Re: [CAD_CAM_EDM_DRO] Re: suggested improvement for EMC

Ian Wright wrote:

> What would be the chances of being able to return to the start of the last
> cut segment or, in other words, the last previous 'Gxxx' instruction. This
> would probably then cut a bit of clear air before hitting metal but should
> pick up cleanly.

I had an Allen-Bradley 7320 running for a while before switching over
to EMC. It had very sophisticated features for moving around within
a program. For instance, at any time, you could hit Cycle Stop, and then
select a handle for program skip block, program skip slew or program
skip search, which would allow you to skip forward or back a single
block at a time; or as fast as the CPU could read program blocks from
the source as long as you held the direction button down; or to search
for a specific program block number. Then, you just hit Cycle Start
to continue from that point. Of course, you had to read the G-code
to make sure you would not run afoul of modal settings, etc. For
instance, if there was a
N120 G01 Xxx
N130 Yyy
N140 G02 Xxx Iii
and you aborted in the G02 and backed up to the N130 line, it would croak
because Yyy is not a properly specified G02 move. But, this was REALLY
nice when tinkering with new programs.

They also had another feature called jog retract, which would allow you to stop,
even with a feed hold in the middle of a move, retract one axis (perhaps to clean
or change a tool) and then hit Cycle Start, and the machine would move the
retracted axis back to the position it had been in, and resume the move that was
in progress when you stopped or paused! I thought this was really exotic, but
could see how it could be useful in a production shop.

Anyway, there are reasons the EMC interpreter is a "One Way" program,
having to do with lookahead, but it is a handicap, and some workarounds to
reduce the handicap would be very useful.

Jon