William Maddox wrote:
I recently picked up an interesting book, entitled
"The Logic Desgign of
Computers: An Introduction", by M. Paul Chinitz. Mr. Chinitz was
Director of Training at Univac from 1953 to 1956, and it shows -- even
though this book was published in 1981. Seriously, the design style in
the book is straight out of the 50's and 60's. Although the book makes
numerous references to standard TTL logic components, he does just about
everything with simple gates and RS flip-flops, and devotes space to
such dated topics (for a book of its scope) as 1's complement
arithmetic, serial arithmetic, and delay lines. Proponents of a
straightforward fully-synchronous design style based on MSI, like
Winkel, Prosser, and Mano, would be horrified to see students taught to
gate clocks in 1981. The book uses a simple 8-bit 1-address
architecture as an example, using a multiphase clocking scheme and RS
flip-flops with preclear. All a wonderful throwback to the days when
every gate and transistor counted.
Don't forget core memory and early 1024x1 Dynamic rams. :)
Fun to interface.
I picked up a copy at a swap meet, misplaced it,
ordered another from an
internet used bookseller, and then found my original copy. I thus now
have an extra copy of this book. It is available for $5 plus shipping if
anyone is interested.
--Bill
Design - Design - Design
I think I'll go for the BIG CPU DESIGN --- A simple 24 bit CPU.
So I will need the book, 14 74172's and 8 74F382's.
Now if I could only drop the ALU logic since I have alot blinking
lights and strange data paths. Since my goal is about PDP-8 speeds
I don't think I will have a problem using LS parts.
Ben.