Homing issue

Home › Forums › TinyG › TinyG Support › Homing issue

Tagged: ,

Viewing 5 posts - 1 through 5 (of 5 total)
  • Author
    Posts
  • December 29, 2014 at 4:57 pm #7232
    Harry
    Member

    After reading many (very many) messages about home and limit switches I cannot find a solution for this homing issue. On my EshapeOko I installed NC switches and when I setup the TinyG according https://github.com/synthetos/TinyG/wiki/Homing-and-Limits-Setup-and-Troubleshooting#homing-setup-and-configuration, the carriage moves until engaging the homing switch, and then stops, without backoff. I double checked that the min switch goes to min and max to max, disconnected all other limit/home switches from TinyG.

    Coolterm reports: {“er”:{“fb”:438.02,”st”:31,”msg”:”Move time is NAN”}} followed by
    {“er”:{“fb”:438.02,”st”:28,”msg”:”Failed to get planner buffer”}}.

    The TinyG board info:
    [fb] firmware build 438.02
    [fv] firmware version 0.97
    [hp] hardware platform 1.00
    [hv] hardware version 8.00

    I put the dump file in: https://www.dropbox.com/sh/m5kupbg459q6anb/AABCrAKvi98lageOqL783fxPa?dl=0

    Any help/suggestion is appreciated.

    Best regards,
    Harry

    December 29, 2014 at 10:38 pm #7233
    cmcgrath5035
    Moderator

    Hmmm, nothing obvious in your $$ parameter setup.

    New machine, or have you used it for a while, without issues?

    You might try a $defa=1, to reset all parameters to factory default.
    Alas, you will then have to reset all you parameters, but it seems some units have some non-visible (via $$) issues sometimes.

    December 30, 2014 at 1:26 am #7234
    chmr
    Member

    Hi,
    because the same issue popped up just a few days ago, and both had a very low latch velocity, I experimented a bit.
    Turns out this problem occurs when both the latch velocity $_lv is ridiculously low, and jerk homing $_jh is ridiculously high. It seems this triggers a numeric underflow and tinyg tries to continue with NaN (not a number).

    Solution: Set $_lv higher (>= 50) and $_jh lower (< 1000). If the jerk is too high, the machine can’t follow anyway without losing steps, and this somehow defeats the purpose of homing to get a *reliable* position.

    Happy New Year to you all!

    December 30, 2014 at 6:36 am #7235
    Harry
    Member

    Problem solved, thanks for your help! Setting the lv and jh as advised solved the problem.

    Thanks and Happy New Year!

    December 31, 2014 at 7:10 am #7237
    cmcgrath5035
    Moderator

    Good find, chmr.
    Looking at current fw build parameters, $_lv=100 but $_jh=10000.
    I’ll suggest that those be reviewed.

  • Author
    Posts
Viewing 5 posts - 1 through 5 (of 5 total)
  • You must be logged in to reply to this topic.