X-axis offset

Home › Forums › TinyG › TinyG Support › X-axis offset

Tagged: 

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • July 2, 2015 at 4:55 pm #7923
    dhyeylu
    Member

    Hi,

    So I have Tinyg v8 cnc controller and I have been using for one of my projects.

    I am trying to parse the JSON text file that is outputed by the controller.

    When I see the file, I observe that the x-axis start point is always printed as 5mm instead of 0 which is where I am starting from (I have homed the system correctly).

    So because of the this what is happening is when I ask the tinyg to move from -1.5cm to +1.5 cm, it moves correctly but the tinyg outputs -1cm to 2cm.

    Any idea how I could resolve this error. I can always post-process this but I am wondering if there is any setting that I missing out on.

    Please Help!!

    July 2, 2015 at 10:20 pm #7924
    cmcgrath5035
    Moderator

    Off the top, looks odd.
    Tells us a bit about
    What firmware build are you running? $fb=440.14 is most up to date
    What OS do you use to control (Mac, Win, Linux, ) ?
    What program to deliver Gcode?(CoolTerm,. Chilipeppr, ) ?

    When you say “JSON text file that is outputed by the controller.”, you mean the streaming status reports that come back while Gcode is running?

    We will likely want to see your full parameter set sooner or later

    July 6, 2015 at 1:21 pm #7942
    dhyeylu
    Member

    The firmware build I am using is 435.10.
    I am using Windows OS.
    I use TgFX to control the robot.

    Regarding the file that is outputed, you are correct. It is the streaming status reports that the controller spits out.

    I am giving a snippet of what I see below:

    {“sr”:{“mpoz”:-30.208}}
    {“sr”:{“mpoz”:-30.125}}
    {“sr”:{“mpoz”:-30.042}}
    {“sr”:{“vel”:0.00,”mpoz”:-30.000,”stat”:3}}
    {“sr”:{“vel”:25.03,”mpox”:5.065,”stat”:5}}
    {“sr”:{“vel”:50.00,“mpox”:5.143}}
    {“sr”:{“mpox”:5.222}}
    {“sr”:{“mpox”:5.305}}
    {“sr”:{“mpox”:5.389}}
    {“sr”:{“mpox”:5.468}}
    {“sr”:{“mpox”:5.551}}

    As you can see, this is the first instance where there is an X-direction translation. And instead of beginning from 0, it begins printing from 5mm.

    At the moment I am using a hack where I postprocess the data based on the above observation and it has dramatically improved my accuracy. But I would like to get a more permanent and portable fix.

    I am not sure which files will be useful for you to debug the issue. If you can let me know, I can add those to dropbox and provide you the link.

    Thanks!!

    July 6, 2015 at 3:26 pm #7943
    cmcgrath5035
    Moderator

    I assume you have read this a few times:

    And this is somewhat a guess.
    I search the wiki, cannot find a definition of mposx or mposz, but guess they are Machine position x and z. (Machine position relative to 0,0,0 after homing cycle)
    I’ll also assume that your Gcode is not running in the G53 coordinate system, rather it is running in G54 or perhaps another.
    Perhaps the “offsets” you are seeing are the offsets of G54 relative to the Machine (Homed, G53) ?.

    Perhaps by defining a different SR string, requesting posx/posy/posz rather than mposx, etc., you will see what you are looking for?

    There is also a possibility of a bug in the reported position reporting, I know one has been identified in $fb=440.14, cannot tell if it is yours.

    You need to know that:
    A. tgFX is very much out of date and no longer supported. Don’t design around it and wean yourself off to something else. I doubt it is contributing to this issue.
    B. $fb=435.10 is very old, if this is a bug, it will be fixed in $fb > 440.14
    You really should update to 440.14, many fixes. High likelihood tgFX will misbehave with 440.14; that will not be fixed.

    July 7, 2015 at 9:21 am #7949
    cmcgrath5035
    Moderator

    Oops, I just realized that I did not scroll down the screen far enough, and that yes, mpox is Machine (G53 coordinate system) x position. See this table:

    So I am sort of guessing that you have $xzb, X zero backoff, set to 5mm and that is what you are seeing.

    For (your) programming, best solution might be to define a SR that provides posx, which would be position in current coordinate system by default G54.
    That should be the 0.00 you think you want.

    As you migrate away from tgFX, give Chilipeppr a look. There is currently a really nice/powerful update being finalized that makes coordinate systems much clearer and easy to use

    July 7, 2015 at 10:12 am #7950
    cmcgrath5035
    Moderator

    Hmm, after reading this item:

    I come away wondering if your observed offset might be related to this bug or an earlier version of it.

    I would strongly recommend you move to FW 440.14 (or greater, a new build is currently being evaluated) before banging your head against 435.10

    July 8, 2015 at 9:50 am #7952
    cmcgrath5035
    Moderator

    Having read this a couple times:

    I am fairly sure this offset you are seeing could be related to the identified bug in fw 440.14.

    After a successful home cycle, my read is that $mpox and $posx should both be 0.0.

    A fix is in-progress, I suggest upgrade to 440.14 and reevaluate.
    Bugfix release will be released soon.

  • Author
    Posts
Viewing 7 posts - 1 through 7 (of 7 total)
  • You must be logged in to reply to this topic.