Tool Changer

i did the essential,

it already was installed

so i still have no clue on fixing it

kind regards Robin

Hi John / Andy

The fact that this compiles OK on your systems is bugging the hell out of me, because if I try either under the default emc 2.4.6 system or from my run in place emc 2.6pre from git, I get the errors below.

Given the code is only 75 lines long, the error occurs on line 77 and it does not contain a get_data_size() function, I have to assume that this is one of the routines added by comp (error checking?)

Any ideas? I don't want to write user space components in python, but that is the only thing that seems to work.

In the mean while I will settle down to read through comp ( python file to generate a C file!) and try to work out where the added function etc are sourced from

regards

OUTPUT
make -C /usr/src/linux-headers-2.6.32-122-rtai SUBDIRS=`pwd` CC=gcc V=0 -o /Module.symvers modules
make[1]: Entering directory `/usr/src/linux-headers-2.6.32-122-rtai'
CC [M] /tmp/tmpnxxUZY/toolchanger.o
toolchanger.comp: In function ‘_’:
toolchanger.comp:77: error: invalid storage class for function ‘get_data_size’
toolchanger.comp:77: warning: ISO C90 forbids mixed declarations and code
toolchanger.comp:77: error: expected declaration or statement at end of input
make[2]: *** [/tmp/tmpnxxUZY/toolchanger.o] Error 1
make[1]: *** [_module_/tmp/tmpnxxUZY] Error 2
make[1]: Leaving directory `/usr/src/linux-headers-2.6.32-122-rtai'
make: *** [modules] Error 2

PS -
As a test I have just looked for the most non-trival looking .comp file in the source tree and compiled it using comp.
I chose bldc.comp (appears also written by Andy) and it compiled and linked OK and the module was copied across to the modules directory.
Given that I had compiled the 2.6pre system from source in the first place I knew there was nothing on my system preventing it, but really does make my inability to compile the toolchanger.comp file whilst others apparently can, very strange.

Thank you John, I'll be *****ed!

All I did was cut and paste from the first page into gedit, but there must have been something there, because I just repeated the process and got a file exactly the same length, but this time it compiles!!!!!!!!!

I will check it loads and I can see the pins etc in a while, but have no reason to suspect that it won't.

The answer is often simple AFTER you have found it!

Hi again

Having got this working I was inspired to try to write my own component to suit my particular toolchanger. ( Boxford 240F)

It works well except for one thing, I cannot find find a way of connecting to stepgen.3.position-cmd
because it is already linked earlier in the .hal file (net apos-cmd axis.3.motor-pos-cmd => stepgen.3.position-cmd).

The stepper rotating the ATC is notionally axis A which allows me to jog it and move it in gcode.

I can track the commanded and actual position through pins linkages to the component but can't find a way to set it

I also seem to recall that in M6 moves, all jogging and other axis movement at a higher level is suspended until M6 is complete, which presumably means the stepper can only be moved by directly accessing the stepgen.

Can you enlighten me as to how this is best achieved?

regards

PS

Think I have answered my own question

I unlinked stepgen.3.position-cmd and stepgen.3.position-fb from axis.3 and linked them to the relevant pins in my toolchanger component, which operated the ATC in simulation.
Looks like there can only be one source of position commands for an axis.

I could run a script which linked them back to axis.3 when manual control was required, or perhaps have 2 stepgens and switch between them, have to investigate further.

I have the loadrt and addf

but now i got the set all the Inputs and outputs

someone know what is the best way to do it and can someone help me getting started with it

greetings Robin