Re: RS274 vs RS274-NIST

Hi Alan, Jerry,

According to my RS-274D spec, G78 is fair game for anything (G76-79
are unassigned). G83 is peck drilling in the same standard, which
TurboCNC doesn't include at this writing.

So why did I write my own cycle, when one already exists? Well, a
G83 cycle positions XY, then pecks in Z.

But on a lathe, the tailstock axis is usually W, not Z. When the
user defines his axis setup, he might want to peck drill in some
other peculiar way, or use the "peck turning" trick to get a short
chip in plastic as well. So, I took the more general approach and
wrote a G78 cycle. The following two code snippets are equivalent:

;EIA 274D Standard
G83 X1 Y1 C0 D-1 K-0.100 F5

;TurboCNC 3.0f
G00 X1 Y1
G00 Z0
G78 Z-1 I-0.100 F5

I'm rather humiliated to point out that the download version on the
webpage has a bug in G78 that will trip the program up if the peck
drilling has to cross the zero point of its axis.

Alan, another release *is* planned for sometime this weekend. In
addition to other things (like fixing the G78 bug), I laboriously re-
wrote all of the time critical motion code into assembly language,
which adds an enormous speed increase when interpolating.

G81 will be in there as well as G92, although I haven't decided on
whether to include some of the other canned cycles.

Suggestions on new things to add or improvments to make to the
software are welcome...

Dave Kowalczyk
Everett WA
TurboCNC machine controller ->> http://www.dakeng.com

--- In CAD_CAM_EDM_DRO@y..., Alan Marconett KM6VV <KM6VV@a...> wrote:
> Hi Jerry,
>
> Might be a little more like "did Dave make his G78 function" match
the
> RS274 NIST, or another "standard". Good write-up in NIST (or one
of the
> papers) for the Drill cycle. I even tried writing the G81 drill,
G82
> drill with dwell, and G83 Drill with peck cycles. That takes some
> work! I'd study the parameters, and watch the cycles run. You
should
> be able to observe the pecking cycles go deeper and deeper, and
retract
> to the appropriate level. You'll also want to observe if G98 (Old Z
> retract mode) is implemented. Throw in Absolute or Relative mode
on top
> of it (makes those 'L' repeats drill a series of holes) and it
REALLY
> gets complicated! (I'm not sure I've got it ALL working...yet).
>
> I was thinking that a digitizing probe "cycle" might be a good one
to
> implement! Unless I can find one to copy.. er, implement.
>
> Perhaps Dave can comment on his implementation! (I've heard he's
> working on another release).
>
> There is a Gcode list on Yahoo that attempts to collect and compare
> Gcodes. Not much activity as of late.
>
> Alan KM6VV
>
>
> Carol & Jerry Jankura wrote:
> >
> > Guys,
> >
> > I've started reading through the NIST release 3 publication, one
chapter of
> > which describes the operation of the various G-Codes that the
NIST system
> > supports. It appears that the NIST version supports some codes
that my
> > controller does not and does not support others that it does. For
example,
> > NIST uses G83 for "peck" drilling while the controller I'm
working wirh,
> > TurboCNC uses G78 for what appears to be the same function.
> >
> > Now, I do understand that there are several "standards" that
masquerade as
> > RS274. Is the NIST G83 code the same as the TurboCNC G78 code, or
are there
> > some subtle differences between the two? Are there any
other "equivalent"
> > codes that port from one machine to another. Would I get
the "correct" peck
> > drilling cycle if I assigned the G-Code G83 to the same procedure
that
> > TurboCNC uses for G78? Does any table that lists all of the known
G-Codes
> > exist? If so, where?
> >
> > Thanks,
> >
> > -- Carol & Jerry Jankura
> > Strongsville, Ohio
> > So many toys, so little time.... (What's a G-Code among
friends?)
> >