Technically in school !:)

Message: 2
Date: Tue, 02 Nov 1999 02:51:10 +0100
From: "Arne Chr.Jorgensen" <instel@...>
Subject: Technically in school !:)

Hi,

Okay, just a few words Andrew:

Well, the schematics exist , you have to try to get hold of someone.
Some must have worked with these, and you know where you got the Beast,
- so you just have to do some detective work.

[The guy I got it from purchased it used, at an auction, and I got
everything he got. The company that made the computer inside- IIC- has
vanished from the face of the earth. And I don't even know if they ever
distributed the schematics for the computer itself- they might have
considered this proprietary information. But if anybody out there's got it,
by all means send it along...]

Here is a simple story about computers:

Most electronics is dumb circuits, but they do a specific job. With the
microprocessor and some code, this is different. In a normal PC, the
cpu can shut it self off from the main board, and let other processors
work with it. Another well known way it can do things, is to give
control of the bus system to a direct memory access unit. (DMA)
This is a software way of doing things, a "hardware" way of doing it,
would be to force it into some of these states. A "real hardware" way
of doing it, could be to remove the CPU, have another micro controller
put in it's place.

[I'd like to be able to do this- preserve the existing I/O stuff, but
replace the tiny brain with a big one. But I'm afraid there would be a lot
of compatibility issues I'm not prepared to deal with.]

Then you could disassemble the software in the EPROM's, and write a new
software for the original CPU.

[Right. Somebody might be able to do this- but it's certainly not me.]

Here is several ways to do things.

Here is a short story of what the beast could do:

CPU, please check if there is any data from the RS232.
<There is >
Please read a string and place it in memory.
< Ok >
Split the string into data for the X,Y,Z axis, and the feed rate.
<OK>
For the X axis, store this data in the position registers, do the same
with the other axis.
< OK>
Place the feedrate in the DAC register.
<OK>
Now if <single step is pushed > then write to the enable port, and
output the command to the servos.

[Thanks for laying it out this way- it does make things a bit clearer.]

This little story says a few things:

First the local CPU, is the one that will read any data from an external
port,
then this one will know what to do, and where to store these data. All
these goes to output ports, like position registers,
DACs, screen, what ever. The main board in your machine have this bus,
and addresses, to all of these registers. It is just like the ServoToGo
card, but everything is already wired up, and the signals match, etc.

You could use a simple IO board in a CPU, change the addresses that EMC
has in the .ini files or wrapper code for the Stg boars, and talk
directly to your main board in the Beast. This is something that could
be done, but you need someone that knows how to do a task like this. (
Here is where I think some students could do some work, they would learn
a lot about EMC and control systems, and dig into the hardware of
things.

[Sorry if I sounded a bit sour on the subject of our "Technical" high
schools- actually we have community colleges here that function in much the
same way as European technical highs, and I might be able to get some help
from one if I did some research and found out who to ask.]

This would be a very valuable work for a lot of people, if they
made the work public. This kind of retrofit could be done on all kinds
of systems. Okay some will reject to this, and say that the DACs on your
machine is just 8bit wide, or they say something else. The thing is
that you will get a system as good as the original one, - if that has
worked, so will this. To get a better result, than what was in the
original machine, - is just wasted energy in many cases. I really need a
lot of knowledge to do that. Let take a Mercedes Benz, and make it into
a hot street car. Add a root compressor stage, a turbo stage, a
sharper cam, wide tires, etc. How many do make a better car out of it,
and at what cost ?
90 % of all attempts, - is just a ruined good car. You really need a
lot of knowledge to make a success out of it. ( Well, if you want to
race Daytona 500, and have the money for it.... :)

Before I would cut any wires, I would evaluate what I got, what options
and to what cost.

[Believe me, I've been doing this. But it's looking like some kind of brain
transplant is required if this is going to do what I want it to- execute
large NC files. At least it has its servos and ball-screws, unlike the
first machine I was considering.]

Or ship the piece to one of the others on this list,
that make a living out of retrofitting.

[It seems like most are fairly far off- and this mill is extremely heavy.
Having gone through moving it once, I'm in no hurry to repeat the
experience. ]

Have a talk with them.

[I've been doing this- talk is cheap, freight is expensive...]

The only
thing I attempted to do, was that you gave it some thought before you
start cutting wires. Have a chat with Jon, he knows what it takes. The
thing is that we often underestimate what it takes.

Best wishes

//ARNE

[I think I might be past the point of underestimating the difficulty of
this project, although I could always be wrong. It seemed like I was
starting out ahead of Jon and his more extensive Bridgeport retrofit. But
if I had it to do over again, I'd look for a mill with the ability to run
directly from tape, in which case a BTR board would work to drip-feed the
code into the machine.]

Thanks-

Andrew


Andrew Werby - United Artworks
Sculpture, Jewelry, and Other Art Stuff
http://unitedartworks.com