Slic3r vs Kisslicer comparison on new printer

[MorbidSlowBurn]

After adjusting my mechanical set-up based on geneb reccomendations, little to no friction in ujoints and arms. I printed the calibration cube using Slic3r. It still had some gaps on the last layer. So I tried Kisslicer using mhackney's setting he posted in his build blog. Only change I made to the config was 3 loops, .2 layers and a "G28" home command on start. Attached you will find the comparison images. The left part is Slic3r the right is Kisslicer.

The only issues I saw on the Kisslicer print was some bridging issues (expected as it did a 45 path not a straight path across the gaps). Also when printing in Kisslicer I did notice that when it was extruding there was less "popping/crackling" coming from the hot end, almost non-existent. All prints have been ABS at 230C. The Kisslicer print was at a slightly slower speed than the Slicer print.

I also want to add a horizontal move at the end before moving to home as the printer left a long string off the part.

[img]http://i1299.photobucket.com/albums/ag6 ... 2ad147.jpg[/img]

[img]http://i1299.photobucket.com/albums/ag6 ... a55a0b.jpg[/img]

[img]http://i1299.photobucket.com/albums/ag6 ... 8882fe.jpg[/img]

[tom10122]

KISS looks great! I am still trying to get it to work with my 4 axis board

[MorbidSlowBurn]

I was impressed with Kisslicer. The feel of the vertical side is smoother and as I said most settings were the same and no mechanical adjustments between the prints. It was printing slower which may have helped but the last layer is what I really wanted to compare. Still tweaking my printer and wanted to see if it was hardware or software.

Right now I am about 25% into a Yoda print. I think it may fail because I forgot to update the firmware with the latest fix. I am running polygonhell's firmware in my Rostock.

Right now I am happy with my quality. If I want more detail I may switch to a smaller nozzle.

[Polygonhell]

KissSlicer by default prints perimeters outside to in which results in the cleaner outside.
The week point for KissSlicer is IMO bridging, it doesn't identify them and it fills bridges the same as it fills everything else. I resort to support for anything that would require a significant bridge.
But generally the difference between the prints you show is typical IMO.

[MorbidSlowBurn]

In my limited experience I will agree that bridging is an issue and supports are required.

However, mine did print the inside perimeters first before the outside of the part. Confirmed it just now with my current print. I think the smoothness is a combination of me gaining experience and a slower feed rate. Up until the Kisslicer print I heard crackling/popping coming from the extruder regularly. Didn't know what it was but I suspect it was that I was printing a little too quick and was getting small blockages in my hot end. There were minuscule gaps in the printed thread. Now no crackling/popping and no gaps.

I do like the tool path better in Kisslicer as it appears to finish an area before crossing a gap. Just more logical. Again limited experience. Just sharing for anyone else new to this like me.

[mhackney]

@MorbidSlowBurn, were there a lot of retracts in the Slic3r print? I hear that popping right after rapid retracts.

I also discovered that the beta KISS does not do the outside-in perimeters by default. I had to uncheck that little box.

[MorbidSlowBurn]

Mhackney- yes there were a lot of retracts. And thinking back I was aggressive with my retract setting it probably was "sucking air" which could cause the pops.

My major trial between the two was to see if the different slicer would fix my last layer internal gaps and it appeared to do that. As you mentioned in my tuning topic-scientific method approach. I just didn't get all the settings the same between the programs. With more tweaks I may be able to get Slic3r to do the same quality but being new I am trying to find the least "painful" method of getting good prints.

[mhackney]

Yes, I've correlated that popping with retracts. Haven't heard them since using the wipe feature on KISS and minimizing/eliminating retracts.

I am not sure if there is a least painful method to get good prints! That is the Holy Grail of filament extrusion printers. The RepRap forums are filled with posts about it. The most workable and least painful approach seems to be a systematic "bottoms up" system. Once you have the basic mechanical and electrical demons excised THEN you start on calibration:

1) extruder adjustment to get the filament moving - in both directions - with enough force to extrude and not slip but not so much so as to distort the filament. This assumes some reasonable setting for hot end temperature. I always start with 225°C for ABS and 180°C for PLA but I am talking actual temperature measured with a thermocouple at the nozzle/heat block junction. Once I know what the delta is I use that offset as a fudge factor for temperature settings in my slicer and host. Advanced Note: I also check the temperature by inserting the thermocouple into the hot end all the way (without filament of course) and ramp the temp up in my host and watch the thermocouple. I have a spreadsheet with the offsets for my Rostock. The next step is to adjust the thermistor table to tune it to my setup but I have not done that yet.

2) print the single wall cube object to work out the bugs in extrusion, filament starving, etc. Don't worry about globing or threads until the basic extrusion calibration and that it is working properly. You also need to measure your filament diameter to use in the slicer and work out other parameters. I also fine tune the temperature to get best results for the specific filament this way (each roll is different) as I posted somewhere else.

3) then work on infill with my "Z calibration thing". This actually attacks two areas; it makes sure the mechanical delta arm assembly has good free movement and it allows you to work on the settings to get the best infill without over- or under- fill.

4) once you have the basics above dialed in, then you move on to eliminating blobbing and stringing. These can be greatly affected by the items above and fall into the camp of the overall "systems" tuning whereas the first 3 items focus on individual parameter tuning.

Then, you go to sleep, wake up the next day and nothing works! The humidity has gone up, it's a full moon, whatever. It is through previous experience that you'll recognize the issue and know how to compensate.

regards,
Michael

[MorbidSlowBurn]

Excellent post. This is what is some of the intent I am looking for in my tuning guide thread. Basically get everyone using the same calibration methods so we can help each other out. I have what I believe the mechanicals/electrical end down. Now it is just those little tweaks. Need to double check to see if I am getting filament distortion.

I like your thermocouple method, but I don't have access to them. Wondering how accurate an electronic kitchen thermometer probe is? I do have a non contact ir reader that I used on RC nitro engines but it is tough to gauge what you are measuring. Also the emisivity of the surface plays a large role in how accurate the reading is.

Are your calibration objects available on thingiverse?

[mhackney]

Kitchen probes are pretty accurate. A non contact reader is not going to give you good results on this.

The Z calibration object is available in my build thread. I need to redo it to post on Thingiverse - which is the plan. The "issue" is that after printing the thing about 100 times I realize you only need to print 3 layers! So, simply follow my instructions to scale it in your slicer to end up with 3 slices. You probably would do this anyway to take into account your layer height.