Probe reporting false readings (sometimes)

Basic Information:

Printer Model: Fysetc Prusa Clone MK3s
MCU / Printerboard: Einsy
klippy.log (1.6 MB)
Klipper version: v0.11.0-175-gcba119db (RatOS)

Describe your issue:

Dear all,

I´m facing a strange issue with my Fysetc Prusa Clone. For bed mesh I´m using a Pepperl+Fuchs probe. From time to time it gives false readings (about 0.2-0.5 from what I have seen) which will let the printer happily smash the nozzle into the bed. When the problem occours 99,9% a FIRMWARE_RESTART will fix the issue and readings are back to normal.
I doubt it could be a temperature issue since printer is always heating at the same position. It happens if I start a second print right after I removed the previous from bed and it happens also if there are several hours/days between prints. For the record I´m using a 7x7 mesh and issue is NOT happening always (had 5 successful prints before the issue occured). I´m having the issue for about 1 or 2 years or so (fluid, MainsailOS, RatOS didn´t make a difference). I´ve asked google about it and also some more experienced 3D printing guys but that didn´t help.

This is a successful mesh:
17:22:35 probe at 0.000,1.000 is z=1.065000
17:22:38 probe at 34.000,1.000 is z=1.222500
17:22:41 probe at 68.000,1.000 is z=1.255000

This is a faulty mesh:
18:17:04 probe at 0.000,1.000 is z=1.280000
18:17:07 probe at 34.000,1.000 is z=1.437500
18:17:09 probe at 68.000,1.000 is z=1.482500

Time between prints were roughly 30 minutes or so. I issued a M112 right after I noticed the false readings.

klippy.log is attached.

Any help would be highly appreciated. If you need more data just tell me and I will happily provide it.

Cheers,
Christian

It could potentially be an EMI issue. Check the homing_heaters function and see if it helps

Thanks for the link, I added the section to my printer.cfg and will do some prints - I will report back with the results.
Btw. if your assumption is correct this issue should not occur if I do a BED_MESH_CALIBRATE before bed and extruder will be heating up right?

Unfortunately it just happened again after 4 prints or so

19:43:08

echo: Pre-heating extruder…

19:44:25 probe at 0.000,1.000 is z=1.652500
19:44:28 probe at 34.000,1.000 is z=1.835000
19:44:30 probe at 68.000,1.000 is z=1.880000
19:44:31 Klipper state: Shutdown

New klippy.log attached.
klippy.zip (938.1 KB)

Anybody else having some ideas?

Cheers,
C