Arcs incorrect

Home › Forums › TinyG › TinyG Support › Arcs incorrect

Viewing 6 posts - 1 through 6 (of 6 total)
  • Author
    Posts
  • February 26, 2017 at 11:58 am #10282
    RichS99
    Member

    Had the TnyG for a couple years, running on a homemade Ox. Things have gone reasonably well.

    After several months being idle, I brought things back to life, put up new firmware, etc. Now I’m having the hardest time getting accurate cuts. I think I’ve narrowed it down to arc mode instructions.

    Previews of GCode look fine, but do not reflect what’s carved. I’ve used CamBam, Aspire and manual GCode to try to check. I’ve updated firmware, disabled limit switches and checked and reset my settings.

    Current settings are here. This is a factory reset with the most minor edits to reassign the four motors properly and the angles.

    Example GCode going wonky is here

    Photo of the cuts going nuts is here

    And finally, I tried a manual arc cut in Coolterm – G2 X0 Y30 I0 J15 F200 – apparently should be a semicircle, but this is what I got.

    Like I said, the stuff seems to render fine in Chilipeppr, Aspire and CamBam. Currently using a Win10 OS.

    Very frustrated and hoping someone can assist.

    – Rich

    • This topic was modified 7 years, 10 months ago by RichS99.
    February 26, 2017 at 3:55 pm #10285
    Zootalaws
    Member

    That looks like one axis is calibrated wrong.

    So, a move +2 x units, for example, is only moving +1.5, while Y is moving +2

    Have you calibrated each axis?

    February 26, 2017 at 4:19 pm #10286
    RichS99
    Member

    Ugh, did that ages ago and now that I’ve reset I guess I need to go back and look how to do that again. Don’t think I have my old .conf files.

    February 26, 2017 at 4:37 pm #10287
    RichS99
    Member

    Wait, I don’t see how you think the X and Y are different…?

    M1 – x, $1tr=60 and $1sa=1.8
    M2 – y, $2tr=60 and $2sa=1.8
    M4 – y, $4tr=60 and $4sa=1.8, reverse polarity

    M3 is the z axis, which probably needs calibration separately.

    February 26, 2017 at 5:01 pm #10288
    RichS99
    Member

    Just measured x and y movement and, minus a bit of slop, are moving the length requested (29.89mm when calling for 30mm).

    Other ideas?

    February 26, 2017 at 7:53 pm #10292
    cmcgrath5035
    Moderator

    “Calibration” issues are likely belt stretch/tension.
    Arc bugs remain in FW, have been there for a while.

    mm node (not inch) works better, but if problem persists, best solution is to send Gcode that does not contain arcs.

  • Author
    Posts
Viewing 6 posts - 1 through 6 (of 6 total)
  • You must be logged in to reply to this topic.