KISSlicer 1.5 Beta 2 - in development

joestefano
Posts: 57
Joined: 14 Nov 2014, 17:53

Re: KISSlicer 1.5 Beta 2 - in development

Post by joestefano »

Windows7 64bit works fine
Attachments
Map.png
inventabuild
Posts: 271
Joined: 09 Nov 2014, 23:03

Re: KISSlicer 1.5 Beta 2 - in development

Post by inventabuild »

Jonathan,

Thanks for allowing narrower crowning paths because they can be useful in some situations even if they are not ideal in every situation. For the most part the KISSlicer v3.8 update cleaned up the gaps I was getting in my text after v2.2. However in the below example one of the crowning paths in the letter "N" is missing in the first layer, but it appears in the second layer. Is this by design or can you make the crowing path appear in both layers?
First Layer - Missing 1 Crown Path In Letter N
First Layer - Missing 1 Crown Path In Letter N
Second Layer - Crown Paths Fill All Letters Including Letter N
Second Layer - Crown Paths Fill All Letters Including Letter N
The paths were created w/ 0.4 mm extrusion width and 5 loops. I attached the files for your reference.
Attachments
SNAPS Box Top - Small Version.zip
(34.42 KiB) Downloaded 193 times
SNAPS Box Top Text - Small Version.zip
(30.52 KiB) Downloaded 207 times
User avatar
lonesock
Posts: 258
Joined: 09 Nov 2014, 18:41
Contact:

Re: KISSlicer 1.5 Beta 2 - in development

Post by lonesock »

Regarding the extrusion time map, I'm guessing that G-code comments are disabled?

Regarding the crowning on the bottom layer, that was definitely not intentional.

I will look into both of these, thanks!

Jonathan
frozen.rabbit
Posts: 132
Joined: 17 Nov 2014, 09:20

Re: KISSlicer 1.5 Beta 2 - in development

Post by frozen.rabbit »

Yes, it does display with comments enabled.
626Pilot
Posts: 33
Joined: 28 Jan 2015, 22:24

Re: KISSlicer 1.5 Beta 2 - in development

Post by 626Pilot »

I did some experimental work with a post-processor that implements coasting: https://github.com/626Pilot/KISSCoast

The current version will stop extrusion near the end of a path. You can specify a different coast length for walls/short walls (which is highly necessary, as they are over-extruded). The results are better than I've seen before. There is much less stringing and pockmarking now.

However, it's not perfect. After a de-string, when it primes, a bunch of extra plastic is deposited. This is not usually a problem on long loop paths. However, for shorter paths, it causes a huge mess! Consider these two effects:
- Dispensing water from a water cooler, and observing air gurgling up into the water jug
- Knocking a glass ketchup bottle on the neck, and observing how easily the ketchup flows after air gets into the neck

I think something similar may be happening here. The filament starvation eventually allows voids to form inside the nozzle, reducing internal pressure and allowing more plastic to simply drop out from the force of gravity. If the air voids percolate up into the nozzle, it could easily explain the deposition of extra plastic.

I'm working on taking the retraction command in the destring block, commenting it out, figuring out the speed ratio necessary to get the same retraction to happen at the end of the toolpath some specified distance before it ends, and then inserting the correct E commands on the final segment(s) of the toolpath.

I know from another thread that you were looking at adding pre-retraction, so perhaps my research can save you some time. After KISSlicer is perfect at pre-retraction, or whatever solution you use, my code can stay on GitHub to serve as a working example for others who want to experiment with post-processing.

Also, I'm REALLY glad you're back. Messing with those other slicers nearly sent me into conniptions. KISSlicer has them all beat, hands down.
User avatar
lonesock
Posts: 258
Joined: 09 Nov 2014, 18:41
Contact:

Re: KISSlicer 1.5 Beta 2 - in development

Post by lonesock »

Hi, everybody!

There is a 1.5 beta 3.9 on Dropbox (Windows builds so far). Not many changes:
  • fixed time map when G-code comments were disabled
  • increased the seam gap limit to 4 (a request to help fix blobbing, will probably go back to 2 once I get coast working)
  • improved the model tab rendering code
  • improved OpenGL view limits for really large printer beds
@frozen.rabbit: The time map should now be available even if the G-code comments are disabled.

@inventabuild: Just playing with my internal thresholds didn't fix that crowning issue, and I don't want to change any core mechanisms right now before the 1.5 release, as I'm afraid of breaking other stuff. I'd like to postpone working on this until after the release. (Btw, did you know that you don't need to cut the text out of the box top? You can just assign the extruder with the text a higher priority and KISSlicer handles the overlap for you.)

@626Pilot: Thanks for working on this! Currently I only have a really old RapMan, and it uses the strange RPM-based extruder control, with only one head and builtin (and non-replaceable) retract / prime, so I have never been able to work on this myself. The good news is that in a couple of months I will have access to a nice new printer, multi-head and direct extruder control (5D-style firmware). I look forward to being able to test this myself!

There also seems to be a problem with the 64-bit version of the Mac build (if you grabbed 3.7 try forcing 32-bit mode), so I will be working on that, but it also means the Mac build will take longer to get out.

The good news is, other than the Mac problem I think this is almost ready for release! If you have found any show-stopper bugs, please chime in and let me know soon! Thanks!!

Jonathan
frozen.rabbit
Posts: 132
Joined: 17 Nov 2014, 09:20

Re: KISSlicer 1.5 Beta 2 - in development

Post by frozen.rabbit »

time map works.
626Pilot
Posts: 33
Joined: 28 Jan 2015, 22:24

Re: KISSlicer 1.5 Beta 2 - in development

Post by 626Pilot »

lonesock wrote:@626Pilot: Thanks for working on this! Currently I only have a really old RapMan, and it uses the strange RPM-based extruder control, with only one head and builtin (and non-replaceable) retract / prime, so I have never been able to work on this myself. The good news is that in a couple of months I will have access to a nice new printer, multi-head and direct extruder control (5D-style firmware). I look forward to being able to test this myself!
I hope it's a Bowden machine. The hysteresis in the tube (compressibility of filament + friction from tube curvature) makes it necessary to use more aggressive retract settings, so you might well try coasting/pre-retraction on a machine where the extruder is right on top of the hot end and see no difference at all. A lot of Bowden-based 3D printer manufacturers are putting the drive motor at the bottom of the printer, and then routing the Bowden tube to the hot end in a big loop. The increased length and curvature both make retraction worse. Confounding this problem, many of us use E3D hot ends, which - while marvelous in several ways - unfortunately have short heat breaks, meaning that long retracts encourage jamming (especially with PLA).

If you don't have access to a machine with a Bowden tube, I can put out some feelers to see if we can get someone to ship you one.
User avatar
Davide Ardizzoia
Posts: 122
Joined: 07 Nov 2014, 13:05
Location: Oleggio, Italy
Contact:

Re: KISSlicer 1.5 Beta 2 - in development

Post by Davide Ardizzoia »

Chances are that Lonesock will have soon a machine with several bowden tubes to play on...
:-)

Davide Ardizzoia
inventabuild
Posts: 271
Joined: 09 Nov 2014, 23:03

Re: KISSlicer 1.5 Beta 2 - in development

Post by inventabuild »

Jonathan, congrats on your soon to have new printer. Davide's printers look awesome so whichever model it is I'm sure it will be great!!!

How is seam gap limit measured? Is it an absolute value or a % of something?
Post Reply