CAD CAM EDM DRO - Yahoo Group Archive

Re: emc (TOME)

Posted by mike grady
on 1999-10-12 18:11:07 UTC
did you see ron gingers post from quite awhile back
he has a interactive system he cloned from the acu-rite
millpower
he said he would be interested in doing a group project
porting it to linux
i quoted it below
mike

Since Ive poked into this list a bit maybe I ought to do a bit
of an
introduction.

Im an Electrical Enigneer, but with a very old degree- 1966-
they still
did tubes when I was in school, transistors were new. For 30
yearss Ive
been in the computer world (DEC) mostly in software
information systems.
For the last 10 years Ive been a Unix Admin Consultant. Ive
been a hobby
machinist for 10 or 12 years. See
http://www.ultranet.com/~ginger for a
bit more.

My CNC experiements started 4 or 5 years ago when someone on
the net
offered a set of stepper motors to fit a Sherline lathe and I
bought
them. For a driver board I bought a 3 axis board from AMSI in
New York
($150). For software I bought INDEXER.LPT($250).

After I got it all to work, and I spent a few 'gee whiz' hours
typing
'move a,250,b,300' and watching the lathe slides move back and
forth, I
realized the real problem here (as in the entire computer
world) is
SOFTWARE. I never did make a usefull part with this setup.

I muddled around, reading a lot, but not doing much (at least
with CNC-
thats when my steamboat got done, among other things), until a
friend
bought a Lagun mill with an AcuRite MillPower system. After
just minutes
of waching it, I was off onto a new project. I copied its
owners manual,
and started a clone of the program using Visual Basic. Before
the
AcuRite lawyers jump all over me, I will note my program is a
good deal
different, since I use Win95, a mouse and keyoard and all the
normal
Win95 GUI actions, instead of the AcuRite special box and
function
keypad. I have a similar concept, not a chineese copy.

In case you dont know about MillPower a brief bit will be
usefull. This
is billed as an interactive system, and gcodes are not used in
any way.
A set of buttons is labeled with common mill functions, like
LINE, ARC,
CIRCLE, BOLT CIRCLE, etc. You pick a button, fill in the
obvious blanks,
and a program is built to run that object. When youve got a
whole
program ready the system runs it. You never see a gcode.

I have written just about all of these fnctions in VB. I first
planned
to use INDEXER.LPT for the driver, but a friend wanted a
program to
drill PC boards, so he wrote a .dll that we can call from VB
that runs
the motors. It implements bresenhans algorithym for straight
line moves,
and circular interpolation. I call it from VB with a simple
struct that
gives it the distance to move each axis. We do acceleration
and
decellaration and the use of the bits is fully software
configurable.

I modified a Sherline Mill (see my web page for photos) in a
somewhat
unsual way- I noted that each lead screw has 2 ends, one has a
handle on
it, the other flaps around in the breeze. For reasons I dont
understand,
everyone seems to want to modify the handle end, making stange
standoff
brackets and shaft couplers and double shaft motors and lots
of cruff,
leaving the other end flapping. I put my motor on the loose
end, and
dont have to modify the handle end at all. I simply made one
mounting
block and one long shaft coupling on each screw. I use a belt
on the
Zaxis. I can also use the handles if I leave the power off,
and I can
use single ended motors, often cheaper. For this system I
bought one of
Dans 2Amp driver kits.

I have made some useful parts with this, and Ive added a
couple features
to it- I can output a MillPower file on floppy, and I can read
g-code
in. As a by product I can read in gcode and write it out as
Mill Power,
something maybe valuable to Mill Power owners. I can also read
the
standard PC board drill file (excelan?) and run that. I also
can mill
TrueType fonts directly, altough a bit lumpy. Im still adding
a geometry
calculator and a learn mode.

My delima here is the .dll. Because of the way Win95 works,
once I call
the .dll to make a motor move, my VB routine is dead until the
move is
complete. I cant abort it, and I cant update my display as the
tool
moves. I dont think a CNC system is usefull with this limit.

To get around this I tried some experiements with
microprocessor
controllers. I used a Tiny Tiger ($79) which I can load from a
serial
port and use to drive the step and direction signals. Ive
written
bresenhans line stuff there, and it works slick. But if I want
to sell
my package, Id have to get into the hardware business of
making these
boards.

so, that brings me to EMC and linux. Id like to port my VB GUI
to linux
(maybe Tk, maybe Java) and use the EMC API to actualy run the
motors. I
think I can write a perl hack that will read the VB source
code and
generate the Tk functions. Then all I have to do is rewrite
the
callbacks from BASIC to someting on linux.

So, if anyone has read this long winded note, thats what Im up
to, and
thats why Im asking about EMC-APIs and such. If someone else
is
interested in this software development, maybe we can take a
page from
Linux and do a group project?

I suspect someone will ask, so, no, I have not decided to sell
or give
away my VB program. I dont want to sell it with the .dll
weakness, and I
dont want to give it away yet, because someday I might decide
to fix the
.dll into a win32 .vxd and sell it.
Jon Anderson wrote:

> From: Jon Anderson <janders@...>
>
> Gar,
>
> Very well put. Ought to say something that with all the commercial
> products out there, folks are willing to spend the time to learn
> something new, and make something that is still evolving, work for them.
> Lots more pride and satisfaction in that, than popping in and disk and
> typing a:install
> I have been reading about Hoyt's software in various places for some
> time and even inquired about the demo. However I'm at the commercial end
> of the spectrum here, time IS money, and I want the most power and speed
> I can get for my limited bucks. EMC and scrounged servos offer that, and
> I'll dink with the software late at night until it's working. When I get
> to making chips, I want to be able to use the G-codes I know and the
> hundred or so programs I've already got. (for the record, I'm presently
> running Ah-ha)
>
> Something I would love to see, would be a PC or Linux based program that
> duplicated the fill-in-the-blanks approach of Proto-Trak and such, and
> generate standard G-code.
> This seems like a perfect brewing ground for input/ideas and beta
> testers, and I rather suspect there would be a significant commercial
> potential.
>
> Anyone?
>
> Jon
>
> Gar Willis wrote:
> >
> > From: garfield@... (Gar Willis)
> >
> > On Mon, 11 Oct 1999 20:53:46, batwings@... wrote:
> >
> > >>From: "Darrell Gehlsen" <darrell@...>
> > >>
> > >>Hoyts real question:
> > >>Why do you guys insist on using free software when you could be buying it
> > >>from me?
> >
> > >The questions was a real one. It wasn't motivated by a desire to sell
> > >anyone anything. I'm in here by direct invitation, by someone who knows the
> > >subject to be interesting to me. I participate in a number of forums and
> > >always offer best advice in them. You OTOH havn't as far as I can see
> > >offered any insights at all or even asked a reasonable question. Your only
> > >purpose was it looks like to flame me. The above will be the last note of
> > >yours that I ever have to decide to TRASH by hand, think about that. Goodbye!!
> >
> > Your choice, of course, but aside from the terse sarcasm (which EVERY
> > vendor in the US anyway has to learn to live with, and rightly so), I
> > have to side with the suspicions raised by the people in the fore on
> > this "Linux...why bother" thread yous guys started. In EACH case, the
> > proponents seem to be someone who's announced themselves as having
> > products to sell or plans thereof. And just to add fuel to the fire, any
> > time someone tells me they "participate in a number of forums and always
> > offer best advice", I wonder if that fellow doesn't have a rather
> > enlarged view of himself. Just a personal observation learned over time.
> > And by the way, in a public forum like this, what the hell do you mean
> > by saying you're "in here by direct invitation". If someone suggested
> > you join in, how does that make YOU some kind of higher lifeform? That
> > sentence and the one following it, lauding your own advice, say
> > mountains about your assumed self-importance and self-worth.
> >
> > When I read the "Linux/EMC is toooo hard" thread when it started, I kept
> > hearing this subliminal message something to the tune of "why are you
> > guys struggling with Linux & EMC, why not let the 'experts' do that for
> > you". Something like that. The one guy that suggested someone should
> > burn a universal CD that would work slam-dunk, which he'd be happy to
> > pay for?... I thot of the old phrase we used to use on our managers, "if
> > someone says to you, I want software I can just plug into any computer
> > and play without difficulty, hand them a lollipop". Heh.
> >
> > But enough of the sarcasm; let's get serious. I'll tell ya why I'm DEAD
> > SUSPICIOUS of anyone saying "Linux/EMC is too hard, let's let the
> > WindozMT (oops, meant NT :) & Commercial CAD/CAM vendors do it for us".
> > Because I think the BIGGEST thing underlying Linux AND EMC together in
> > spades is CONTROL of our destinies thru *accessible* KNOWLEDGE. With
> > Linux and EMC, all that's keeping any one of us from being in near
> > complete control of bug fixing and basically our OpSys and CAM
> > programming destiny, not to mention a free/great/deep education in the
> > process, is our own native interest and intelligence. NO FRIGGIN "trade
> > secrets" crap and boy's club barriers. THIS I consider nothing short of
> > a MAJOR revolution. Now certainly it doesn't TAKE that depth of plunge
> > to just USE these tools, but that's the depth of it's POTENTIAL.
> >
> > In the early days of computing (yeah, my first was an IBM 1620),
> > everyone I think actually believed there needed to be a Hitler to keep
> > everything compatible, and without the equivalent of the Gates of Hell
> > to keep everyone "inside", pandemonium would ensue. We should have had
> > more confidence in ourselves, frankly. What drives people to collaborate
> > and agree on standards (whether it be Linux or EMC) IS the desire for
> > ease of use and interchangeability. That's a NATURAL drive/urge; we
> > don't need some "Gawd" imposing their "benign dictators" will on us in
> > order to make that happen. Obviously Linux has (and will have) it's
> > genetic deviants, but has this caused the whole effort to dissolve into
> > undifferentiated pond slime? I trow not. Successful adaptation and the
> > coalescence of many able collaborators is what drives and feeds the main
> > artery.
> >
> > In sum, I thot Linux's emergence was miracle enough in this age of
> > magesterially authoritative 'entremanurial' dominance, but the existence
> > of such a stellar piece of work as EMC coupled up with Linux, is just
> > WAY too magnificent an opportunity of the first order in this just
> > barely emerging age of computerized mechanization, to let some piddling
> > concerns about a few problem parts configurations causing some tantrums,
> > get in the way. Good gawd, men, stand back and look at the FUTURE
> > potential of things like Linux and EMC. Have you no vision?
> >
> > I'll put it to ya simply; when you have the source code you're no longer
> > a slave to anyone/anything but your own sloth. And when a bunch of
> > people of similar notion collaborate together, such a voluntary
> > consortium, like this group as an example, has just *enormous*
> > potential.
> >
> > And if you complain there's currently too much LinuxEMC traffic, ignore
> > it and start some thread on what YOU think is so much more important.
> > But of course, if it happens that your pet subject get's little
> > response, I DO hope you'll be the gentleman and take a hint.
> >
> > Gar
> >
> > > Welcome to CAD_CAM_EDM_DRO@...,an unmoderated list for the discussion of shop built systems in the above catagories.
> > To Unsubscribe, read archives, change to or from digest.
> > Go to: http://www.onelist.com/isregistered.cgi
> > Log on, and you will go to Member Center, and you can make changes there.
> > For the FAQ, go to http://www.ktmarketing.com/faq.html
> > bill,
> > List Manager
>
> > Welcome to CAD_CAM_EDM_DRO@...,an unmoderated list for the discussion of shop built systems in the above catagories.
> To Unsubscribe, read archives, change to or from digest.
> Go to: http://www.onelist.com/isregistered.cgi
> Log on, and you will go to Member Center, and you can make changes there.
> For the FAQ, go to http://www.ktmarketing.com/faq.html
> bill,
> List Manager

Discussion Thread

garfield@x... 1999-10-12 15:51:03 UTC Re: emc (TOME) stratton@x... 1999-10-12 16:21:56 UTC Re: emc (TOME) Jon Anderson 1999-10-12 16:22:58 UTC Re: emc (TOME) batwings@x... 1999-10-12 07:34:53 UTC Re: emc (TOME) batwings@x... 1999-10-12 08:26:55 UTC Re: emc (TOME) garfield@x... 1999-10-12 17:59:55 UTC Re: emc (TOME) garfield@x... 1999-10-12 17:59:57 UTC Re: emc (TOME) mike grady 1999-10-12 18:11:07 UTC Re: emc (TOME) garfield@x... 1999-10-12 18:40:27 UTC Re: emc (TOME) batwings@x... 1999-10-12 11:03:57 UTC Re: emc (TOME) garfield@x... 1999-10-12 22:10:52 UTC Re: emc (TOME) PTENGIN@x... 1999-10-13 01:20:45 UTC Re: emc (TOME) Carlos Guillermo 1999-10-13 04:57:18 UTC RE: emc (TOME) Marshall Pharoah 1999-10-13 04:59:01 UTC Re: emc (TOME) batwings@x... 1999-10-12 20:55:12 UTC RE: emc (TOME) Ray Henry 1999-10-13 07:18:17 UTC Re: emc (TOME) Fred Proctor 1999-10-13 07:33:52 UTC Re: emc (TOME) Jon Anderson 1999-10-13 07:59:47 UTC Re: emc (TOME) Jon Anderson 1999-10-13 08:42:05 UTC Re: emc (TOME) batwings@x... 1999-10-13 02:06:23 UTC Re: emc (TOME) Jon Anderson 1999-10-13 12:35:07 UTC Re: emc (TOME) PTENGIN@x... 1999-10-13 14:48:56 UTC Re: emc (TOME) batwings@x... 1999-10-13 06:17:42 UTC Re: emc (TOME) garfield@x... 1999-10-13 17:04:22 UTC Re: emc (TOME) Darrell Gehlsen 1999-10-13 17:18:00 UTC Re: emc (TOME) Jon Anderson 1999-10-13 17:35:47 UTC Re: emc (TOME)