Alt-BEAM Archive
Message #09722
To: beam@sgiblab.sgi.com
From: "John A. deVries II" zozzles@lanl.gov
Date: Wed, 02 Feb 2000 16:46:52 -0700
Subject: [alt-beam] Re: Solarbotics Hextile Designs
At 11:45 AM 2/2/00 , Dave wrote:
>Solarbotics is getting close to starting active design work on our Hextile
>boards package.
>We're looking for your input on whether you want "Mark Tilden" (tm) style
>Hextile boards, exactly like the ones he uses, or if you would like to
>request different types ot Hextiles as well.
>Our initial run will be using DIP-style (regular sized) chips, on Hextiles
>and perhaps some square tiles too. SMT? Well, perhaps in the future....
Well, since it isn't likely that you'll be producing the SMT Bicore 2.x
style "chips" (can you imagine how expensive THAT would be???) I would suggest:
a) Don't copy Tilden's boards. They aren't sufficiently general even
though they seemed to be when he first drafted them -- for one thing, there
aren't usually enough holes for connections to a given pin/signal unless
what you REALLY want to build is a microcore with a pulse neutralizer or a
hexcore.
b) Make a board that can easily be wired as bicores or individual Nv/Nu
neurons.
c) Do make a board like his power conditioner/startup signal chip.
d) It might be a good idea to ignore the '14 IC. I expect some people will
disagree.
e) And the real toughy -- come up with a layout that requires a minimum of
manual wiring. I don't include required resistors and capacitors in this
category -- just plain old from-point-to-point wiring.
Just my 1 3/4's cents worth,
Zoz
Home