[BUG] Sensorless homing with SKR 1.3/TMC2209 failed #17057

gsxr2001 · 2020-03-03T09:31:16Z

Bug Description

Marlin 2.0.4.4 on SKR 1.3 with TMC2209 UART mode, sensorless home for X and Y, 3DTouch for Z. QUICK_HOME enabled

If I home axes individually, all work perfectly. If I home all (just G28) it goes just for 2 or 3mm. Then goes to opposide side (see video)

Got back to 2.0.3 all work again

My Configurations

Configuration.zip

Steps to Reproduce

Go home with menù or G28 command

Expected behavior: [What you expect to happen]

It goes home

Actual behavior: [What actually happens]

It goes to opposide side

Additional Information

https://youtu.be/rguBzUyfv5Y

gsxr2001 · 2020-03-03T15:47:36Z

Disabling QUICK_HOME it works fine

drewzh · 2020-03-03T18:59:52Z

Same experience here. Glad I'm not the only one - I've been diffing files trying to find some configuration schema change but came up with nothing. Disabling QUICK_HOME in the meantime. Thanks for the heads up @gsxr2001

b00gie-dev · 2020-03-04T10:30:39Z

Also have the exactly same issue. I will try to figure out what the problem is. I have an backup of the firmware (~2 weeks) which doesn't cause this problem, unfortunately only the compiled bin file. But I guess I will find the revision by file-date.

cropduster · 2020-03-07T08:29:14Z

I also have issues with sensorless homing on my SKR v1.2. It sometimes does not jome correctly after a ptint has finished andd another print is about to start. This seems to be random. In one case it didn't move any axis and hence 0 for x was where the extruder just happened to be and the print started in the right corner of the bed.

onek24 · 2020-03-07T18:06:40Z

Same issue with SKR mini e3. With SENSORLESS_HOMING and QUICK_HOME enabled.
Occurs with Marlin 2.0.4.4
Last working version is Marlin 2.0.4.3
Disabling QUICK_HOME solves this for me.

Seems to be related to this change:
#16872

brettvitaz · 2020-03-07T19:23:23Z

wow...

Disabling QUICK_HOME does not work for me.

Good thing I keep the working copy around...

thinkyhead · 2020-03-08T04:32:27Z

Occurs with Marlin 2.0.4.4
Last working version is Marlin 2.0.4.3

Ah, well we can just run a comparison on these then.

thinkyhead · 2020-03-08T05:04:35Z

There are only three changes between 2.0.4.3 and 2.0.4.4 that seem likely to affect sensorless homing. (I have not looked at changes in the same time period in the TMCStepper library, but let's assume it's fine for now.)

These lines were added to G28.cpp (lines 109-110). Try removing them:

do_blocking_move_to_xy(-0.5 * x_axis_home_dir, -0.5 * home_dir(Y_AXIS), fr_mm_s / 2);
safe_delay(100);

A bug fix changed your MINIMUM_STEPPER_PULSE to 0 if you have Trinamics. The (assumed wrong) previous value was 2. Try setting MINIMUM_STEPPER_PULSE to 2 or 1 and see if it has any effect.
A bug fix changed your MAXIMUM_STEPPER_RATE to 5000000 if you have Trinamics. The (assumed wrong) previous value was 250000. Try setting MAXIMUM_STEPPER_RATE to some lower value like 2500000 … 1000000 … 500000 … or 250000 and see if this has any effect.

Thanks for helping to improve Marlin and to wrangle these interesting new hardware combinations!

onek24 · 2020-03-08T17:47:03Z

There are only three changes between 2.0.4.3 and 2.0.4.4 that seem likely to affect sensorless homing. (I have not looked at changes in the same time period in the TMCStepper library, but let's assume it's fine for now.)
1. These lines were added to `G28.cpp` (lines 109-110). Try removing them:
   ```c++
   do_blocking_move_to_xy(-0.5 * x_axis_home_dir, -0.5 * home_dir(Y_AXIS), fr_mm_s / 2);
   safe_delay(100);
   ```

2. A bug fix changed your `MINIMUM_STEPPER_PULSE` to 0 if you have Trinamics. The (assumed wrong) previous value was 2. Try setting `MINIMUM_STEPPER_PULSE` to 2 or 1 and see if it has any effect.

3. A bug fix changed your `MAXIMUM_STEPPER_RATE` to 5000000 if you have Trinamics. The (assumed wrong) previous value was 250000. Try setting `MAXIMUM_STEPPER_RATE` to some lower value like 2500000 … 1000000 … 500000 … or 250000 and see if this has any effect.
Thanks for helping to improve Marlin and to wrangle these interesting new hardware combinations!

Commenting out lines 109-110 in G28.cpp definitely solves this issue for me.
Therefore it seems like this merge #16872 broke it.

gsxr2001 · 2020-03-08T19:30:20Z

There are only three changes between 2.0.4.3 and 2.0.4.4 that seem likely to affect sensorless homing. (I have not looked at changes in the same time period in the TMCStepper library, but let's assume it's fine for now.)
1. These lines were added to `G28.cpp` (lines 109-110). Try removing them:
   ```c++
   do_blocking_move_to_xy(-0.5 * x_axis_home_dir, -0.5 * home_dir(Y_AXIS), fr_mm_s / 2);
   safe_delay(100);
A bug fix changed your MINIMUM_STEPPER_PULSE to 0 if you have Trinamics. The (assumed wrong) previous value was 2. Try setting MINIMUM_STEPPER_PULSE to 2 or 1 and see if it has any effect.

A bug fix changed your MAXIMUM_STEPPER_RATE to 5000000 if you have Trinamics. The (assumed wrong) previous value was 250000. Try setting MAXIMUM_STEPPER_RATE to some lower value like 2500000 … 1000000 … 500000 … or 250000 and see if this has any effect.
Thanks for helping to improve Marlin and to wrangle these interesting new hardware combinations!
Uncommenting lines 109-110 in G28.cpp definitely solves this issue for me.
Therefore it seems like this merge #16872 broke it.

What are 109-110 lines? Because in Marlin 2.0.4.4 at that lines I don't have uncommentable lines

onek24 · 2020-03-08T19:45:34Z

There are only three changes between 2.0.4.3 and 2.0.4.4 that seem likely to affect sensorless homing. (I have not looked at changes in the same time period in the TMCStepper library, but let's assume it's fine for now.)
1. These lines were added to `G28.cpp` (lines 109-110). Try removing them:
   ```c++
   do_blocking_move_to_xy(-0.5 * x_axis_home_dir, -0.5 * home_dir(Y_AXIS), fr_mm_s / 2);
   safe_delay(100);
A bug fix changed your MINIMUM_STEPPER_PULSE to 0 if you have Trinamics. The (assumed wrong) previous value was 2. Try setting MINIMUM_STEPPER_PULSE to 2 or 1 and see if it has any effect.

A bug fix changed your MAXIMUM_STEPPER_RATE to 5000000 if you have Trinamics. The (assumed wrong) previous value was 250000. Try setting MAXIMUM_STEPPER_RATE to some lower value like 2500000 … 1000000 … 500000 … or 250000 and see if this has any effect.
Thanks for helping to improve Marlin and to wrangle these interesting new hardware combinations!
Uncommenting lines 109-110 in G28.cpp definitely solves this issue for me.
Therefore it seems like this merge #16872 broke it.
What are 109-110 lines? Because in Marlin 2.0.4.4 at that lines I don't have uncommentable lines

I'm sorry, i wrote that wrong. I mean "commenting out lines 109 - 110".
The blocking move in line 109 moves the X and Y axis too far back.

adamrb · 2020-03-10T04:28:33Z

Also seeing this on an SKR 1.4 with TMC 2209's.

Removing these lines fixed it for me as well: https://github.com/MarlinFirmware/Marlin/pull/16872/files

gsxr2001 · 2020-03-11T18:09:58Z

The problem has not been resolved! Not completely. Commenting on the 2 lines, there is no longer the problem of moving in the opposite direction, but the quick home still does not work! At this point it is better to disable QUICK_HOME. So the problem is not solved.

onek24 · 2020-03-11T18:13:22Z

The problem has not been resolved! Not completely. Commenting on the 2 lines, there is no longer the problem of moving in the opposite direction, but the quick home still does not work! At this point it is better to disable QUICK_HOME. So the problem is not solved.

How does the quick home not work?

By the way, the issue here is that the motors move in the opposite direction, which actually can be resolved by removing this 2 lines. Therefore THIS problem is solved. If you have another problem regarding a different (perhaps similar) topic, feel free to create a new issue so that the developers can keep track of the different issues.

gsxr2001 · 2020-03-11T18:19:08Z

The problem of opposite direction can be solved commenting (disabling) QUICK_HOME. The result is the same

thinkyhead · 2020-03-11T22:08:27Z

If QUICK_HOME doesn't work for your particular tuning of TMC drivers, it is recommended to disable it. For many setups it does work at this point.

github-actions · 2020-07-03T03:04:39Z

This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.

thisiskeithb mentioned this issue Mar 7, 2020

QUICK_HOME and SENSORLESS_HOMING are not compatible #17084

Closed

thinkyhead closed this as completed Mar 10, 2020

github-actions bot locked and limited conversation to collaborators Jul 3, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[BUG] Sensorless homing with SKR 1.3/TMC2209 failed #17057

[BUG] Sensorless homing with SKR 1.3/TMC2209 failed #17057

gsxr2001 commented Mar 3, 2020 •

edited

Loading

gsxr2001 commented Mar 3, 2020

drewzh commented Mar 3, 2020 •

edited

Loading

b00gie-dev commented Mar 4, 2020

cropduster commented Mar 7, 2020

onek24 commented Mar 7, 2020 •

edited

Loading

brettvitaz commented Mar 7, 2020 •

edited

Loading

thinkyhead commented Mar 8, 2020

thinkyhead commented Mar 8, 2020 •

edited

Loading

onek24 commented Mar 8, 2020 •

edited

Loading

gsxr2001 commented Mar 8, 2020 •

edited

Loading

onek24 commented Mar 8, 2020

adamrb commented Mar 10, 2020 •

edited

Loading

gsxr2001 commented Mar 11, 2020

onek24 commented Mar 11, 2020 •

edited

Loading

gsxr2001 commented Mar 11, 2020

thinkyhead commented Mar 11, 2020

github-actions bot commented Jul 3, 2020

[BUG] Sensorless homing with SKR 1.3/TMC2209 failed #17057

[BUG] Sensorless homing with SKR 1.3/TMC2209 failed #17057

Comments

gsxr2001 commented Mar 3, 2020 • edited Loading

Bug Description

My Configurations

Steps to Reproduce

Additional Information

gsxr2001 commented Mar 3, 2020

drewzh commented Mar 3, 2020 • edited Loading

b00gie-dev commented Mar 4, 2020

cropduster commented Mar 7, 2020

onek24 commented Mar 7, 2020 • edited Loading

brettvitaz commented Mar 7, 2020 • edited Loading

thinkyhead commented Mar 8, 2020

thinkyhead commented Mar 8, 2020 • edited Loading

onek24 commented Mar 8, 2020 • edited Loading

gsxr2001 commented Mar 8, 2020 • edited Loading

onek24 commented Mar 8, 2020

adamrb commented Mar 10, 2020 • edited Loading

gsxr2001 commented Mar 11, 2020

onek24 commented Mar 11, 2020 • edited Loading

gsxr2001 commented Mar 11, 2020

thinkyhead commented Mar 11, 2020

github-actions bot commented Jul 3, 2020

gsxr2001 commented Mar 3, 2020 •

edited

Loading

drewzh commented Mar 3, 2020 •

edited

Loading

onek24 commented Mar 7, 2020 •

edited

Loading

brettvitaz commented Mar 7, 2020 •

edited

Loading

thinkyhead commented Mar 8, 2020 •

edited

Loading

onek24 commented Mar 8, 2020 •

edited

Loading

gsxr2001 commented Mar 8, 2020 •

edited

Loading

adamrb commented Mar 10, 2020 •

edited

Loading

onek24 commented Mar 11, 2020 •

edited

Loading