grblHAL probing with BSMCE04 (STM32F1)

[PolyEng]

hi , i'm having this issue w probing on this hw :BSMCE04 (stm32 mounted is genuine , at least it seems so no clones)

essentially probe triggers correctly i can see it in the GUI of IO sender and on the console of IO sender but when i issue for example a G38.2 machine doesnt move and reports probe position where it is (probe IS NOT triggered ) checked with oscilloscope for noise .. everything is fine.
console output says "ok" and cycle continues

if i manually TRIGGER the probe and issue a G38.2 the cycle errors out (as it should) since probe is already trggered before starting the probing operation.

so probe is fine electrically and switches correctly but fw has some bug i think?

(before you blame my wiring or "noise", the probe is isolated through opto and the same hw setup of the probe is working correctly if i swap the controller for a PI PICO)

thanks in advance for whoever will look into it

i'll include a snippet of gcode and relative output from console in verbose mode

G91F2
G0Y5
G0X-6
G0Z-3
G38.3F2X10
 
starting from this coordinates i have to do Y+5   X-6 and Z-3
**<Idle|MPos:-3.500,23.508,9.996|Bf:100,1024|FS:0,0|Pn:HS|WCO:-3.500,23.508,9.996|WCS:G54|A:|Sc:|MPG:0|H:0|T:0|TLR:0|FW:grblHAL>**
Probing...
G91F2
ok
PM:ok
G53G0Y28.508
ok
PM:ok
G53G0X-9.5
ok
<Run|MPos:-3.500,23.528,9.996|Bf:98,1024|FS:90,0|Pn:HS|WCO:-3.500,23.508,9.996|WCS:G54|A:|Sc:|MPG:0|H:0|T:0|TLR:0|FW:grblHAL>
PM:ok
G53G0Z6.996
ok
PM:ok
G38.3F2X10
**<Run|MPos:-9.496,28.508,6.996|Bf:99,1024|FS:0,0|Pn:HSP>**
[PRB:-9.500,28.508,6.996:1]
ok
PM:ok


machine does all the G0 moves and arrives to destination... now G38.3F2X10 i should move X+10 till i find the probe .. but the machine doesnt move because **Pn:HSP **Probe is triggered, problem is probe isnt triggered **```

[terjeio]

When I test here probing works as it should (tested by triggering the input manually).
Please enable Run substatus in General > Status report options and post a complete log of a probing sequence.

[PolyEng]

When I test here probing works as it should (tested by triggering the input manually). Please enable Run substatus in General > Status report options and post a complete log of a probing sequence.

rn i have changed the board , it will take some time for me to swap again the board , i'll keep you updated , thanks in advance

[CBR900cc]

hello @PolyEng I've some issue with probe pin, in my case if I use z-limit pin with inductive sensor, all working fine, but if I connect inductive sensor to pb4 (probe pin) nothing work, board not read any signal of triggered or not... can you share you pinout connections/photos?

[terjeio]

@CBR900cc If you short the probe input to ground does it change the probe state?

[CBR900cc]

@terjeio yes, if shorted probe is triggered, I'm using this sensor LJ12 A3-4-Z/BX, can you explain me how wiring it?

[terjeio]

can you explain me how wiring it?

Blue to GND, Brown to 5V and Black to the probe input. Since the sensor is rated 6-36V it could be that it fails to pull down the probe input enough when supplied with only 5V? Can you measure the voltage at the probe input when it is triggered vs. not triggered?

[CBR900cc]

can you explain me how wiring it?

Blue to GND, Brown to 5V and Black to the probe input. Since the sensor is rated 6-36V it could be that it fails to pull down the probe input enough when supplied with only 5V? Can you measure the voltage at the probe input when it is triggered vs. not triggered?

yes, if is not triggered, output 24v if triggered 0.7v

[terjeio]

output 24v if triggered

Ouch - the inputs on the header are not protected by optoucouplers, there are only voltage dividers to step down 5V signals to 3.3V. If you are lucky you have not damaged the MCU. Either add an optocoupler, an additional resitor in series with the input or supply the sensor with a voltage that does not exceed the maximum allowed voltage on the MCU pin. IIRC it is 5V tolerant so 12V or less should be ok since the dividers uses 1K (to ground) and 2K to the input.

[CBR900cc]

ops... @terjeio now it is connected with a 1k between signal and gnd and 1k between signal and input header, but now if i do a probe cycle, cycle working correct and grbl see probe is triggered or not with iosender (if I use openbuildscontrol that see probe trigger and crashes), but sometimes it crashes and i need to replug usb cable

[terjeio]

but sometimes it crashes and i need to replug usb cable

Perhaps you should increase the 1K resistor between signal and input? 2K is better?

[CBR900cc]

2K

thank you, I try it and update you!