Undrestanding G38.2 probe file output

[bogflap]

Can anyone interpret the contents of the file referenced by a PROBEOPEN statement. I have a 3 axis mill, the x and y values make sense but I can make no sense of the z value. Also there are nine entries per G38.2 statement executed. I assume this is a feature, sorry bug?

[Big John T]

AFAIK, it would be one for each possible axis.

John

[bogflap]

As far as I can make out this is the code that outputs the x,y,x,a,b,c,u,v and w coordinate values. Possibly this called once per gcode co-ordinate system. Have not got that far yet. However the Z value is at odds with the displayed value in the AXIS opengl window. Whereas axis reports a value of say -0.0057 (inches) the probe output file contains a value of -0.001322. Even worse I get an axis value of -0.0040 (inches) the probe output file value is 0.000278. Thats a change of sign so this is not simply a matter of scaling. I am still trying to get my head around the code so any help in this specific area would be welcome

[Big John T]

I did get a verification on that and it is X Y Z A B C U V W as you have discovered. And I'll add that to the manual...

I would have to assume you have some kind of offset set.

John

[bogflap]

Had the same thoughts about having an offset set.Trouble is what offset?
I am trying to find the bit of axis code that populates the Z axis GUI field and see how it doe it. I am not concerned about whether axis displays the right answer or the probe file has the right answer. Hey whats right and whats wrong, all I want to know is why they are different.

[bogflap]

Ok. So ran up AXIS 2.3.0-beta1 and used the log file functionality )LOGOPEN, LOG, LOGCLOSE). This all works nicely.

[Big John T]

In 2.3 you can clear the offsets from the Machine menu.

John

[bogflap]

Have started using 2.3 beta 2. Had a 'doh' moment when I realised that the probe values for Z are when the probe switch trips. However the value displayed by AXIS is the Z position after EMC has stopped decelerating the probe. I did a few measurements and found that the difference between the two constant down to the fourth decimal place. Phew! Anyway having read the 2.3 docs and found that there are four variations on the G38.2 (3,4,5) command I am now looking at taking measurements when the probe approaches the workpiece and also when moving away from it. I wonder if the mean of the two will be any more accurate.
Thanks for your help (and patience) big john t

[Big John T]

The extra probing codes allow you to do faster probing as I understand it. Take a look at the smart probe pointed to here

http://www.linuxcnc.org/docview/deve...de.html#r1_1_4

John

[bogflap]

Like the smartprobe.ngc code. Keeps the probe tip a 'deceleration distance' away from the workpiece during the x direction traversal. The G38.3 detects if the probe collides with the workpiece when doing a lateral move.
Like it and many thanks Big John T

[Big John T]

Your welcome

John