CAD CAM EDM DRO - Yahoo Group Archive

Re: Postprocessor

Posted by Fred Smith
on 2001-05-13 21:10:39 UTC
--- In CAD_CAM_EDM_DRO@y..., "Tim Goldstein" <timg@k...> wrote:
> To elaborate on what Bob said, your post processor is really
already in
> Vector. When you insert a NC object you are really specifying the
post to
> use. You can also go in and modify the output of any of the NC
objects and
> rename it for your customized post. Fred has available
preconfigured NC
> files (posts) for EMC and CNC Pro that I know of in addition to a
myriad of

I'm going to jump in here and say a word or two about post
processors. By the way John S., you are TOO kind. Make sure to let
me know when you want to do some 3D stuff ;-)

Post processors were originally seperate programs that were designed
to eat the cl or cutter language files produced by APT programming
systems. APT is a numerical, geometric, description language that
has been pretty much replaced by Cad-Cam in most cases. It is STILL
a very good way to do 5 axis work, and is often found in aircraft and
aerospace applications. The CL files are a standard format to
describe a tool & its movement. It was input to a post processor
program that would format & correctly produce the G-code required for
any particular machine. You could simple repost with a different
post processor to have a program for a different brand of controller.

The thing is that most of the expensive cad-cam systems trace their
roots back to some common code that today would be called "open
source", kind of like EMC is. I have seen a couple of efforts to get
access to this code, but apparently it is written in fortran and
stored on ancient DEC PDP8 or 11 tape format(or worse). It is also
still probably covered by US export regs that prohibit export of any
Cad-Cam or machine tools that exceed 3 axes. Maybe not since EMC is
now spreading all over with 6 on a hexapod.

In the early days of Bobcad(say DOS version 10 or 12), it had an
internal format that was written to a temporary file. It was very
similar to the CL files that were produced by apt. I never asked
Bob, but he was at MIT about the time or right after, some of this
stuff was happening, and may have had direct access to the code, for
sure he had direct access to APT.

Today both Bobcad and Vector "post" directly from the drawing
database to the NC code. There are no intermediate files that are
available to post differently for one machine or another. You just
regenerate your G-code, or rather CNC code file with a
different "filter". Filter is probably a better way to describe the
process. remember in the "old" days 4K of ram memory was a lot &
everything was processed sequentially from one deck of punched cards
to the next, or to mag tape reels. The 4 mhz IBM XT computer was a
technological wonder, and it wasn't available until about 10-15 years
after APT came out.

In Bobcad the drawing and the CNC file are seperate files, in Vector
they are stored in the same OLE container as a unit with multiple
drawing and NC objects possible.

The "posts" or "filters" in Vector have been set up to provide as
direct access as possible to all the commands that one would normally
use to program any of the low cost controllers. At last count I had
completed about 10-12 different ones. Since the drilling cycles are
configurable in Vector, I have customized them to each machine, the
later ones all interactively ask for the necessary parameters. If
the controller supports looping, macros, variables, and other kinds
of programming constructs besides standard G-code, I have tried to
place those in custom blocks so that it is easier to learn the
language of your machine, and if you can't remember what the code for
dwell is, you click dwell & it says "Please enter how long to
dwell?", and tells you to enter the necessary value in seconds, ms,
or whatever units your controller requires.

If cutter comp is possible there are G41 & G42 selections that will
automatically turn cuttercomp on and off, if you use approach &
departs. If it's not available for that machine I have removed it
from the machine specific configuration. I have spent probably at
least a half to a full day on the documentation that I could get for
each machine & developed the posts based on that kind of study, & how
I would program a CNC machine. Not all use G-code. Some machines
can understand HPGL or 3D DXF. Vector will also "automatically"
output ordered geometry in these kinds of files so that you can
control the order and directin of the action in file formats besides
g-code.

We also have a public user forum with 3 years of Q&A, text string
searchable, & available 24/7/365.

Best Regards, Fred Smith- IMService
Listserve Special discounts and offers are at:
http://209.69.202.197/cadcamedmdro.html

imserv@... Voice:248-486-3600 or 800-386-1670 Fax: 248-486-
3698

Discussion Thread

Jerry Kimberlin 2001-05-13 18:30:04 UTC Postprocessor Bob Campbell 2001-05-13 19:18:09 UTC Re: [CAD_CAM_EDM_DRO] Postprocessor Tim Goldstein 2001-05-13 19:50:02 UTC RE: [CAD_CAM_EDM_DRO] Postprocessor Fred Smith 2001-05-13 21:10:39 UTC Re: Postprocessor Tim Goldstein 2001-05-13 22:36:36 UTC Taper Lock bushed pulleys machines@n... 2001-05-14 00:10:38 UTC Re: Taper Lock bushed pulleys machines@n... 2001-05-14 00:13:41 UTC Re: Postprocessor Smoke 2001-05-14 21:39:49 UTC Re: [CAD_CAM_EDM_DRO] Taper Lock bushed pulleys wanliker@a... 2001-05-14 21:59:32 UTC Re: [CAD_CAM_EDM_DRO] Taper Lock bushed pulleys machines@n... 2001-05-15 00:20:15 UTC Re: Taper Lock bushed pulleys Sven Peter, TAD S.A. 2001-05-15 04:49:39 UTC Re: [CAD_CAM_EDM_DRO] Taper Lock bushed pulleys timg@k... 2001-05-15 12:42:02 UTC Re: Taper Lock bushed pulleys Jerry Kimberlin 2001-05-17 18:16:51 UTC Re: [CAD_CAM_EDM_DRO] Re: Postprocessor Ray 2001-05-18 06:52:27 UTC Re: Re: Re: Postprocessor Jerry Kimberlin 2001-05-18 07:24:50 UTC Re: [CAD_CAM_EDM_DRO] Re: Re: Re: Postprocessor dougrasmussen@c... 2001-05-18 07:29:55 UTC Re: Postprocessor machines@n... 2001-05-18 08:52:01 UTC Re: Re: Postprocessor Alan Marconett KM6VV 2001-05-18 18:39:58 UTC Re: Postprocessor ballendo@y... 2001-05-21 18:34:54 UTC Re: Postprocessor