[Dale Grice]

In the attached pic, there is an oil grove that starts at the end, goes in roughly an inch in 270 degrees and then backs out in 270 degrees. In the pic you can't see the exit, but is 180 degrees from the entrance.

I could write a threading G-code, or use a canned Mach4 threading program to do the inbound groove and then do a second program to do the outbound. But how would one "connect" the grooves together? How does Mach determine at what point in the circumference to start?

Another option would be an inbound G-code followed by an outbound G-code. Then move the tool in the X direction. then do the same in and out G-code. But would the tool start at the same spot like threading?

Thanks for any insight. I guess I need to go experiment.

Dale Grice

[Superman]

It nearly seems to be a milling operation. Anything with a fixed start AND finish point would be that type of op.
If quantity is low, possibly done with a manual die grinder ( it is only an oil groove).

Your RPM (if threading) would be ridiculously slow, just to control the lead ins & outs

[joeavaerage]

Hi,
Mach does not thread per-se, it is the motion controller that does it. The motion controller gets at the very least an index signal from the spindle
and the controller uses that index signal to start the thread. This is how multiple cuts always start in the same position. Mach can't do that, the delay for the signal
to get to Mach and then for the signal from Mach through a motion buffer of several hundred milliseconds is just too greater delay.
Ergo the motion controller must do it, because it and it alone can act fast enough.

If you have a rotary axis that is position capable the you can do as you require.

For instance if you have a A axis the:

G1 X25.4 A270 F720

would mean the X axis would advance one inch (25.4mm) while simultaneoulsy the A axis would rotate 270 degrees at a rate of 720 degrees per minute. So 270 degree toolpath would take 22.5 seconds.
If this were followed by:
G1 X0 A540
then the X axis would withdraw one inch back to its start position while the A axis rotated another 270 degrees. At the end of both moves the part would have rotated 1.5 rotations and the end of the groove would
be 180 degrees from the start of the groove.

Does this sound like it would work?

I took the liberty of throwing a little piece of brass in the mill which has a position capable fourth axis and running it with a 1.5mm endmill. Note that this does not rely on any special property
of the machine or controller but just relies on two axes being coordinated, and that is after all what CNCs do. In this case one axis is rotary and the other linear....but does that not just
mean a spiral toolpath?, and is that not what you want?

Craig