3

I'm new to 3D printing, but my printer supports Linear Advance. I heard that it offers improvements in print quality. I used Marlin Linear Advance Pattern Generator to generate a print with horizontal lines at a variety of k-values.

Which K-Value would be best from my below image?

enter image description here

geekgeek4
  • 103
  • 7
  • 2
    0.65-ish. Though Linear Advance on a Bowden printer may impact print speeds due to the high K value. – towe Feb 04 '22 at 06:36
  • 1
    @towe: Because the Ender 3's extruder is ungeared, it can handle wild speed and acceleration needed to make the impact on print speeds "not too bad". Just make sure you have non-zero E-jerk (maybe 5?) and E speed and accel limits set to something like 150-200 and 5000-10000, respectively. If you have them set slow, everything will become ridiculously slow as soon as you turn on LA. – R.. GitHub STOP HELPING ICE Feb 04 '22 at 20:32
  • Just for experiment try [this K-factor tower generation tool](https://github.com/ArtificalSUN/Kcalibrator/). It might give you a better insight. – dma_k Oct 11 '22 at 21:04

2 Answers2

5

As a general answer to evaluate the effectiveness of the K-factor, when the K-factor Calibration Pattern generator output print is inconclusive (probably not in this case), printing a tower at various K-factor values might give you more insight, e.g. like:

enter image description here

To vary the K-factor with height, a similar procedure as in How does one use a heat tower? can be followed to insert a new K-factor with G-code M900.

0scar
  • 32,029
  • 10
  • 59
  • 135
4

Go with the sort of tower 0scar suggested and look for corner bulge/rounding. It's a lot more useful than the single-layer test pattern from the Marlin site, which I've found can be misleading.

With that said, just looking at your test print, I would go with the lowest K factor that gives acceptable-looking results, so around 0.5 or so. Overshooting is more likely to harm your print quality than undershooting, and will limit your print speed too. When I used the original bowden on my Ender 3, my calibrated K factor for PLA was 0.6, so I think this is in the range of what's expected.