I'm facing weird "pillars" of underextrusion on outer walls of my XYZ test cube.

On the pictures below I`ve printed PLA test cubes with a 0.4 mm nozzle, 0.2 mm height and 210/50 °C hotends/bed temperature.

Gaps are appearing in walls parallel to both X and Y sides. The pictured side is parallel to Y face.

I'm slicing with Cura, my printer is a homebuilt around Anycubic Kossel with Marlin 2.0 onboard.

What have I tried already:
1. Temperature from 190 to 210 °C
2. Retraction from none to 6 mm 60 mm/sec
3. Tuning down Jerk in Marlin from [10,10,0.3] to [5,5,0.3]
4. Tuning down acceleration from 3000 to 1000
5. Tuning print speed from 60 to 30 mm/sec 6. Checking belts, nozzle and extruder.

Now I just ran out of ideas. Delta is calibrated by G33 autotune. Mechanics looks just fine. What am I missing? enter image description here

enter image description here

UPD1: tried removing combing - it did not help. But I noticed that my printer accelerates strangely at this parts

UPD2: could it be stuttering? I have 320 segments per second with block buffer size of 8 bytes(?)

UPD3: lowering segments count to 120 and raising block size to 32

  • 1
    $\begingroup$ Welcome to 3dPrinting.SE! Could you help us out here and orient the print for us? What side exactly are we looking at in the images? Being specific about which direction the print is printed in (ie: X/Y/Z axes) will help in the diagnosis. $\endgroup$ – Pᴀᴜʟsᴛᴇʀ2 Feb 17 '20 at 23:02
  • $\begingroup$ Are you sure it's actually underextrusion? It looks like the bulges in your walls coincide with where the infill lines hit. Is it possible that you have too much play in the delta mechanisms such that these lines overshoot from momentum of the head? $\endgroup$ – R.. GitHub STOP HELPING ICE Feb 18 '20 at 14:43
  • $\begingroup$ Not sure about the extrusion, actually. I've checked the play and the head stands still. $\endgroup$ – Luka Baryshych Feb 18 '20 at 15:06

Localized underextrusion usually means you're oozing material somewhere it wasn't supposed to go. Since you have Marlin 2.0, linear advance might help solve that, and in general right retraction amount is important and turning off combing may be needed (combing over infill allows material to ooze).

Also there are reported bugs in Marlin 2.0 such as https://github.com/MarlinFirmware/Marlin/issues/15473 that might be related, though I would think you'd not be using junction deviation since it's supposedly incompatible with deltas.

  • $\begingroup$ Thanks a lot! I'll try to tune Linear advance and combing. Yes, I'm using classic jerk as it is required for deltas $\endgroup$ – Luka Baryshych Feb 18 '20 at 10:14

Well, I could not track the problem and it lasted till two major updates:

  1. I have changed rods for ones with proper lenght
  2. I have reset all Cura settings to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.