Newsgroups: comp.robotics
Path: brunix!uunet!decwrl!world!gkulosa
From: gkulosa@world.std.com (Greg A Kulosa)
Subject: Boards for robot control (Was: Re: gcc cross compiler for the 6811?)
Message-ID: <Bvt4Ao.54x@world.std.com>
Organization: The World Public Access UNIX, Brookline, MA
References: <ld4tb4INN9eq@exodus.Eng.Sun.COM> <Dy90RB1w165w@beltrix.guild.org>
Date: Thu, 8 Oct 1992 14:30:21 GMT
Lines: 66

peters@beltrix.guild.org (Peter Sleggs) writes:

>cmcmanis@pepper.Eng.Sun.COM (Chuck McManis) writes:

>> In article <72y3RB4w165w@beltrix.guild.org> (Peter Sleggs) writes:
>> >I have most of what you want designed already [ except for the Memory -
>> >8Kx8 was the affordable part then (1989) ]
>> >peter
>> 
>> Well there is designed and there is "I'll sell you a PC board, documentation,
>> and some source code." I've got my own Z80 design that while its low power
>> and has good expansion options, is not nearly as impressive as what you
>> can do with one of the high integration controllers like the 'HC11.

>If there was a demand I would run a new board with newer memory devices
>and the like, however having seen a number of other boards out there I
>was not going to jump into the fray, unless there was something special
>to provide a distinguishing feature.

What is wrong with the 6.270 design by Fred Martin?  I don't know if
any of you have read the docs, but it definetely deserves some
attention.  It uses a 68HC11, has 32k of battery backed RAM, something
like 16 or 24 analog inputs, controls 6 motors, has a small LCD screen
for user interaction (along with two buttons and a knob that you can
use to make menu choices, calibrate, etc. with).  It also has a
prototyping area with access to the 6811 bus.  

The REALLY nice part of the system, however, is the Interactive-C that
Fred and Randy Sargent designed for the board.  It hooks to the board
over an RS-232 line from a PC or unix machine and allows you to
execute C commands INTERACTIVELY with the board.  If you type (For
example):

motor(4,on);

the Interactive-C will compile, download, and execute the command and
immediately turn motor 4 on.  You can see how useful this would be for
troubleshooting, debuging, and just plain prototyping of code and
hardware.  IC also includes libraries which have functions for
dealing with Servos (The ones from Radio-control models), infrared
sensors, PWM routines for the 6 motors, circuitry to tell if the main
4 motors are stalled, and a few more that I can't remember.  IC is
also a multi-tasking environment.  You can run C functions in the
background with different time-slice values, and communicate with them
through global variables.  You could , for example, run a routine that
checks sensors for a collision, and then notifies the main routine to
change course.  It should be rather easy to develop a
subsumption-style robot using multitasking functions cooperating in
some manner. (I know this is something that I would like to try).

Please note that this has all been from memory (My docs are at home),
so I may have left out some important features, or may have gotten
some stuff slightly wrong.  In any case, FTP the docs from
cher.media.mit.edu and look into this design.

Also, please don't go bothering Fred about this with lots of E-mail, I
know he is pretty busy supporting the MiniBoard, and he will get mad
at me if I give him more to do :^)


Disclaimer:  I have no direct connection with MIT or Fred Martin.  I
am just envious of their designs.

-- 
Greg A. Kulosa
gkulosa@world.std.com
