Added rudimentary support for delta printers

[armageddon421]

Most changes should be enclosed in #ifdefs. Please check.

[kliment]

Can you explain this?

[armageddon421]

This messed up the separation between cartesian and delta coordinate space. The printf was to find out where the change came from, then I just commented the disturber out. I just noticed this mysel and I am changing the plan_set_position function to do the necessary calculations. Also, I will clean up the printfs.

[armageddon421]

I'm not really happy with changing the plan_set_position to reacalculate the delta position.
This would only work well for x=y=0, because otherwise the delta calculations would assume wrong xy-positions.

[armageddon421]

I guess I'll Ignore the command if x and y are not zero or the current values or something.

[kliment]

Well, that's no good for cartesian setups. It's used quite a lot. Delta users are going to expect the same functionality I guess, but we definitely should not be breaking it for cartesian.

[armageddon421]

Of course. Least I can do is add #ifdef. What exactly is it used for?

[kliment]

G92 is used to set position after manually moving the machine. A number of people with no endstops use it for homing (adjust position by hand or with G1 commands and then issue G92). It's used to restart a partially failed print by setting the last known printer position. It's also used (on E) to set filament position to a known value after changing filament mid-print. Other than the restart function all of these are possible with x=y=0. That limitation on delta would be better than making it a no-op.

[armageddon421]

I'm breaking my head over how to avoid incostistencies in behaviour.
If I force it to xy zero and the user entered something other than 0, The
printer will just move to the center with the next G1 X0 Y0 instead of the
set position. If I do the recalculations anyways, the delta coordinate
space might be messed up. If I ignore the command for xy != 0, well
obviously the command is ignored... Which is what the user least expects.
However, if it is really only used to set "sane" positions and not to set
an offset or something, this could maybe used to continue prints - even on
delta printers. Decisions, decisions,...

On Mon, Apr 29, 2013 at 10:21 AM, kliment [email protected] wrote:

In src/gcode_parser.c:

@@ -385,7 +421,8 @@ static int gcode_process_command()
st_synchronize();

                GET_ALL_AXES(current_position,float);

•               plan_set_position(current_position[X_AXIS], current_position[Y_AXIS], current_position[Z_AXIS], current_position[E_AXIS]);
  

•               printf("from G92\n\r");
  

G92 is used to set position after manually moving the machine. A number of
people with no endstops use it for homing (adjust position by hand or with
G1 commands and then issue G92). It's used to restart a partially failed
print by setting the last known printer position. It's also used (on E) to
set filament position to a known value after changing filament mid-print.
Other than the restart function all of these are possible with x=y=0. That
limitation on delta would be better than making it a no-op.

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/14/files#r3994832
.

Robert

[kliment]

Okay, I'm going to add you as a contributor. You clean up the printfs and make the second config file and then you can merge it yourself once you feel it's ready.