GregValiant

I printed on my Ender 3 Pro with the 1.1.5 8 bit mainboard and Marlin 1.1.8. I've been thinking on this and I MIGHT (big might!) have an explanation. Gcode is gcode and the printer just goes where it's told. The toolpaths within the gcode are different, but that really shouldn't cause a quality dropoff like you are getting on the 5.1 print. 5.1 has one difference that maybe your printer doesn't like, and it is in the gcode. The default Wall Order in 5.1 is "Outside to Inside" and equates to the 4.x setting "Print outside walls first". I'm not sure why Ultimaker made the new default that way, but my Ender is not fond of it. I've made the default "Inside to Outside" like 4.x was. Outside to Inside might be OK for some things, but overhangs certainly hate it as there may be nothing for "Outside first" to connect to when the outer wall goes down.

I don't know how far along you are with Cura. Turn on the Custom settings. Next to the Settings Search box is an icon with three lines. It's the Settings Visibility button. Set the visibility to "All". Click on it again and collapse the settings as it makes it easier to find things. In the Support section will be "Support Structure" and you can choose Normal or Tree. That's about it for types. On the left side is a toolbar with the move, scale, rotate tools and the 6th button down is Support Blockers. They can be configured to modify where they overlap models, block support, or print as support or print as a model. They can be scaled and moved like any other mesh. I downloaded the chess set and took a look at the X wing rook. The pods on the ends of the wings are barely attached to the wings. That would be really tough to print. Every model has needed repair as well. Here is the other knight. You can see that the axle caps on the back aren't actually connected to the main model. They will fail for sure. Those models have issues. Regarding supports, in the Material section of the settings I've been playing around with lowering the Support Flow and the Support Interface Flow. I've had some good luck running them both at 90% flow rate. It makes for weaker supports that are easier to remove.

I'm guessing it's the way it was modeled. What CAD software did you use? The horseshoe shape appears to have been placed resting on the surface rather than being extruded from the surface. There may be an infinitesimal gap between the base and the horseshoe. There should be no skin between the horseshoe and the lower base. That area where the infill is exposed doesn't look right either. That infill area should not go to the slot. On that layer the only place infill should be exposed is directly beneath the horseshoe. I think you might have an inverted surface there. Here is a similar model. The infill should be a continuation up from the base and into the interior of the horseshoe feature. The top layer of the base should look something like this...

I looked at the extrusion rates using MSExcel. They look fine but it's hard to tell with 5.x because the line widths change and I can't account for that with my macro. So I printed them. One of these is your 4.13 file and the other is your 5.1 file. I made no changes and didn't tune anything during the prints. I'm sitting here staring at them under a magnifier and if I hadn't marked which is which I could never tell them apart. So what does that tell us? I have to let Cura off the hook on this. The 5.1 on the left (or is it the right!!) has no ringing while the 4.13 print (which is actually on the right) has very minor ringing, That is the only difference I can see between the two prints. The startup gcode of each file is the same as well. There just doesn't seem to be anything about the two gcode files that should cause your printer to hiccup on one and not the other.

On the Cura menu bar go to "Settings / Printer / Manage Printers" and then "Machine Settings". Check the box for "Gcode Flavor" and if you have Marlin make sure it is set to Marlin and NOT "Marlin(volumetric)". Click on the Extruder tab and check the "Compatible Material Diameter" and make sure it is set to whatever your printer uses (1.75mm or 2.85mm).

Those "X Wing" rooks look tough. Stringing aside - proper support is a must. In regards to the stringing, make sure you have your "Retraction Minimum Travel" and "Max Comb Distance with No Retract" lower than the shortest distance you need a retract on. If there is a move of 15mm and you tell Cura you only want a retraction if the nozzle travels 20mm then yes, there will be a string. Here is my effort.

How is your layer cooling fan working? I ask because the tops of the windows of the second deckhouse are really sagging and things like the radar dish on the roof don't look too good either. I thought I had posted these but it was late and I must have forgotten to hit the submit button. Neither of these have the "saggy" windows that you are getting. I have a single 5015 bolted to the side of my Ender 3 Pro. These are with some old PLA and it could have stood some drying out. PrusaSlicer Cura

You know I hate having to explain... Oreo cookies are the best in this arm of the galaxy. Here is the proper way to eat an Oreo. Oatmeal cookies are second best (IF they have nuts and raisins). Animal Crackers may be called crackers but they are actually cookies. And welcome back.

Animal? Oreos? Oatmeal?

Hello. As far as I know nothing changed in Pause at Height. I use it a lot but I don't use the retract or prime functions. Post that gcode and someone will take a look. If you read the tooltip for "Extrude Amount" you will see it is for UM2's. In most cases the retractions and primes are taken care of in the regular gcode and there is no need for further retractions and primes. There can be a prime move by the "regular" gcode after the pause. Whether there is or not is dependent on whether there was a retraction in the regular gcode before the pause.

That's it. That part won't print anywhere near as faceted as the first model.

#12808 and #12956 are the Github bug reports I was referring to. They aren't the only ones.

I would say just go with "Fine" and export that model you printed and then change it to "Coarse" and export the same model again. That's what I did for that comparison view. I used MS 3D Builder but you can open the STL files in Cura and put them side-by-side. In the "Prepare" screen you should see that the Fine model isn't nearly as faceted as the Coarse model. Curves and cylinders really show the difference.

Hello @gr5 and @princefemo01 There are bug reports on GitHub regarding this behavior. Something is broken with "Concentric" as both regular Infill and as Support Infill. It looks like the developers are addressing the problem for the next release but for now, it's a problem with certain combinations. One combination is that if "Infill Layer Thickness" is different than "Layer Height" things go wonky. It might be operating system specific as well. On my Windows system if I set the Infill Layer Thickness to 2X layer height the result is 0% infill in the preview and it isn't generated in the gcode even though I have my infill density set to 50%.

In the AutoCAD STL export utility the setting is called "Facet Resolution". The model on the left was exported with the FACETRES at 0.25 and the model on the right had the FACETRES set to 10.00. The model on the right would print as a "smoother" model as the sides of the triangles that comprise the chords of the diameter are much shorter. A high resolution STL gives a high resolution print. The gcode files can be much larger and in the case of these two - the low res gcode was 4.5mb and the high res gcode was 9.7mb. The print time was about the same.

@Cuq - He's running Klipper and an SKR mainboard. @kc6331 - I get the same results when I slice the model. The gcode contains (1) M104 line at the beginning immediately followed by M109 and those set the hot end temperature. At the end of the file there is a single M104 line to turn the hot end off. Octoprint can be configured to send some commands on it's own. I wonder if your Klipper install is doing something similar. What's happening to your temperature is not the result of any command contained within the gcode.

By "Low resolution" I meant the STL file. Low resolution results in larger facets on curves and cylinders but a smaller file size. Higher resolution results in more triangles, smaller facets, but larger file size. It's a function of the export utility from CAD.

Hello @TT_Vert and @Smithy. Sometimes it's hard to figure out what Cura is doing with the supports. That particular model is kind of low resolution as witnessed by the coarse triangles that comprise it. I think that is part of it. It appears that the triangles of the top and bottom overhanging features line up the same but the triangles of the middle feature don't line up with those of the top and bottom. So Cura is seeing a line going across that is sticking out past the points (vertexes) of the middle feature and Cura wants to support it because that's what the math says to do. It looks like it's supporting air because certain settings (min XY distance for instance) is keeping the support away from where Cura has designed it to go. Another problem with that part is the trapped supports that go behind the middle "T" feature. It will be tough getting those suckers out. Unfortunately, they are required because the vertical portion prints first and several layers later the connector to the main body prints. Talking about this would likely just get more confusing. Post that model file here and I'll take a stab at it. I can tell you now though, one of my suggestions might be to make those "T" sections separate details that would glue into place. As Smithy alluded in regards to getting rid of supports...Just because we can print difficult models doesn't mean we have to.

"...it has to be an issue with Cura." That could well be. Cura is not without bugs. I'm just saying that the particular gcode you posted is not the problem. The temperature gets set to 215 at the beginning and is turned off at the end. That's all that is in there. It appears you are running a direct drive hot end (retraction 0.8mm) and your StartUp and Ending Gcode aren't from a Creality definition file. Did you change the mainboard of your printer? I ask because I see you are running gcode written for RepRap firmware and a stock CR-10 would have Marlin installed.

That is a gcode file. Please look at my previous post and provide a Project File. In lines 14 and 15 of that gcode file the hot end gets set to 215°. There are no other temperature lines in the file until it gets to "M104 S0" and shuts off the hot end at line #660817. So the Gcode is not affecting the printers hot end temperature.

I don't know how the new parts mounted. It sounds like the centerline of the nozzle moved further towards the front of the printer. If that is true then the only thing that should be affected is the Home Offset of the printer. The printer will put the 0,0,0 origin of a Gcode file at the Home Offset 0,0,0. That point has to be over the bed and not somewhere off the build plate. If the parts made the print head bigger, then you might want to change the Print Head settings. Use a ruler and measure from the center of the nozzle to the left most part of the print head (MinX) and that will be a negative number in the settings. Measure from the nozzle to the right-most part of the print head (MaxX) and that will be a positive number. In front of the nozzle is negative Y and the behind the nozzle is positive Y.

Can you post one of the problem models? It would be even better if you could load one of those models, set Cura up the way you like, and use the "File | Save Project" command and then post the 3mf file here. I've moved to 5.1 exclusively and I haven't had any issues like that. I do tend to print more mundane mechanical things and they don't generally have spires or minarets but they often have several islands that the nozzle has to move back and forth in order to print them.

I was going to ask you for a project file next. That isn't to say I would have found that but from 2 to 6 is a jump. There was a poster here who was wondering why the prints were so slow. After going back and forth with a couple of comments he produced a project file. 7mm retraction at 1mm/sec. So his prints were spending 14 seconds on every retraction. No wonder they were slow. Congrats on figuring out your problem.

It looks like Layer 38 (Cura Preview) is the last full loop around the hole and Layer 39 is the first horseshoe shaped path. Could it just look like a hesitation there? It's where the head has to stop and move back around to the other side of that opening that develops.

Recv: ok T:210 /212 B:60 /60 T0:50 /0 T1:210 /212 @:0 B@:0 That looks like the printer response to M105. My guess is that Octoprint is sending "M105 T0 T1" and the "ok" part of the response indicates to Octoprint that the port is still open. I don't see anything in the gcode that could be causing the problem. Usually when something like this happens there is an M104 or M109 (M133 for MakerBot) line that sets an extruder temperature to the Standby temperature which is too cool to print at. That doesn't happen in your gcode. The print starts with T1 and the only mention of T0 is to disable the fan at the start of Layer:0 and enable it at the start of Layer:1. I save the ends of filament rolls for situations like this where something is occurring and I can't figure it out and it's driving me nuts. I changed my Home Offsets to make my Ender an Origin at Center printer and swapped out your starting gcode for mine. The rest of this print is your gcode. There were no extrusion stoppages but there was some funkiness on the backside. I don't know what to make of that but the overhang was close to needing support. After a couple of layers (which did extrude correctly) the print finished fine. The only thing I'm sure about here is: If that is a window ice scraper then you have a really tiny car.