[bogflap]

Where does the output from these statements actually go, is there an output file or something ?

[Big John T]

It is in the manual.

http://www.linuxcnc.org/docview/deve...l#sub:Messages

John

[bogflap]

However this poses the next question The 'message display device', Could anyone let me know what is or where it is or how to I define one

[Big John T]

While I have not tried it I assume that you need to run EMC from a terminal window and the "default console" will be the terminal window you started EMC from.

Which leads to the next question... What are you trying to do?

John

[bogflap]

Basically using the grid probe ngc file to measure the 'flatness' of a piece of blank pcb. This is for isolation routing of a circuit board. Given that I can measure the relative 'flatness' I can then adjust a gcode file so that the isolation routing depth is constant (or 'more constant'). This avoids all the hassle of having to get the blank pcb absolutely flat using mechanical means.Simply probe the board whilst held in place using staright forward nuts, bolts and washers then probe using a fairly small mesh size and adjust the gcode input file in the z will compensate in software. By the way nigh on thirty years of coding means I see that none trivial task you thought of as quite achievable.
What I basically need is a simply formatted text file with all the Z values in it so that I can calculate all the relative changes across the probed mesh.

[bogflap]

Using the LOG functionality in AXIS 2.3.0-beta1means I can pick and choose what I output and it gets written to the log file. I will go with this.

[Big John T]

Are you checking the flatness of a wavy part or just getting the angle of a flat part before you start to machine it?

John

[bogflap]

The basic problem is avoiding using expensive bits of kit to hold the pcb flat e.g. a vacuum plate. Basically securing the pcb along the edges still gives some slight bowing in the pcb especially towards the centre of the pcb. Some folk even seem to super glue the boards down to a known flat surface. Anyway bowing seems to occur even on quite small pieces of pcb (checked with a dial gauge). I am assuming that first of all the mill table is flat and also that the pcb manufacturing produces some pretty accurate pcb thicknesses. Either way it does not matter if I can correct the g code to take into account any variations from whatever source.
Many thanks Big John T

[guru_florida]

If you start emc from a terminal the output will go to that window. I am not sure if it is writing to stdout or stderr though. You could easily redirect any of this to a file with simple shell redirection using 'emc > myfilename'.

Also, as powerful as linux is, I am sure you could setup some sort of redirect to a terminal type window. For example, redirect output to 'netcat' (apt-get install netcat) using a random port on the localhost and with netcat in listen mode. Then use a Gnome telnet program to connect to the netcat port you configured. EMC2 output will then be shown on the telnet screen. Put the emc2/netcat command in your desktop/menu shortcut so this setup runs by default and you can telnet at any time to see emc2 output.

The command would be something like this:
/pathto/emc | netcat -l -p 4000 --source=127.0.0.1

If emc2 happens to be writing to stderr instead of stdout then the pipe (|) requires 2 for stderr fileno:
/pathto/emc 2| netcat -l -p 4000 --source=127.0.0.1

Remove the --source=127.0.0.1 and you would be able to telnet to the emc output over another networked computer to monitor progress! hehe

[Big John T]

Basically using the grid probe ngc file to measure the 'flatness' of a piece of blank pcb. This is for isolation routing of a circuit board. Given that I can measure the relative 'flatness' I can then adjust a gcode file so that the isolation routing depth is constant (or 'more constant'). This avoids all the hassle of having to get the blank pcb absolutely flat using mechanical means.Simply probe the board whilst held in place using staright forward nuts, bolts and washers then probe using a fairly small mesh size and adjust the gcode input file in the z will compensate in software. By the way nigh on thirty years of coding means I see that none trivial task you thought of as quite achievable.
What I basically need is a simply formatted text file with all the Z values in it so that I can calculate all the relative changes across the probed mesh.

You can actually write your g code file from a g code file while probing! I had a discussion on that the other day. You LOG what you want to the new g code file like this:

#8=-.1 (Z probe)
(LOGOPEN,probe-results.ngc)
(LOG,G1 X#5061 Y#5062 Z#5063)
G38.2Z#8
(LOGCLOSE)

The (LOG,G1 X#5061 Y#5062 Z#5063) will write the values in the variable file from the most recent probe along with the text...

http://cvs.linuxcnc.org/cgi-bin/cvsw...e=text%2Fplain


John