Relay computers

[John Lawson]

On Fri, 24 Sep 2004, Dwight K. Elvey wrote:

>> From: "Ade Vickers" <avickers at solutionengineers.com>
>>
>> At 22:50 23/09/2004, you wrote:
>>
>> Cheers,
>> Ade.
>
> Hi Ade
> I think your counts are way off. As I recall, I did a schematic
> a while back for a single bit of an alu and it only took about
> 4 or 5 relays. I did use DC and diodes for steering.



   I also think that these relay counts are a bit excessive.  We are 
forgetting the huge amount of wisdom available on relay logic design, 
starting mainly with AT&T and the Bell System, and many other authors who 
wrote on this subect from the 20s thru the 60s.  There are several 
branches of mathematics and combinatorial logic devoted to the analysis 
and optimization of relay logic systems - which of course is being rapidly 
forgotten today for obvious reasons.

   That makes these books cheap / free :

Engineer's Relay Handbook, various authors (Hayden)

Switching Circuits and Logical Design,  Caldwell (Wiley)

Design of Switching Systems, Ritchie* (Bell System)

and those are just the three cluttering up my desk - there are probably 10 
more on the shelves.

   Point being - the design of a modern Relay Computer must start with the 
Boolean design and equations, to which are then applied the well-known and 
well-documented principles of Optimization as they concern 
electro-mechanical logic.  Then the practical (also well-documented) 
aspects of: speed, timing, current draw, contact protection, etc.  Layout 
and wiring optimization come next...  not the least being: when it breaks, 
can you find it and fix it?


  Cheers

John

Note: This book is by the "C" language author's Father....