Lin Advance setup

Hello, I am looking into linear advance and I am just interested in figuring out what information is relevant here in late 2019.

1. To enable linear advance we much change "USE_ADVANCE" to 1 and it is no longer recommended to use "ENABLE_QUADRATIC_ADVANCE" (1, leave at 0), is this correct?

2. Possibly a stupid question, but I've seen K referred to as quadratic. Is it bound/related to the "quadratic advance" shown in the question above? If so, going under the assumption quadratic advance is not recommended. Is the 

3. The way to calibrate linear advance is to use this calibrator, download the g-code and use 'find & replace' to change all "M900 K" with "M233 X" or "M233 Y" 
  -Sub question, the X and Y in M233 has nothing to do with the axis X and Y, am i correct?

4. Marlin apparently has deprecated the L value and now seem exclusively use the K value, thus removing a lot of the now irrelevant information about calibrating the L value linked from this forum. Is it still recommended running the calibrator with the L value or is the K value a better choice?

5. Is repetier linear advance comparable to Marlin's V1 or V1.5? Which does one pick in the calibrator (point 2)? Is "M233 Y" still relevant or is it "M233 X"

6. I have seen @Repetier several times mentioning that an L value of 40 is a good starting point for direct drive setups. Is this in combination with a calibrated K value or is it a good starting point for running the calibrator (for example running M233 Y, test with the values 40-80 or something)


  • Yes, quadratic is not good to use. So set that part 0. M233 Ylinvalue is used to set it.
    Typical range for Y on direct extruder is 30 - 150 depending on free length.
    Use 1.0.4dev for best result with linear advance. There was recently some improvement I think it is only in dev tree.
    I guess it is more V1 marlin kompatible also that seems to only affect the factor anyway, not how it is handled.

  • Ah, Thank you. So to summarize

    1. Correct. USE_ADVANCE 1 and QUADRATIC 0

    2. Seems correct. Use L not K. M233 Y not X

    3. Use calibrator in 1.0 mode not 1.5, value typically between 30 -150. Replace M900 K with M233 Y

    4. Using L instead of K still recommended for Repetier-fw

    5. Linear advance V1.0, M233 Y, Use repetier 1.0.4dev for best results.

    6. 40 May be good, direct drive value typically between 30-150.

    Thanks a lot. :) Really helped clear out a lot of the confusion I had since a lot of info was from early 2018 and I was unsure of how/if things had changed.
  • Yes, summarization is correct.
  • Hello 

    i was looking around , where can i find a tutorial for repetier linear adv ?
  • There is none. Important thing is use 1.0.4dev version for it as it has some fixes regarding advance. Makes only sense for direct extruder and value lies between 40-120 for these depending on length nozzle-drive and temperature and material. It corrects the flow when decelerating/accelerating to be more equal so edges and infills look better and seams are less prominent. But we do not have a tutorial on our own, also I guess some will exist on internet.
  • Why does it only make sense for direct extruders? What's wrong with bowden here?
  • The amount you have to advance depends on filament length between nozzle and the drive. Direct extruders have maybe 5-7cm and that is already hard to correct without retraction. Bowden has 50-70cm or more so it needs to correct 10 times more in the same time. In fact it would need even to retract filament very quickly and that just does not work well. Plus with bowden you normally print faster which is why you use bowden - but increased acceleration also needs faster retraction in addition. So nothing wrong with bowden just physics get to limits here.
Sign In or Register to comment.