Smooth linear advance breaks linear advance calibration

[ThomasToka]

Did you test the latest bugfix-2.1.x code?

Yes, and the problem still exists.

Bug Description

As many know i maintain a Marlin fork on latest Marlin bugfix-2.1.

I noticed that smooth linear advance breaks any configuration possibilities and with this the prinquality is not like it is without it.

Smooth linear advance enabled:

Smooth linear advance disabled:

Bug Timeline

enabling Smooth Linear Advance

Expected behavior

• visible pattern change on K change

Actual behavior

• no change

Steps to Reproduce

Simple to reproduce:

• do not use FTM as i dont use it.
• enable smooth linear advance with or without e sync, K0.035 U 0.020 (or whatever you have)
• print the smooth.gcode Pattern attached (it has U with 0.020 added when k i not 0)
• make pic
• disable smooth linear advance, keep linear advance enabled
• print the classic.gcode attached
• make pic

Version of Marlin Firmware

bugfix-2.1.x

Printer model

Ender 3 S1 Pro

Electronics

STM32F4

LCD/Controller

Dwin

Other add-ons

many ;)

Bed Leveling

None

Your Slicer

Prusa Slicer

Host Software

None

Don't forget to include

• A ZIP file containing your Configuration.h and Configuration_adv.h.

Additional information & file uploads

pressue_advance_orca.zip

[Nuck-TH]

bugfix-2.1.x

just for information of reviewers - it isn't mainline bugfix, but bugfix automegered into tomas's fork.

[ellensp]

If you cannot reproduce this on stock bugfix, we cannot help you. We cannot maintain your fork for you.

Please test on unmodified Marlin Bugfix

[ThomasToka]

bugfix-2.1.x

just for information of reviewers - it isn't mainline bugfix, but bugfix automegered into tomas's fork.

sorry stop spreading fud. you should know better with you experience if you would have looked in it.

"it isn't mainline bugfix, but _bugfix automegered into tomas's fork" -> haha

[ThomasToka]

please do the simple test: print one pattern with SLA one without. whatever settings you use. simple as that.

for the non believers i i have put up the squashed fork on top of todays bugfix-2.1.x

https://github.com/ThomasToka/Marlin/tree/MARLIN-E3S1PROFORK-BYTT-RELEASE-CODE-v036_squashed_bugfix_21x

There is not one change in the Smooth LA implementation done by me.

Its the easiest to say its the fork. I removed the smooth la for now and am fine.

Others will report in future if it was not me.

[ThomasToka]

just compiled the default Ender 3 S1 Pro configs from upstream with todays virgin bugfix-2.1.x.

configs attached. Pictures will follow. first print was with sla enabled. same result. will now print it without sla and then post again pictures.

Configurations.zip

[ThomasToka]

Upstream same result.

with sla:

without sla:

[dbuezas]

In this pr #28076 of your authorship, this change sets k to zero if smoothLA is active.

const float advK = ftm_active
          ? TERN0(FTM_HAS_LIN_ADVANCE, ftMotion.cfg.linearAdvK)
          : TERN0(HAS_ROUGH_LIN_ADVANCE, extruder_advance_K[E_INDEX_N(extruder)]);

[ThomasToka]

Great you found it. But this wrong code was not contributed by me.

I contributed working code and it was demaged by someone and not reviewed at all by the maintainer of smooth linear advance:

const float advK = TERN(FTM_HAS_LIN_ADVANCE, (ftMotion.cfg.active ? ftMotion.cfg.linearAdvK : extruder_advance_K[E_INDEX_N(extruder)]), extruder_advance_K[E_INDEX_N(extruder)]);

[ThomasToka]

After a deeper dig in the variables used here my one liner is not perfect.

Solution to cover all possible situations as a oneliner:

const float advK =
  TERN0(FTM_HAS_LIN_ADVANCE, (ftMotion.cfg.active && ftMotion.cfg.linearAdvEna))
    ? TERN0(FTM_HAS_LIN_ADVANCE, ftMotion.cfg.linearAdvK)
    : TERN0(LIN_ADVANCE, extruder_advance_K[E_INDEX_N(extruder)]);

or like proposed in the PR with the bool.