Mark

it will homing itself and move the nozzle to the zero position and then calibrate itself if the collision_calibrate:1 after M102 S-6 . about the collision : https://github.com/markniu/Bed_Distance_sensor/wiki/Collision-sensing

correct, you can use this value: ``` homing_speed: 5 second_homing_speed:3 homing_retract_dist:5 homing_retract_speed:2 ``` like in this video: https://www.youtube.com/watch?v=RuPoXbrSPDc&feature=youtu.be and need to careful, and don't do bed mesh or z tilt...

the reason is that the nozzle was still touch on the bed after it move 0.6mm while calibrating you can see the data by M102 S-5 too seems you let...

one more test you can do, read the raw data at different position directly by M102 S-7, that can help to understand how it works 

This is strange. especially the Raw data 640 at 0mm , how about set the second homing speed from 3mm/s to 4mm/s? > I tried reverting back to the configuration...

what's the data return by M102 S-7 after " 1015 at 0mm" ? yes, may be there is a miscommunication , for test you can also increase the delay value...

is the 0.3mm returned after G28? like `at 0 mm, BDsensor to bed: 0.3 mm` ,if it is that is not the max difference, but the mount distance of coil....

Recommend to decrease the delay value in BDsensor section from 20 to 10, because it can slow down the time consumption of klipper software

here the actual error is the timer too close, and I will change this message `fail: update z must be >0 ` as a warning message later. you can add...

Seems the klippy.log file was not updated successfully. > Different error this time Uploading klippy.log (13).txt…