KISSlicer Overhang-Bug (All Versions + Stable Release)

Beefi
Posts: 14
Joined: 07 Feb 2015, 07:00

KISSlicer Overhang-Bug (All Versions + Stable Release)

Post by Beefi »

Hi,

there is a bug in the Overhang-Feature:

For example, if your Perimeter has a flow rate of 2.0 and the "Cool-Parameter" (for overhangs) has a flow rate of 1.0 then
the flowrate is never going down to 1.0 in the overhang-area.
At the beginning of the overhang-area the flow goes down and up (results in a blob) an the overhang-area prints at a normal Perimeter-flow (not 1.0).
I analyzed this with the gcode and some graphical gcode-viewer.

Only if you set the Cool-Parameter to the same flow-rate as the Perimeter-flow, then there is no stop and go and no blobs (and no overhang-feature).
If you set the Cool-Parameter to another value as the Perimeter (for example 2.1 oder 1.9), there is a stop and go and also NO Overhang-Feature,
because the Cool-Flow will not be used!

This Bug also exists in stable Version 1.6.3.

Sorry for the bad english :D

I attached some Fotos and the STL-File (in the ZIP):
Attachments
Benchy-Cut Testfile.zip
Benchy cut Testfile
(710.73 KiB) Downloaded 51 times
Cool-Flow <> Perimeter-Flow
Cool-Flow <> Perimeter-Flow
Cool-Flow = Perimeter-Flow
Cool-Flow = Perimeter-Flow
Screenshot
Screenshot
Screenshot.png (82.08 KiB) Viewed 2416 times
User avatar
pjr
Posts: 692
Joined: 05 May 2015, 10:27
Location: Kamnik, Slovenia

Re: KISSlicer Overhang-Bug (All Versions + Stable Release)

Post by pjr »

Hi

Can you please also upload the project file?

Thanks

Peter
Beefi
Posts: 14
Joined: 07 Feb 2015, 07:00

Re: KISSlicer Overhang-Bug (All Versions + Stable Release)

Post by Beefi »

Here is a Project-File and a Gcode-File (in the ZIP).
You can search in the Gcode for "overhang" here you see that the Cool-Flow is never used.
Attachments
Kisslicer.zip
Project and Gcode
(113.74 KiB) Downloaded 55 times
User avatar
pjr
Posts: 692
Joined: 05 May 2015, 10:27
Location: Kamnik, Slovenia

Re: KISSlicer Overhang-Bug (All Versions + Stable Release)

Post by pjr »

Thanks. Looking at it now...

Peter
User avatar
pjr
Posts: 692
Joined: 05 May 2015, 10:27
Location: Kamnik, Slovenia

Re: KISSlicer Overhang-Bug (All Versions + Stable Release)

Post by pjr »

There's a problem with your settings in the G-code :(
Capture.JPG
Capture.JPG (66.31 KiB) Viewed 2370 times
You are limiting the flow rate to 1mm^3 /sec and then slowing to the same vlalue for overhangs. So there should not be much of a speed change, and I see only a very small variation.

Also, when viewing the G-code file, Craftware doesn't show any deformation where you are getting it:
Capture1.JPG
Capture1.JPG (49.36 KiB) Viewed 2370 times
Also, please remember that the speed decrease (and fan increase) for overhangs does depend on the amount of support from the previous layer.

I will keep looking.

Peter
Beefi
Posts: 14
Joined: 07 Feb 2015, 07:00

Re: KISSlicer Overhang-Bug (All Versions + Stable Release)

Post by Beefi »

Hi pjr,

ok thanks, that could be the reason. This printer has a Dual-Extruder...when I'm going to print with ONE Extruder then I load this "LEER-Profile". (Leer in german means empty).
With this "Empty-Profile" I want to avoid to heat the second extruder.
Empty-Profile to print with ONE Extruder
Empty-Profile to print with ONE Extruder
Ext Map Settings.PNG (43.52 KiB) Viewed 2366 times

Then I have to take another solution for this...is it ok to switch the printer profile from 2 to 1 extruders if I like to print with one extruder? Or does this solution has other sideeffects?
Individual switching between 2 or 1 extruders?
Individual switching between 2 or 1 extruders?
Printer Settings.PNG (42.11 KiB) Viewed 2366 times

You have to check the stop and go with another gcode-viewer. With a gcode viewer which is showing speeds and not only coordinates.
Like this one http://gcode.ws/.
Here is a screenshot...there you can sie the blob graphically:
Gcode-Analyzer.PNG
Beefi
Posts: 14
Joined: 07 Feb 2015, 07:00

Re: KISSlicer Overhang-Bug (All Versions + Stable Release)

Post by Beefi »

I made a testprint...absolutely no change in code and in the result.
Also with the choice of only one extruder there is a stop and go on the begin and the end of an overhang area.
Also, please remember that the speed decrease (and fan increase) for overhangs does depend on the amount of support from the previous layer.
I hope you understand what I want so say:
It's okay when the printhead slowdown into an overhang area and speeds up after an overhang area. But Kisslicer slows a little bit (not to the cool-flow parameter) down and speeds up again before an overhang area (and after).
Even tough the Fan can fully cool the filament or the overhang angle allows to only slow down a little bit...cause of the pressure in the nozzle, the blob will be set with this behave.
User avatar
pjr
Posts: 692
Joined: 05 May 2015, 10:27
Location: Kamnik, Slovenia

Re: KISSlicer Overhang-Bug (All Versions + Stable Release)

Post by pjr »

OK, so layer 7 at 1.85mm.

Here is the perimeter path:

Code: Select all

; 'Perimeter Path', 0.7 [feed mm/s], 20.0 [head mm/s]
G1 X160.329 Y45.397 E8.15652 F12000
; 'Destring Prime'
G1 E8.55652 F2400
; head speed 20.000000, filament speed 0.672792, preload 0.000000
; unsupported segments 3
; Unsupported degree: 0.233921
; Overhang(42.1)
; Unsupported Fan = 76 (from 75)
G1 X160.391 Y45.391 E8.55862 F1200
G1 X160.792 Y45.354 E8.57216
G1 X161.564 Y45.283 E8.59824
G1 X162.707 Y45.177 E8.63687
G1 X163.918 Y45.128 E8.67763
G1 X165.163 Y45.15 E8.71953
G1 X166.373 Y45.242 E8.76035
G1 X167.412 Y45.376 E8.79558
G1 X168.483 Y45.569 E8.83218 F1189
G1 X169.548 Y45.817 E8.86897 F1178
G1 X170.111 Y45.98 E8.88868 F1166.9
G1 X170.449 Y46.077 E8.90051 F1155.9
G1 X170.561 Y46.109 E8.90445 F1144.9
; fan %*255
M106 S194
G1 X170.688 Y46.153 E8.90895 F1133.9
G1 X171.067 Y46.284 E8.92244 F1122.9
G1 X171.698 Y46.503 E8.94492 F1111.8
G1 X172.202 Y46.71 E8.96325 F1100.8
G1 X172.801 Y46.955 E8.985 F1089.8
G1 X173.865 Y47.462 E9.02468 F1078.8
G1 X174.884 Y48.02 E9.06376
G1 X175.947 Y48.685 E9.10594
G1 X176.92 Y49.373 E9.14601
G1 X177.845 Y50.108 E9.18575
G1 X178.757 Y50.923 E9.22691
G1 X179.619 Y51.79 E9.26803
G1 X180.431 Y52.706 E9.30922
G1 X181.163 Y53.635 E9.34899
G1 X181.873 Y54.651 E9.39069
G1 X182.509 Y55.681 E9.43142
G1 X182.632 Y55.984 E9.44242
G1 X182.664 Y56.264 E9.45188
G1 X182.625 Y56.545 E9.46145
G1 X182.5 Y56.841 E9.47225
G1 X181.836 Y57.911 E9.5146
G1 X181.123 Y58.925 E9.5563
G1 X180.388 Y59.85 E9.59605
G1 X179.573 Y60.764 E9.63723
G1 X178.708 Y61.628 E9.67836
G1 X177.756 Y62.472 E9.72117
G1 X176.827 Y63.202 E9.76092
G1 X175.85 Y63.885 E9.80099
G1 X174.826 Y64.519 E9.84154
G1 X173.795 Y65.079 E9.88099
G1 X172.715 Y65.588 E9.92116
G1 X172.129 Y65.826 E9.94242 F1089.8
G1 X171.636 Y66.026 E9.96031 F1100.8
G1 X171.029 Y66.234 E9.98192 F1111.8
G1 X170.665 Y66.36 E9.99488 F1122.9
G1 X170.543 Y66.402 E9.99921 F1133.9
; fan %*255
M106 S191
G1 X170.424 Y66.436 E10.00337 F1144.9
G1 X170.066 Y66.538 E10.01588 F1155.9
G1 X169.471 Y66.708 E10.03673 F1166.9
G1 X168.467 Y66.94 E10.07137 F1178
G1 X167.393 Y67.132 E10.10809 F1189
G1 X166.317 Y67.269 E10.14458 F1200
G1 X165.145 Y67.356 E10.1841
G1 X163.934 Y67.378 E10.22483
G1 X162.69 Y67.327 E10.26673
G1 X159.924 Y67.071 E10.36016
G1 X156.293 Y66.731 E10.48286
G1 X152.667 Y66.396 E10.60535
G1 X148.998 Y66.059 E10.72929
G1 X145.304 Y65.722 E10.85409
G1 X141.432 Y65.37 E10.98488
G1 X137.796 Y65.037 E11.10768
G1 X137.487 Y64.979 E11.11828
G1 X137.188 Y64.858 E11.12911
G1 X136.91 Y64.662 E11.14055
G1 X136.708 Y64.441 E11.15062
G1 X136.552 Y64.18 E11.16085
G1 X136.445 Y63.87 E11.17189
G1 X136.397 Y63.523 E11.18368
G1 X136.395 Y59.44 E11.32101
G1 X136.395 Y49.063 E11.67009
G1 X136.433 Y48.691 E11.68268
G1 X136.532 Y48.374 E11.69386
G1 X136.707 Y48.063 E11.70584
G1 X136.98 Y47.782 E11.71902
G1 X137.291 Y47.597 E11.73121
G1 X137.608 Y47.495 E11.74241
G1 X138.266 Y47.425 E11.76469
G1 X141.759 Y47.106 E11.88266
G1 X145.247 Y46.787 E12.00049
G1 X148.948 Y46.45 E12.12551
G1 X152.542 Y46.122 E12.24692
G1 X154.418 Y45.949 E12.31028
G1 X156.035 Y45.799 E12.36492
G1 X157.365 Y45.675 E12.40984
G1 X158.449 Y45.574 E12.44646
G1 X159.287 Y45.495 E12.47478
G1 X159.811 Y45.446 E12.49247
G1 X160.153 Y45.414 E12.50405
G1 X160.315 Y45.398 E12.50951
G1 X160.329 Y45.397 E12.50999
; Blending Fan to Cool (76.6% blend)
; Blended Fan at 79%
; fan %*255
M106 S201
There is a slight slow-down, but it is very slight and there are no pauses. Unless the fan speed changes are causing your printer to pause.

I wonder whether it would be possible for you to edit the G-code and print the result:

1. Remove the fan speed changes
2. Remove the extruder feed changes at the pointy end
3. Remove both fan and extruder feed changes

As for your profiles - dual vs single - it would depend how your printer G-code is set up. I think that selecting an "Empty" material profile is probably your best option.

Peter
Beefi
Posts: 14
Joined: 07 Feb 2015, 07:00

Re: KISSlicer Overhang-Bug (All Versions + Stable Release)

Post by Beefi »

Hi pjr,

I generated a gcode-file with KISSlicer and removed ALL M106-commands in the file...then no blobs are occuring.
So the problem is definitely at my printer/firmware.
When I don't remove the M106-commands, then you can notice a very very very short break at the M106-command-position (at the beginning of the overhang).


But...
I googled some benchy boats and there are more boats sliced with KISSlicer which has the same problem. So I believe that this behave could be normal or in other words, many printers have this very little break.
I guess that most peoples don't care about these blobs or they don't notice it. Did you print the STL from my first post on your printer? Do you have this blobs, too?

Is there the possibility to deactivate this fan control at overhangs to avoid fan changes during a perimeter-print?
It would be nice to force overhangs to the parameter "Cool-Flow" and let the fan force to the perimeter-fan-setting...originally I expected this behave from this parameters.

Thank you very much for the support!
User avatar
pjr
Posts: 692
Joined: 05 May 2015, 10:27
Location: Kamnik, Slovenia

Re: KISSlicer Overhang-Bug (All Versions + Stable Release)

Post by pjr »

OK, I don't know what your printer is, but I don't get this issue when printing with an "original" Prusa (Mk2/Mk) They run PR Marlin on a 2560 MCU.

With this, I know that the fan commands are implemented immediately upon receipt of the command and are not buffered, so any blobs would occur up to 10 moves before the actual position in the G-code.

Just wondering whether setting the Material "Cool" value to 60 (the same as the loops) would prevent your problem?

Peter
Post Reply