[BUG] PRUSA_MMU2 prevents endstop detection (also for sensorless)

[kursatu]

Did you test the latest bugfix-2.0.x code?

Yes, and the problem still exists.

Bug Description

Build the latest bugfix

• Home X or Y

• Observe that the movement does not stop and t it hits the end. You need to power down the machine.

• Undefine PRUSA_MMU2 , it works as expected.

• This defect happens with both optical sensor and sensorless homing.

• Additionally obstruct the sensor and type M119, observe that they are triggered.

• Home and it tries to home even though the sensor is already triggered.

• Same thing happens with the sensorless homing. You can't stop the movement by holding the head.
  But if you don't define PRUSA_MMU2 both sensor & sensorless homing work fine.

I've initialized the Eeprom/settings all the time while trying to get it to work.
Same issue with sensorless homing only is reported here and in other referenced issues in that defect.
#22012
If I go back just before the commit referenced in this defect then things work.
JohnnyTheOne@dd49902

Bug Timeline

It used to work,. It seems like a regression at JohnnyTheOne@dd49902

Expected behavior

I expected the movement to stop when the endstop is detected.

Actual behavior

No response

Steps to Reproduce

No response

Version of Marlin Firmware

SENDING:M115 FIRMWARE_NAME:Marlin bugfix-2.0.x (Jun 19 2021 16:23:47)

Printer model

Custom CoreXY

Electronics

SKR2
TMC2209

Add-ons

Custom MMU2

Your Slicer

SuperSlicer

Host Software

Pronterface

Additional information & file uploads

Marlin_defect.zip

[thinkyhead]

It sounds like some kind of pin or interrupt conflict. Are you using ENDSTOP_INTERRUPTS_FEATURE at all? Can you enable PINS_DEBUGGING and send M43 to check for any pin / timer conflicts?

Of course, please test the bugfix-2.0.x branch to see where it stands. If the problem has been resolved then we can close this issue. If the issue isn't resolved yet, then we should investigate further.

[thinkyhead]

I'm not sure how #21953 will play into the issue, as that was a pretty major refactor that resulted in a lot of regressions. I suspect there is some kind of plus-one bug somewhere in that pile, but related to the MMU2. I'll review that PR and see where changes related to MMU2 and/or endstops were made, and see if there are any suspicious-looking LOOP or REPEAT constructs.

[Minims]

Hello, I have the same issue and ENDSTOP_INTERRUPTS_FEATUREis not enable on my side.
i have an SKR Pro 1.1 on a CoreXY + MMU2S
I have juste tried with the last bugfix2.0 and the result of M43 is :

M43
         <unused/unknown>                       Input  = 1                       
PIN: PG12        M42 P110         E2_CS_PIN                              Input  = 1
PIN: PG13        M42 P111                                Input  = 1                       
PIN: PG14        M42 P112                                Input  = 1                       
PIN: PG15        M42 P113         E1_CS_PIN                              Input  = 1
PIN: PH0         M42 P3                                  Input  = 0                       
PIN: PH1         M42 P4                                  Input  = 0                       
ok

M119
Reporting endstop status
x_min: open
y_min: open
z_min: open
z_probe: open
filament: TRIGGERED
ok

[thinkyhead]

You both have SKR boards in common. I wonder what that could mean.

The MMU2 uses some serial port defined by MMU2_SERIAL_PORT, and there are pins associated with that serial port, and maybe something got broken in relation to that. So far my scanning of commits specifically related to the MMU2 are not yielding any clues.

Try defining MMU2_DEBUG in your configuration and then there will be extra log output related to the MMU2, and maybe this will give some indication where the MMU2 is interacting with homing.

[thinkyhead]

As another test, even though your MMU2 won't function with this, try changing your MMU2_SERIAL_PORT setting to see if endstops are still broken by the feature. This will at least tell us whether the serial pins are strongly implicated.

[JohnnyTheOne]

It has to do with 5 extruders not mmu2 #22012 (comment)

[Minims]

I have already tried to use anoyher serial port and the endstop issue is the same.

But on a prusa style (not corexy) with skr1.4 turbo-tmc2209 sensorless is working with mmu2.

so it might be related to

• corexy
• 5 extruders

[Minims]

If i disconnect the mmu2, serial says mmu2 not responding so disabled.
but sensorless is still not working