried with 4.8. An oddity is that when I brought the 3mf model into Cura 4.9, the printer is a Voxel with 4 extruders. When I bring the model into Cura 4.8 it is a Ghost4 with 4 extruders, and the file name seems to indicate it's a CR10. Either way there is no offset between the nozzles but "Extruders Share Heater" is not selected. What printer is that?

@MadNess69 it was another post that provided the clue. When I export models from the detail files they have no relationship to each other and so don't merge correctly and have to be manually placed. If I localize the models into the assembly file and export them from there they do have a relationship to the origin and merge correctly. @3D_BOY CHEP or some other video master might have a basic video on slicing. Most STL files are solid. Hollow models (of characters and so forth) present problems because the wall thickness has kind of been set by whoever created the model. @kmanstudios might have some insight into that. If you have a particular model you'd like some pointers on, Load it in Cura, choose the settings you think work, and then use "File | Save Project" and post the 3mf file here.

When Creality went to the 32 bit boards (4.2.2 and 4.2.7) they missed a trick. The firmware for the TFT style LCD screen didn't play well with the new boards. It appears to be in the firmware since there is a specific switch in Marlin that needs to be enabled for TFT style screens. Since M0, M1 and M117 send messages to the display, and since the display didn't want to play nice, the printer ignores all three commands so in your case M0 doesn't work. On the Marlin site you will find that M0 is defined as "Unconditional Stop" and it sends the message "Click to Resume" to the display. Now you know the problem is in the firmware and I think it can be resolved someway through the firmware. I'm just not sure of the exact fix. Neither Creality nor Marlin (who are really upset with Creality) would offer any help/advice/info regarding what was going on. I do have a workaround. Within the Cura Pause at Height dialog choose "By Layer" and set the Disarm Timeout setting to 1800 (that's the max without changing a line in the plugin). In the box "Gcode after Pause" enter "G4 S180". G4 is "Dwell" and the S parameter is in seconds so the 180 would give you 3 minutes to do something. If you were to make it S600 you would have 10 minutes. When the printer gets to the Pause - the routine will initiate, the head will park where you want, the M0 will be ignored, the G4 command will "dwell" the print for however many seconds. At the end of that time the print will restart by itself. The printer will not accept any input while in "Dwell" so you must wait for it to resume on it's own. There are two obvious drawbacks to this. #1 you have to be standing there because the restart is on a timer and #2 because it's a timer you better choose the time correctly or it can either restart too soon or take forever to get going again.

Random will leave a dot at any point on the exterior which does make it look like there are zits all over. The best you can really do with round objects is to allow the Z seam to be a line. That way you can turn the piece around and your opponent has to stare at it. I add 0.1 wipe distance, set the Z seam to User Specified, then Back, Front, or whatever and then set the Seam Corner Preference to None. There's going to be a Z seam and there isn't any place to hide it on round objects.

"I took precaution to cut the red wire (vcc+) in the usb cable" That shouldn't be necessary. On many printers that voltage does power the board with the printer turned off, but it also provides the "External Reset" that is sometimes required to straighten out a "situation".

Wow. Good one! With the model in place in Cura, and with your settings as you wish, use "File | Save Project" and generate a 3mf file and post it here. Somebody will take a look. This is a 4 year old thread though. It would have been worthwhile to start a new topic.

At the top of the settings dialog and to the right of the Settings Search box is a drop down list for Settings Visibility. The bottom option is "All". Select that so all settings appear. One of the Settings in the Mesh Fixes area is "Remove all Holes". Make sure that is NOT checked (enabled it will make your slot will disappear). Another one to look at is "Make Overhangs Printable" which also should NOT be checked. Adaptive Layers works well with curved features and not at all with straight vertical features (like your rectangular solid). When you print a sphere, as the print head rises and the sides of the sphere start rolling inward towards the centerline, Adaptive Layers will note the increasing angle and make adjustment to the layer height so the "steps" are shorter and consequently the finish looks better. The max thickness of any layer will be "Layer Height" but there will be layers that are thinner. When there is no angle, there is no "topography change" and the layers stay the same height. The end of this model is essentially spherical. There are other radii (not shown) that cause some local variation in layer height, but denser areas indicate lower layer height. When approaching the top of the spherical area, you can see that Adaptive Layers decreases the layer height. This also has the effect of turning a 4 hour print into a 7 3/4 hour print. It looks nicer though. Within Cura there is no option to vary the layer height according to the Z height. When I want to vary the layer height I slice multiple times and generate a gcode each time (a slice for each layer height change) and splice the Gcode files together at the Z height where I want the change. I do enough of that I wrote my own post-processor to automate it. For that model, I would print a lower piece with the slot, add the tensile specimen, and glue on an upper piece to trap the specimen. Too often we want to print things complete when an assembly is indicated and may be a better choice. If the "top" piece was printed with a slight raised area over the specimen location it would tend to clamp the specimen in place. I don't know if you would want that, but it's an option. For location, I put matching 2.2mm holes in each piece and set "Horizontal Hole Expansion" to 0.1. That allows me to super-glue pieces of filament into one piece which are my locating pins to align the second piece.

"At least according to the marlin config file from creality". Over the course of the last year or so, Creality ran out of various parts and made substitutions. Among the parts that they substituted were mainboards. They grabbed what boards they had available, did quicky patches to the firmware, installed the boards and sold the machines. There were issues and one of them was that G2 / G3 was not enabled. Another issue was the M0/M1/M117 glitch that killed the ability to use the "Pause at Height" plugin. Yes, it's a 32 bit mainboard but memory is memory and I've seen no mention of memory space in regards to any Creality mainboards. The problem was severe last summer and into the fall. Since then it seems to have gotten better but other than tracking complaints (like the G2/G3 thing) it's hard to tell. Creality is quick to blow their own horn in regards to their positives and completely silent about any negatives. My Ender 3 Pro with the 8 bit board reads and prints Arc Welded gcode perfectly. Cura however does not (but I think I read someplace that somebody was working on that). If your prints look like current Cura previews then G2/G3 is not enabled.

Hi @geert. I switched it to PETG and it's been fine. It does have hooks that go into the dash vents and yes, that was the main failure point. It's here in sunny Florida where it will be 90F° again today. Thinking more on the topic of annealing (which we used to call "normalizing")... 3d printing involves so many disciplines it's amazing. Mechanical engineering, electrical engineering, thermodynamics, computer engineering, software design, material science, geometry, trigonometry, physics, (I'm sure there are more) and a firm users knowledge of a few different pieces of complicated software. They all come together in a little rectangle measured as LineWidth x LayerHeight which in a lot of cases is just .08mm². There sure is a lot going on in that little area. I have an aluminum heating plate, a glass plate that floats on it (it's edge located so no clips), and a plastic part that may be made from a few different materials and that is bonded to the glass plate (at least for a while) and each has it's own coefficient of expansion. Once the aluminum and glass are up to temperature (admittedly not very high for those materials) the plastic gets put down at 210° to 240°, sometimes with a fan running across it and sometimes without. So it goes through a quick freeze. The bottom layers are at the build plate temp and as the model grows, the point source of 210-240 travels around but the build plate temp doesn't get through the plastic very well (poor thermal conductivity) and the part is cooler at higher layers. So where is the stress and why does it build up? Is it a function of part geometry as different mass areas (10% infill vs @helmut92's 40% infill) have local areas of thermal inertia? Is it affected by height off the build plate? Do large horizontal holes affect it? "Maybe", "Could be", "Possibly" and "I don't know" are all good answers. I tell ya, Everything affects Everything and 3d printing involves almost everything.

A G0/G1 could be done as a final move in the Gcode file before the cooling routine. The G28 is not a good idea as it would raise the Build Plate and occasionally a part might be in the way. G0 X0 Yqqq Zqqq ;change the qqq's to what's appropriate for your printer. Your Max Y and Max Z would work. If you were to put a move into your End Gcode it would always show up in gcode files, but you could use Cura keywords. G0 X0 Y{machine_depth} Z{machine_height} which for my printer is entered into the gcode as Y230 Z250.

Set the Build Plate Adhesion Type to "None", and the Horizontal Expansion to "0", and slice the model. Generate the gcode file. At the beginning of the gcode file will be MaxY and MinY. MaxY - MinY + LineWidth = Y Dimension from Fusion For a 25 x 25 calibration cube sliced for my printer it would be 127.3 - 102.7 + 0.4 = 25. That would be correct. If it matches the model then you need to look at the printer again.

This is with a support blocker configured to "Print as Part" and the SplitBoard is moved away a tad. The Horizontal Expansion is at .26. The block forms a bridge between the parts.

It's a narrow gap and your Horizontal Expansion is .26. So .26 is added to both parts and the gap closes. It would work if you moved the two parts farther apart. I would suggest a brim as well. The small footprint of SplitBoard1t.stl will likely fail without a wider base. If you set the Brim Distance to .1 it will come off easier.

In Cura 4.7, Creality gave their printer definition files a default Maximum Resolution of .025 which was much too fine for the printers. There were just too many calculations required and the print head would stutter as the processor tried to keep up. The fix was made in 4.8 when the Creality definition files were updated (by Creality) to a Maximum Resolution of 0.5. That eliminated a lot of very short line segments and printing was much improved. So in Cura, check "Mesh Fixes | Maximum Resolution" and set it to 0.5. Maximum Travel Resolution is tied into it. Maximum Deviation can stay at .025. Also having an effect on circles is the Accel and Jerk settings. Once they get much over 500 Accel and 8 Jerk (Creality Ender 3 Pro with the 1.1.5 mainboard), stuttering can start. The Arc Welder post processing plugin was developed to change the line segments (that comprise any circle) into G2 and G3 commands. Instead of the printer/processor having to calculate 200 gcode commands to get around a circle of radius 32 it would be 1 command with Arc Welder. This has a huge effect when printing via USB as not near as many commands are needed to be sent and processed. There is a lot of overhead required in the firmware (36kb?) when G2/G3 are enabled so of course Creality simply disabled them and Arc Welder doesn't work on some newer Creality models. At any rate, my printer is an Ender 3 Pro with the 8 bit board. Printing from the SD card is my preferred method as USB printing from Cura is no longer maintained (no current Ultimaker printers use it). USB Printing from Cura also has a peculiar drawback due to buffer over-runs (the opposite problem of stuttering) and printing will halt as a Cura generated M105 temperature check gets truncated and is read by the printer as an M1 which is "Unconditional Stop" (Pronterface/Printrun suffers from that same problem). There is also the problem on Windows systems of Microsoft pushing out updates in the middle of the night causing the computer to re-boot which re-boots the printer and your print that looked terrific when you went to bed is sitting on the build plate like a lump.

With a model loaded in Cura, and with your settings configured, select "File | Save Project" and post the resulting 3mf file here. Post one of the "good" gcode files as well. It may be something in the "Start-Up G-Code" that is causing the problem.

The USB Printing option in Cura is un-maintained (Ultimaker printers don't use it anymore). Connection is supposed to be automatic so there are no options to pick a port, or a baud rate. Check in the Cura Marketplace under Installed and scroll down to "USB Printing" and make sure it's installed and checked. When you are sure it's installed and active... Turn off the printer. Close Cura. Disconnect the USB cable. Start the printer, let it finish booting, and re-connect the cable. The printer should re-boot when it sees the 5 volt signal on the USB cable. Make sure no other running software is accessing the port. The only thing connected via USB should be the printer. Start Cura. Go to the Cura Monitor tab. Sometimes it takes a minute or two for the connection as Cura cycles through the available ports and tries different baud rates until there is a response that Cura understands. That all works fine with my Ender 3 Pro.

There are times when the models are exported from an assembly. Those models should have a relationship to each other and as Ahoeben says, Merge works well. There are other times when the models are exported from their individual detail files and so there is no spatial relationship from model to model. In a sense it is like putting an assembly file together in a CAD program. In a CAD assembly program, each part has to be moved into place separately, constrained to the position, and the location data and assembly constraints are contained in the assembly file while the part itself is included only as an external reference. In the image below all 4 parts were brought into Cura, but each came from the detail file. They were merged but are co-mingled. The "feet" of the assembly are the same part (not shown and opp) and so merged exactly into one another. In this image the parts were brought in from an assembly file then selected and merged. They are in the proper relationship to each other, but they were located somewhere around the front passenger seat of a car. Moving this assembly to the build plate, it won't slice yet. The assembled models fit the build plate but the current bounding box does not. Ungrouping and re-grouping allows the model to be placed and sliced. It isn't always necessary to do tricks. But sometimes...

This is just for general information. Cura always imports STL files by center of geometry (the bounding box). Consequently it can be fooled. If the center of geometry for each part in an assembly is the same in the XY then all models come in wherever, but you can re-locate them to the center of your build plate. They can be aligned where they belong in the Z but the XY dimensions of the bounding boxes need to be the same. The X Y dimensions can be up to the size of your build plate. Take your base purple piece. All the other pieces will fit in it's "shadow". For each other piece (you would need to consider the green pieces separately since they are at different locations) add a feature that is the shape of the purple part in the plan view, and extrude it from 0 to -1 in the Z. The X0 Y0 for every part would be the same. When they come into Cura, move each one down -1 in the Z to put the new feature below the build plate. Here is a screenshot from AutoCad. Three pieces have the base added (they don't need the chamfer). Here is the assembly in Cura. The base purple piece is at 0,0,0. The other three pieces came in off the plate but I moved each one to 0,0,-1 (to hide the base below the top of the build plate.)

Geez, and I thought the lines were showing air turbulence as the Benchy went speeding along.

The second part shows a lot of under-extrusion. Since the lighter area isn't near as dense it shows as a lighter color. I always ask myself "What's the last thing I did?". If it was that brass driver then check the set screw again. There is also the ever popular "partial clog" in the nozzle to consider. Something interfered with the filament flow, and it started all at once.