BL Touch/Z-Probe is not registering properly – starting 1.21RC5



  • I did not notice this because it doesn't happen during normal operation.
    I am re-configuring the printer after some gantry changes, and at first it looked like maybe wiring related, but now I pin-pointed it to a change after RC4.
    The BL touch specific config values are here

    M558 P9 X0 Y0 Z0 H7 F100 T5000 A10 S0.02
    G31 X-30 Y-31 Z1.4 P25
    M280 P5 S90

    When i do G32 –
    here is bed.g
    G91 ; relative axis movement
    G1 Z4 F400 S2
    T0
    G90
    G30 P0 X520 Y279 Z-99999
    G30 P1 X40 Y279 Z-99999 S2 ; probe left point

    it works fine as it's moving to middle of bed at Y279, but when I tried to do G29 to redo the map, it says probe is already triggered.
    Basically somewhere Y 40 or below it sometimes says probe is already triggered when I do G30 S-1 or G29 which starts at that point. Anything below 35 -- will only work 20% of the time.
    I was going to run a new wire for the z-probe pair, but then tried to roll back to RC4 and no false positives anymore, took a lot of troubleshooting -- with RC4 I can go to Y20 (safe position to probe) and it works fine every time.


  • administrators

    Try adding a probe recovery time, R parameter in M558. I suggest 0.5 seconds to start with. If that fixes the problem, then lower values may work.



  • I tried recovery (first thing I did – it was somewhat helpful, but it was still unpredictable, sometimes it worked, sometimes it didn't ) -- I went as high as 2S - also changing
    G31 X-30 Y-31 Z1.4 P5 -- lowering from P25 -- helped sometimes, I thought actually that solved it until I tried this morning and it was glitchie again, that's when I tried rolling back to RC4 and boom -- fixed, works with P25 and no need for recovery.



  • PS – I put an scope-probe on the ground/signal pins, and there is no difference to how the BL touch responds based on Y position, which is why it doesn't make sense that wires would be an issue -- it is almost as if something invalidates the Y position in the newer firmware versions, and refuses to probe at a given point.


  • administrators

    1. What voltage does your scope see between signal and ground, when the bltouch isn't triggered, and when it is? Can you see any noise or glitches on that line?

    2. P25 in the G31 command is very low. What happens if you use P500?



  • @dc42:

    1. What voltage does your scope see between signal and ground, when the bltouch isn't triggered, and when it is? Can you see any noise or glitches on that line?

    2. P25 in the G31 command is very low. What happens if you use P500?

    1. As I recall the voltage was correct at 0 and 3.3v -- normally it reads 0 -- and says triggered, when it's probing, it pulses very rapidly, so I'd have check the frequency with which it triggers, but the I don't see any noise present based on Y position...

    2. I tested P500 briefly, and saw no change -- it worked in the Y positions it did before, and didn't where it didn't.

    I was testing with v21 final I guess, and I saw it report already triggered at for G30 S-1 Y=37, then Y=38 would be fine. It also seems somehow connected to Y min and U min settings -- I have Y min set to 0 and U min at the time was set to 37... U is secondary Y axis -- also there are additional X axes coordinates, and they're configured to not be able to reach this point, which shouldn't matter. Not sure if you started checking min/max of various axes for probing -- I don't see why that would be necessary and why it would be rather selective, basically confusing why it works on all points with RC4.



  • Sorry for the noise. I redid more tests, and found that the probe signal was disappearing as the gantry moved along Y axis – I will re-test the wire to check where it had a problem, but there was definitely a problem in the wire, maybe the new firmware has different value of a pullup, or just dumb luck -- either case I replaced the wire and now the BL touch signal is solid -- not sure what happened -- silicone insulated wire in a cable chain -- I don't see any breaks in it -- but there is definitely an issue.


  • administrators

    I'm glad you solved it. These sorts of problem are usually caused by a bad crimp connection, typically a fracture in the wire where it enters the crimp.


Locked
 

Looks like your connection to Duet3D was lost, please wait while we try to reconnect.