Having problems with Anycubic Kossel Linear Plus calibration

infiniteloop

Probing speed is F120, should not be too fast.

Indeed, that’s rather slow.

+/- 0,16 should be compensated without major problems with bed mesh

That depends. The distance between probing points may ignore some details. I don’t know what your current hight map looks like, but according to your latest test prints, it might be interesting to have a closer look at the two areas where the filament delaminates.

For that, you can define a special grid to probe, just one area at a time, with a very narrow grid spacing. The resulting hight map is like a look through a magnifying glass.

Of course this will not cure your problem, in the best case, you only get a better understanding of what happens.

An alternative explanation for the observed delaminations might be a damaged surface, damaged in the sense that the coating in these areas does no longer stick well with the filament.

SnakeSP

@infiniteloop
The mesh is dense (305 points).
As for the delamination reasons - looks like the nozzle is a bit too low at that points. Babystepping +0.05 solved this, but this again is not permanent solution...

infiniteloop

@SnakeSP

The mesh is dense

Too lazy to do the math, but the resolution is at approx. 10 mm - not enough for a close look. But talking about babystepping: Do I understand you right that this is only true for prints in these two areas, or does this work for the whole print bed?

SnakeSP

@infiniteloop
Babystepping +0.05 works for the whole print. At least it did not make the other parts being printed worse.

infiniteloop

@SnakeSP

Babystepping +0.05 works for the whole print.

Then, you simply subtract 0.05 (mm) from your Z probe height - which is exactly the same as to set the babystepping to +0.05. It might be worth trying some even finer adjustment by subtracting one or two more hundredths. Good luck!

SnakeSP

@infiniteloop
Did that before and after autocalibration with new Z probe height (correct direction - subtracted 0.05, double-checked that) and creating new bed mesh ended with printing mid air... Which is strange.

SnakeSP

@infiniteloop
Did just as you've suggested (Z probe height - 0.05) and surprisingly it did work (i have not changed heightmap.csv this time).
Print comes out with almost acceptable quality:
photo1
The only thing that bothers me is that defects at the start of each print:
photo2 photo3 photo4
it happens after retraction and moving to new coordinates, so i suspect this is due to incorrect retract settings in slicer.
For now i use (bowden tube 640 mm with ID 2 mm):
Retraction distance = 8.0 mm
Extra restart distance = 0.4 mm
Retraction vertical lift = 1 mm
Retraction speed = 100 mm/s
Coasting distance = 0,25 mm (was not able to make Pressure Advance work yet).
Z vertical speed = 40 mm/s, movement speed = 200 mm/s

infiniteloop

@SnakeSP Ah, spaghetti printing! Maybe you should switch to some eatable filament?

infiniteloop

@SnakeSP

Print comes out with almost acceptable quality

First things first: Congrats! One problem solved. What happens when you create a new height map remains to be seen, that’s the next step.

Your photos look familiar to me, I’m in the middle of the same fight with my own printer (just a mean cartesian, bowden setup). So, I can’t provide a solution yet, only two observations:

Print results became better when I reduced the retraction distance (currently at 6 mm).
Your retraction speed looks way too high - I don’t dare to go beyond 30 mm/s.

And yes, pressure advance is a hot topic in the forum, I just try to learn about that from the wise guys over here…

SnakeSP

@infiniteloop

infiniteloop

@SnakeSP Sorry for that…

SnakeSP

@infiniteloop
As for retraction speed - i tried low speeds (30-40 mm/s) and got way too much stringing. But will try to lower down for experiments.
Distance - yes, looks too much, will lower it.

infiniteloop

@SnakeSP

…got way too much stringing

With PETG, in my case it helped to lower the hot end temperature a bit (-5 deg.).

SnakeSP

@infiniteloop Yes, lower temperature reduces stringing, but i'm already at 240 С for the first layer and 220 C for all higher layers. 220 C is the lowest temperature recommended by my filament manufacturer.
So looks the only 3 parameters i can play to reduce stringing are Retraction length, retraction speed and coast at end.

infiniteloop

@SnakeSP

220 C is the lowest temperature recommended by my filament manufacturer.

My PETG is rated at 210-220 C. Worked for my Aero clone the printer came with, but when I mounted an original E3D water-cooled design, I had to go with 235-240 C instead. Everything is relative…

Veti

@SnakeSP said in Having problems with Anycubic Kossel Linear Plus calibration:

For now i use (bowden tube 640 mm with ID 2 mm):
Retraction distance = 8.0 mm

you might want to build flying extruder.

SnakeSP

@Veti That's interesting, but not in nearest future as it is rather complicated.

Danal

I have a printer with about 1.1meter bowden. I run 4 to 5 mm retract at 40.

SnakeSP

@Danal On what plastic? I use PETG now and it is very prone to stringing. What extruder are you using? I'm on stock Anycubic one, which is not very accurate. Extruder modification is next thing to upgrade in my list after making magnetic arms.

dc42

@SnakeSP said in Having problems with Anycubic Kossel Linear Plus calibration:

Measured Z offset at all points from bed.g with G30 S-1 and updated bed.g with H for each point

If you mean that you just used G30 S-1 to see what Z value it was triggering at, that is NOT what the G30 H parameter is for. It is for correcting for variations in trigger height at different points, if your Z probe doesn't have a consistent trigger height. See https://duet3d.dozuki.com/Wiki/Calibrating_a_delta_printer#Section_Adding_trigger_height_corrections_to_the_bed_g_file.