Klipper: pressure advance vs linear advance

[leungtech]

In marlin 1.5, there is a function call linear advance, in klipper, there is pressure advance. Are they the same thing?? In marlin, linear advance will not retract, does pressure advance retract??
http://marlinfw.org/docs/features/lin_advance.html
Thx!!

[theopensourcerer]

There is documentation:

https://github.com/KevinOConnor/klipper/blob/master/docs/Pressure_Advance.md

[wizhippo]

Any chance we could modify or make use of something like http://marlinfw.org/tools/lin_advance/k-factor.html to tune pressure advance instead of printing the square such as replacing M900 K0.4 ; set K-factor with SET_PRESSURE_ADVANCE ADVANCE=0.4.

[dragonnn]

@wizhippo why not... just open the g-code in any test editor and replace M900 K with SET_PRESSURE_ADVANCE ADVANCE=
This isn't rocket science.

[wizhippo]

@dragonnn I'm going to try it, I was just hoping for constructive input/thoughts in regards to if the marlin testing method is just as good or if testing with a corner like the docs show would be better. Perhaps one is more suited then the other for how they are implemented.

[KevinOConnor]

@wizhippo - when I first started thinking about pressure advance, I ran tests on my printer using a mechanism that seems similar to the marlin link above ( https://github.com/KevinOConnor/reprap-test/blob/master/extruder-spring/test-extruder-spring.py ). Unfortunately, I found that method has two big drawbacks:
1 - minor changes in bed leveling have a significant impact on the results. If the bed is just a little off, the test isn't useful.
2 - it's difficult to compare the results of two separate runs, thus making fine tuning harder.

In contrast, I've found the simple square test to be a very robust. It takes a little longer to run, but the results are easy to interpret and very reliable. The square could probably be a bit shorter to make it print faster, though.

Separately, the pressure advance documentation could use some updates. I recently got my delta printer (with a long bowden tube) working well with a very high pressure advance (0.7). (It's possible to get good results with a high pressure_advance if one lowers max_accel.) Unfortunately, the ramps board on my delta melted ( https://imgur.com/a/FhHxV ) before I could finish my work on that.

[wizhippo]

Just for fun I ran it. Bit hard to see as my bed is a mirror but it looks like high values work best. My acceleration on the print is 2000mm/s. You can see it here https://drive.google.com/open?id=0BxUx95JrlxN3bndGdXlVelRxczd3NVNvM1lpUjd5M2w2UTVF

Is it worth using such high values?

[Alexdad76]

@KevinOConnor

Unfortunately, the ramps board on my delta melted ( https://imgur.com/a/FhHxV ) before I could finish my work on that.

Always solder the power wires! This is a GRP (Good RAMPS Practice :-).
https://3deshnik.ru/wp-content/uploads/2015/10/IMG_5418.jpg
https://3deshnik.ru/wp-content/uploads/2015/10/IMG_5413.jpg
I already know several cases of ignition of printers and even apartments ...
Be more careful!
We need you alive and healthy!

[wizhippo]

@KevinOConnor Thanks for pointing out that you used a high value. After doing the test print I mentioned above 0.7 was looking good for me too. I've since done multiple prints and have much better quality and less stringing. I have reduced my retract distance from 6mm to 1mm.

[KevinOConnor]

Great. I'm closing this topic though, as the original request (description of pressure advance) is already answered in the documentation.