BLTouch Clone only works sometimes (rarely)

General Discussion
1

Basic Information:

Printer Model: AnyCubic I3 Mega
MCU / Printerboard: Raspberry Pi / TriGorilla
klippy.log (50.4 KB)

Describe your issue:

TLDR: BLTouch only works every ~10 tries

Hi there, i am trying to get a BL-Touch clone to work for BedMesh leveling.
It is connected directly to the RPI with this configuration:

[bltouch]
sensor_pin: ^rpi:gpio0
control_pin: rpi:gpio27
z_offset: 2
x_offset: 12
y_offset: -25
#pin_up_touch_mode_reports_triggered: true
#probe_with_touch_mode: false

using the commands Sending BLTOUCH_DEBUG COMMAND=pin_up /down works perfectly everytime when the pin is not over the bed and can completely retract.

if the pin is above the bed and i trigger the pin to lower (BLTOUCH_DEBUG COMMAND=pin_down) it lowers as expected but goes directly to some kind of error state (blinks fast blue/red). in this state i cannot do anything until the pin is retracted again by BLTOUCH_DEBUG COMMAND=reset. than i can do the same again.

when i do “PROBE” it lowers the pin but reports “BLTouch failed to deploy” and "Failed to verify BLTouch probe is raised; retrying.
" then resets the pin and i can start it again without an actual measurement.

interestingly 2 times it did actually report a measurement. but current i cannot reproduce it.

reading the documentation i also went through BL-Touch “clones” section and tried
pin_up_touch_mode_reports_triggered and probe_with_touch_mode with both fals and true without any change in the behaviour.

The Probe itself is around 2 millimeter above the nozzle in the retracted state.

What am i missing here? is there anything else i could check?

Cheers and thanks a lot!!

IMG_3599
IMG_3599640×480 38.5 KB

2

I would move the probe a little bit up.

3

Hi there,

thanks a lot! i did make it to work, but i am kind of unhappy with the accuracy.

probe accuracy results: maximum 0.975000, minimum 0.802500, range 0.172500, average 0.868250, median 0.832500, standard deviation 0.067897

for me that looks like i might be more acurate by using the good old “paper-methode”

4

It seems your clone has a very bad quality. Typical standard deviation are at least a factor 10 smaller than your value, e.g. just tested on one of my printers:

 16:37:45  // probe accuracy results: maximum 2.232681, 
minimum 2.225971, range 0.006710, average 2.228405, 
median 2.228233, standard deviation 0.001997
5

Maybe running 5V Hardware in 3,3V Pins?