Notifiche
Cancella tutti

Linear Advance Calibration Cube  

  RSS
christopher.d3
(@christopher-d3)
Estimable Member
Linear Advance Calibration Cube

Which calibration cube have people been using for Linear Advance tests? I’d like to come as close as I can to replication their results. Thanks.

Postato : 24/11/2017 10:34 pm
Sebastian650
(@sebastian650)
Active Member
Re: Linear Advance Calibration Cube

Calibration cubes are an outdated way to determine the K value. During my implementation of LIN_ADVANCE for Prusa i3, I created a single line zig-zag pattern, which allows for much faster result comparison and nearly no material waste. Luckily, some smart people created a pattern generator and since some days, the Marlin docs are updated:

http://marlinfw.org/docs/features/lin_advance.html

See the explaination of the pattern there, and also the link to the gcode generator.
As the change in result with different K is gradual, the easiest way is not to search the perfect line, but to look for the line with lowest K where you can see imperfections on speed change and then the line with the highest K, where imperfections start to become visible again. The best K value is in the middle.

Postato : 25/11/2017 8:21 am
christopher.d3
(@christopher-d3)
Estimable Member
Topic starter answered:
Re: Linear Advance Calibration Cube

Thanks, I’ll chase that down. One thing that confuses me - Josef’s blog article gives typical values of K in the 10s, except for MK2MM which has a value of 200. Is that difference because the MK2S is a direct extruder and the MK2MM has Bowden drives? Which would mean all the low values people toss about are irrelevant to me? Thanks again.

Postato : 25/11/2017 8:51 am
Sebastian650
(@sebastian650)
Active Member
Re: Linear Advance Calibration Cube

Yes that's true. K is nothing else than a spring constant. If you have a spring made out of one specific material (PLA, in most of our cases), you have to compress it more to get x Newton of pushing force in the nozzle if it's longer. It might be not linear due to other effects (friction etc.) but think about doubling K when doubling the free filament length.

As a result, the best thing to get perfect prints is to use a extruder where the hobbed drive is as close to the nozzle as possible, and to use as stiff filaments as possible 😉 Therefore I don't like bowden systems..

Postato : 25/11/2017 9:16 am
christopher.d3
(@christopher-d3)
Estimable Member
Topic starter answered:
Re: Linear Advance Calibration Cube

Thanks. Good to know. Another Bowden vs. direct drive trade-off. For now, at least, I’ll stay with the MultiMaterial version of the MK2S just to make my life as complicated as possible. 🙂

Postato : 25/11/2017 7:49 pm
Condividi: