GregValiant

You are sure that "Extruders share heater" is checked? In Cura 4.8 it is in the "Printer Settings" which must be loaded from the Marketplace.

"...it is perfectly centrated in the 0,0,0 point." Yes, that is Rhino. If yours worked then mine would be wrong. When I export a 3D solid from AutoCad it must be wholly in the positive octant or it won't export. It doesn't care where in the octant and no matter where I place it in the CAD software it comes into Cura with the Center of Geometry (the bounding box) at the mid point of the build plate. Since my STL generator prohibits the origin from being within the model then where is the STL generator putting the origin? Why would it care where the origin is? I have no idea how many different STL generators are in use, but I would think that none operate exactly the same as others. When a model comes into Cura WWAAAYYYY over there, I move it. If I need to have two parts in an exact relation to each other on the build plate then I fool Cura so that when I move both to 0,0,0 they are in the correct relationship. If a designer wanted to model the ball on the trailer hitch on my truck - the center of the ball is at about X=6100, Y=0, Z=450. I wouldn't want it to come into Cura at that location. The left and right door handles are not symmetrical. The right door handle would come in at Y=1000 and the left door handle at Y=-1000. My point is that I would still end up moving things around.

Just to make it more complicated - Sketchup is notorious for producing STL files that are poorly constructed and in need of repair. You can upload your STL to service.netfabb.com for repair. It's free but you need to make an account. I only had 1 model that was so bad it couldn't be repaired. Have you downloaded and installed Cura? Because that is the next step to process the STL file into a Gcode file that the printer can understand.

I have Cura 4.6 as my oldest install. I checked on Github and didn't see any obvious references to the speed problem you are having. The newest version is 4.8. Can you update to that and see if the problem still occurs? You can keep your older installations of Cura when you update.

It works for me in 4.6, 4.7, 4.8, and Arachne. I think something else must be going on between your computer and your printer.

I'm constantly dazzled by the Cura Team's (and FieldOfView, SmartAvionics, et al) ability to keep Cura on the cutting edge while juggling the needs of: multiple operating systems, computer platforms, gcode flavors, and printers. As with any extremely complicated system, people (who ALL run operatorware "dumb1") will do something totally unforeseen that will cause it to break. With the installed base of users/debuggers out here, the number of what I would call "valid" bug reports is amazingly low. I decided to make a color change to the dust/rain covers on my cafe bike. Using Arachne and Arc Welder they came out superb. They were very good before (using 4.6 I think), but between the new software capabilities and my own advancing slicing skills, they came out looking "store bought". While doing the above, I think I found a minor bug (duly reported on Github). There is some satisfaction in making a microscopic contribution to the massive effort. The bug had nothing to do with wanting to run Cura on my Cray super-computer running Windows 3.0 with 5 monitors and my home built dual-extruder printer (one is a 2-in-1-out and the other a 3-in-1-out) that has firmware developed by NASA and is connected via parallel port. Should I re-level the bed? Should I wait for the wife to get up before I do that?

It would be helpful to know what printer as well. Those are long retracts and primes. Is it a "3-in-1-out" hot end? That would require "Extruders Share Heater" to be checked. Selecting each extruder in turn, you should be able to set the Printing Temp, Initial Printing Temp, and the Standby Temp for that extruder. I set up a 3in1out machine in Cura. Setting all 9 temperature settings results in a single M104 at the start of the file and "M104 S0 ;turn off hot end" at the end of the file. That is with each extruder assigned to a separate part.

The Cura max Z speed is in the Printer Settings (you may have to load the plugin from the Market Place). That setting should place a limit within Cura. If the person in the first post had it set to 2mm/sec (default is 10) then he can input 120mm/sec, but the Gcode would output 120mm/min. Within the printer, M203 stores the max feedrates. No matter what you enter into Cura as a setting, the max speed is governed by M203. echo:Maximum feedrates (units/s): echo: M203 X500.00 Y500.00 Z25.00 E50.00 The above is a response to M503 on my Ender. You can see that the max Z feedrate is 25mm/sec. If I enter 1000mm/sec in Cura, it's going to be limited to 25mm/sec when the printer sees the Gcode. At the beginning of a gcode file, if you have: M203 X900 Y900 X50 E50 M500 Then you will change the max feed rates of the printer to near their max possible. M500 saves the change to memory (you only need to do it once). Then check the Cura Printer Settings and make all the speed settings there match the M203 settings.

If the extruder is skipping then the E number in the Gcode is being met, but filament isn't getting pushed. The result is under-extrusion. If the under-extrusion was to fall off to say 50%, then yes, the Cura calculation will appear to be way off, but the reality is that it's the print that is off. Whether it's Cura doing the calculation, or you with your calculator, the amount of filament is pretty easy to figure out. With Cura, there might be a discrepancy due to certain settings (like Coasting), but it's going to be pretty close and certainly closer than 412 vs 290. I always clean out my hot end before long prints. It has always been the problem with my Ender and although it's now a Micro-Swiss, I don't trust it to go much beyond 40 hours without a good cleaning and a trim of the bowden tube. With Silky PLA, that's about 20 hours. That stuff is just sticky.

Cura knows the layer height and line width and the total length of extrusions in a model. The filament diameter allows a calculation of the length of filament used. From there the density determines the weight. If the filament diameter or density are inconsistent in "extruder settings" vs "printer settings" or from material to material then I suppose variations could/would occur in the total weight of filament used. The error you're reporting is huge. Your Esteps/mm matters to print quality but I don't think Cura cares. It is noted in the "printer settings". If that isn't present with the other settings you will need to load the plugin from the marketplace.

Open the gcode file in a text editor like Notepad. Go down to "Layer:0" and see what the "Z" is. It should be at your "Initial Layer Height". It should be noted just above or below Layer:0 line. If it isn't your Initial Layer Height, then maybe you have a Z-offset in Cura? Was the model located on the bed or above it? If that Z in the gcode is at the Initial Layer Height, then are your Home Offsets correct? Is the bed in the correct Z relation to the nozzle? Getting the Z to zero is part of the leveling process. If you have an ABL maybe the Z-offset there is incorrect.

Oh joy. Another learning curve. Torgeir, with so much work having been done under the hood, don't you think congratulations should wait until after the test drive? I know there was a lot of effort put into Arachne, but will that translate into performance? Will it be like putting a high lift cam and headers on? Tuning the fuel injection? Re-mapping the ignition curve? Or will it be more like adjusting the seat. Only time will tell sir... only time will tell.

An alternative to ahoeben's suggestion is to try the "rotate" tool option to "lay flat". Sometimes the model comes in at a slight angle (but Lay Flat doesn't always work if the problem is in the model). The top and bottom patterns can be changed. "Concentric" might work well on that. If you use "Lines" you can change the angle in the setting "Top/Bottom Line Directions" by entering something like "[0,90]". The default is 45° and 135° ([45,135]) which is indicated by empty square brackets. If I have to rotate a model on the build plate to get it to fit, I often change the line directions and the infill direction.

My only printer is an Ender 3 Pro. I stay out of the Ultimaker Printer threads because I don't own one and don't know enough about them to try to help with the hardware. It is the same sort of situation with this thread. I have never printed with a .8 nozzle and so I wouldn't know where to begin trying to develop a profile. I also didn't want to give advice that would be (at best) uninformed, and possibly completely wrong. You did not mention your hot end. I do know that the stock Ender 3/5/CR-10 hot end is not very capable. Using a .4 nozzle and line width at .2 layer height means it is filling a rectangle under the nozzle that is .08mm². At a print speed of 50mm/sec that is a flow rate of 4mm³/sec. That is good enough for printing minis and other small items. That is right in the wheelhouse of what Creality was shooting for. Moving to a .8 nozzle/line width at a layer height of .56 means you are filling a rectangle of .448mm². At the same 50mm/sec you are asking the hot end to deliver 22.6mm³/sec. If you do have a stock hot end then that little sucker has to put out 5.6 times as much plastic as its design median. Getting the flow rate down to 4mm³/sec would require a print speed of 8.9mm/sec. That's where the stock hot end/extruder combination would be happiest. Retraction distance? Retraction speed and prime speed? I have no idea. You have put yourself at the pointy end of the stick and are going where few have gone before. I would think you'll need to make things up as you go. It's possible that at .4 layer height you found the upper limit of what your hot end / extruder combination can deliver. I bought the Ender because it was cheap. I replaced the mainboard when it cooked the Estep driver, and then again when it burned the 24v connector. Two fans died prematurely. I had to replace the cheesy plastic extruder with an aftermarket aluminum unit. The layer cooling fan was a joke and so that got replaced. I run a Micro-Swiss hot end as the first TWO stock units didn't get the job done. The stepper motors are fine and the machine is well calibrated and makes good parts. But the tune is on a knife edge. The littlest thing can knock it off tune, and then the trouble shooting starts over again. I think that if I was going to run some sort of production that required a .8 or 1mm nozzle that I would either get a printer configured for 2.85mm filament, or contract the parts to have them injection molded.

When things change due to a models orientation it is often something in the model. If you would upload the model to Service.Netfabb.com it will be processed and most errors get fixed. It's free, pretty good, and fairly fast. Just wait a bit and then download the model. You do need to create an account and sign in. If nothing else it will take "model errors" out of the loop.

When I moved one of the things I decided I no longer needed was a nice hobby air brush kit. I hadn't used it in 30 years so I figured I wouldn't miss it. Wrong again Greg.

Layer shifts are commonly caused by loose belts. You have a lot of belts on that machine. When you say you already tried lowering the print speed - from what number to what number? What speeds did you try? You might also try lowering the Accel and Jerk settings to soften the starts and stops. With Accel at 200, Jerk at 5, it will be slow, but it should provide a starting point and you can increase them until the layer shift shows up and then back them down. As Ahoeben suggests, it's likely a machine problem so the manufacturer is the place to start.

"...would not introduce unwanted features/or rather features that are not designed onto the sliced model." The radius on the far side doesn't extend to the upright like the one on the near side does. There is a jig-jog there. The square "hole" starts at a different layer on one side than it does on the other side of the opening. Above about layer 125 it is no longer a hole, but rather a gap between the two uprights. The reason for that is the radius is not square to the model but rather runs at a slight angle (see the second image). The feature indicated is symmetrical and appears on both ends of the model. If this was a design intent, then it is very subtle and would require finer slicing. I think Cura made an attempt to print this feature but at .12 layer height nad .4 nozzle, it was too fine to describe properly. The second image was with a .2 nozzle at .12 layer height. You can see that there is a double line at the left bottom and the outer line is at an angle as Cura tries to make the feature because it is in the model. The angle is 0.57755886°. This is the .2 gcode file opened in AutoCad. The angle is apparent. There is also a subtle angle in the counterbored pocket in the inside as witnessed by the non-parallel double lines.

You will need to post one of the gcode files that doesn't work and one of the ones that does work. Make the file names something that clearly indicates which is which. The tool to upload a file is below ("choose files"). Do the printers have an Auto Bed Leveling system? Gcode is Gcode no matter what slicer is used but the order in which the gcode is placed as well as certain commands (that a printer doesn't like) can cause problems. Sometimes it's in the "Start-Up" gcode and sometimes it's in the machine definition. Somebody will need to view and compare the gcode files. You have a third party printer with a fourth party mainboard and fifth party firmware. That's a heckuva party.

I did some work designing tool cutter paths for molds. I would scale the part by the shrink factor and then as Geert says, subtract if from a solid block. At the determined split line, cut the mold into two pieces and they become (usually) the top and the bottom. Alignment pins, pry slots, vent holes, fill holes, pockets for inserts, etc., all get added and there it is. A mold. The part might require the addition of draft angles to some areas so it could be removed from the mold. It's worse than useless to make a real nice mold that you can't get the part out of. The locating features you have there may require removeable inserts. After the part hardens the inserts get pulled out and the part can be removed.

I did a bit of a test for "support angle". With a 0.4 line width and 0.2 layer height, the maximum theoretical angle is 63°. There is enough "stickiness" to the filament that 65° printed well. An unsupported 70° surface printed, but drooped as the outside loop was printing over air. It might be acceptable on a gusset but not for any more important feature of a part. Did you try a support blocker for that rectangular feature that is surrounded by the radiused feature? It only needs to be 1mm thick and located correctly to negate those "hanging" supports. That shouldn't affect what is above and your hole should still have support in it.

Well...it was my bad. I had forgotten to remove "Pause At Height" from the list of plugins. The leak was at the "resume" layer and not at the transition between base and spiral. So the only thing the flow test showed was that I should have slowed the print head down (the extruder started dropping steps when the flow went to 110%) and it also reaffirmed that I'm a dummy.

The KISS principle at work. The housing is my own design.

I don't see anything in the response either so maybe the Vref is adjustable on the board. "Could I just and slowdown the Jerk/Accel/Speed of the X axis?" I don't think it's the starts and stops that are the problem. If the diagnosis that the "X" quits working mid-print is correct (and it sounds like it is) then it's going to fail. The stepper will quit if it doesn't receive the signals from the mainboard, or if it has an internal problem. Either the motor is cooling down and coming back to life, or the driver is over-heating, or gr5 was right when he mentioned a bad solder joint or something on the mainboard that is opening up when it warms up. I can't recall ever hearing of a stepper failing on a Creality printer. On the other hand Crealty mainboards are a well documented failure point. I have one sitting on a shelf. If you take the X connector on the mainboard and plug it into the Y socket, and then plug the Y wire into the X socket, and run a print then the failure will point to your problem child. If the X still quits working it's the motor. If the problem moves to the Y it's the mainboard.

Yes, M25 requires an M24 to resume. It never occurred to me that choosing pause (M25) and resume (M24) on the LCD would work. Nice find.