10 minutes ago, maht said:Hey there,
Have you experimented with the 'Z seam' feature in custom mode? This allows you to manipulate where the Z seam is placed.
Randomizing the Z seam may be a solution in this case - this will place the seam in a random position for each layer.
That's not a bad idea!
This didn't do much to help it tho 😞
Got a 3mf file you're willing to share? or a thingiverse/youmagine/other-online-repository link?
6 minutes ago, maht said:Got a 3mf file you're willing to share? or a thingiverse/youmagine/other-online-repository link?
I have pm:d the file!
Just had a quick try myself. By turning off 'Spiralize outer contour' and randomizing the seam I got (imo) a slightly better result. 3mf will hit your inbox shortly.
The geometry here is a difficult one to make a z seam disappear from completely as each layer is a circle, so there's no corners to hide the seams in.
The alternative is to post-process the model, e.g by creating a seam at a constant point so that it can be removed with a knife/sanded/etc. But given you are using translucent material I'm not sure how feasible that is.
That's actually a pretty good idea to slice it with 0% infill, but i don't know if that's gonna be an improvement with all that solid infill along the wall. Probably gonna be pretty visible once printed. Ideally what i need is just one line all the way up.
I was thinking of using the Surface Mode: Surface to to get it seamless, this removes the bottom. But what i could do is first print the bottom 0.4 mm high, add a 0,4mm or so z offset in the rest of the model, and print it like that.
Theoretically giving best of both worlds, isn't that what Surface Mode: Both is supposed to do?
One obvious question is: why is there a visible seam in "spiralize" mode at all?
Shouldn't it be a continuous extrusion?
1
19 minutes ago, tinkergnome said:One obvious question is: why is there a visible seam in "spiralize" mode at all?
Shouldn't it be a continuous extrusion?
As i understand it, all vase mode print have seams, its's just that slicers are usually really good at hiding them. This model being particularly hard with circular layers, with no corners to hide a seam. I guess Surface Mode: Surface could be interpreterad as a raw no compromise vase mode, however this removes the bottom layers, which i need.
Just now, ManlishPotato said:As i understand it, all vase mode print have seams, its's just that slicers are usually really good at hiding them. This model being particularly hard with circular layers, with no corners to hide a seam. I guess Surface Mode: Surface could be interpreterad as a raw no compromise vase mode, however this removes the bottom layers, which i need.
But i should say i'm no expert in this area.
No, there should be no z-seam in spiral mode. This is clearly a bug that still haven't been fixed.
This was posted in January and there hasn't been a follow-up. I wonder how he/she made out. Cura has been updated twice since then. Sometimes I find it helpful to just think on stuff for my own knowledge...and I did.
+1 on No Z-Seam. The Z value changes incrementally and constantly as the nozzle goes around the shape. Since it is a single wall construct, there is no sideways step either (i.e.: to a second line of wall). There really shouldn't be a Z-seam. Another way to look at it is there is only 1 layer. I've printed a 3 piece 650mm sailboat hull. There was no z-seam with Cura 4.6.1.
Looking at the Gcode - within every 360° movement of the nozzle around the part there were 140 discreet steps in the Z and at that point it was 1 layer height (.20mm in my case) higher. A .0014mm step every 2.5° isn't really noticeable.
This is also 4.6.1 from today and nothing I do remove this seam. It is spiraling up, but it still makes a seam without any reason.
I wrote a bit of Visual Basic code I use to read gcode files into AutoCad. I brought in one of the boat files and it shows a helix all the way up. It's pretty close to a perfect helix too. No jig-jags at all. Looking at the 3d print from that boat file under a microscope I don't see anything that could be taken for a seam. I was pushing a lot of plastic to make a thicker hull but that shouldn't make any difference. A seam line is a seam line.
I'd like to stare at that Gcode file for a bit. Can you put it up here?
Kegle1.stlKegle1 11-05-2020.gcode Here are the .stl and the .gcode.
Hi Minimalfort,
As I have an UM2 ext, I'd just uploaded the gcode file to Cura ver. 4.6.1. And sure I could see the seem..
Then I loaded into Cura your stl file, and sliced it with my "spiralizer" setting but cannot see any seem there.
Here is the file:
UM2E_Kegle1_Cura_4_6_1_May_12-20_T.gcode
Thanks
Torgeir
Hi Folks,
I've been using the spiralizer mode since Cura introduced it in their slicer, over time there have been some variants.
First, if you like to analyze the gcode file, you need to use the file that's saved to be printed and not using the preview function after slicing your stc file!
So, it is important to select the additional file types when installing Cura, normally you would have the default, but you will need to select 3mf and gcode as a minimum IMO.
The most important setting I found to be set is:
Combing to off.
Z-seam alignment to shortest.
Seam corner reference to none.
Select single wall.
Top and Bottom layers to 0.
Layer height typically 0.1 mm (or extra fine 0.06 mm)
Analyze your gcode file, -that's saved for printing. In here you will see the final result of your extruder, how it move and an indication of amount of filament that's used during travel.
This last point is important, cause it will give you an idea of how you printed object will look like.
I.E. In spiralizer mode, the extruder shall move steady, without any stop, or stop -then a little back before starting move steady.
The "h" mark is often caused by this little "back" movement.
I'm using the gcode viewer every time I'll print something new, I do not have make a test print if I'm using the filament brand I'm used to. When checking printing in spiralized mod, I'll move the line bar all the way down, then starting play. Those very first lines to be printed will indicate if this is a true spiralize mode.
This is a great tool in Cura.
Here is tree pictures from the gcode viewer at this critical start point (Using Cura 4.6.1):
Here is the start of the gradually climb in order to do spiraling without stop.
Just before next layer.
Another one and so on...
There is some easy way to improve the h stepping, this is to select from 8 bit to 16 bit stepping.
But this is perhaps a bit beyond this thread.
Good luck.
Thanks
Torgeir.
I like this kind of stuff though. Keeps the brain loose.
I read your gcode file into AutoCad and here it is
You can just see the magenta line running up the cone. They are short dashes (G0 moves) in the XY plane. There is no Z movement in any of them.
Here is a front view of the same model with the G0 layer isolated. You can see the G0 moves going up the cone.
Here is the one I sliced from your stl file.
and the front view of the same one.
In the one I sliced - you can see that all the G0 moves are at the bottom in the mesh. Nothing is running up the cone. There is no seam in the one that I sliced (Cura 4.6.1)
If you take a look at the statistic text in the images, you can see that the one I sliced has twice as many lines in it as the one you sliced. It's almost as if I was slicing a different model.
I haven't had time to mull this over yet. Something is going on though. But I was right when I said no Z seam. What you have there is an XY seam.
Hi again Folks,
I've forgot to show how Minimalfort's own file looks in Cura, -also sure agree with GregValiant that this is an X/Y issue..
Still using the best Cura version ever IMO (thanks to Ultimaker and all involved people in this development) Ver:4.6.1
Here is the picture:
This "little" correction movement is what's makes the "famous Z seam".
Thanks
Torgeir
Edited by Torgeir
All the short lines in the seam are about .15mm long. The longish ones are about 1.3mm long. I don't understand why the one I sliced didn't do it. The same version of Cura should produce the same results. The total movement inboard is .134mm for each jog and is consistent up the model. Another thing is that the Gcode file you provided is 2.2MB and the file I generated is 4.1MB. It is attached.
EDIT: Here are some lines of code from MinimalFort's gcode file:
G0 F9000 X42.581 Y47.395
G0 X42.527 Y47.627
;TIME_ELAPSED:397.392266
;LAYER:5
;TYPE:WALL-OUTER
G1 F1200 X44.104 Y46.187 Z2.002 E550.2832
And here is the same area from the one I created:
G1 X108.211 Y53.26 Z1.199 E3897.18335
G1 X110.361 Y53.145 Z1.2 E3897.25663
;TIME_ELAPSED:2464.219053
;LAYER:5
M106 S191.3
;TYPE:WALL-OUTER
;MESH:Kegle1.stl
G1 F2237.6 X112.554 Y53.106 Z1.201 E3897.33128
As you can see, there are no G0 lines in the one I created and in fact there are none anywhere in the entire file.
When I have AutoCad read in a gcode file the color of the line is determined by words like "Mesh" or "Wall" or "Infill". The Gcode file of MinimalFort came in as red which is Outside Wall. The one I did came in as yellow which is Mesh. (The M106 line is because I had the fan set to step up speed by height.)
Another thing is the feedrate. In the first file it is consistent, layer by layer, at F1200. In the second file it constantly varies by layer going from a max of F2237.6 at layer 5 down to F600 at layer 498.
All in all it doesn't seem possible that these were sliced with same software with near identical settings.
On another topic Tongeir - How do you get that fancy view in Cura. Mine looks flat.
Edited by GregValiantHi GregValiant,
I've been looking at your "cut cone" when using the 3MF file of it with your setup.
Well, -the only difference here is the printer used..
During the time since Cura geared up and introduced this "new" format into their slicer, things changed as this format have much more information about your object and your printer.
For this reason I'm using a few versions of Cura, even one of the oldest version to compare with.
There are previous Cura versions that's makes 3MF files, which makes newer versions of Cura crash if you try loading these "old" 3MF files.
This 3MF format is quite new (5-6 yr or something like that), so I'll "guess" there is some developing going on all the time..
For this reason I've tried and compared all the version there is and found my favourites.
So I'm using Cura version: 15.04.5, 2.3.1, 2.5.0, 3.6.0 and 4.6.1! Well I'm also using burtoogle's version of Cura @burtoogle.
I found that the different versions may slice object differently depending of shape and form etc.
Almost forgot your last question, I'll just zoom in from the outside of the model and using the gcode viewer (using the gcode file saved ready to print). There is a few thing you cannot see if you look in the preview mode right after slicing.
Thanks
Torgeir
I hope we are not talking apples and oranges here. Minimalfort sliced with 4.6.1 and got the seam. You sliced it and there was no seam. I sliced it with 4.6.1 and there was no seam. The Torgeir and Greg gcode files are extremely similar. Minimalfort's gcode file is very different.
It must have been a settings thing. Time to move on.
Ok Greg,
I thought this file was yours, -well that's life.
ArunC posted a topic in UltiMaker Cura,
.thumb.jpeg.0b7a05eafc09add17b8338efde5852e9.jpeg)
Dustin posted a topic in Firmware,
Recommended Posts
Guest maht
Hey there,
Have you experimented with the 'Z seam' feature in custom mode? This allows you to manipulate where the Z seam is placed.
Randomizing the Z seam may be a solution in this case - this will place the seam in a random position for each layer.
Link to post
Share on other sites