Dave McGuire wrote:
On Feb 6, 2007, at 11:48 PM, Richard wrote:
I did a bunch of that too, on a supercomputer project at P.U...it
was lots of fun. :)
-Dave
Our company, Irvine Computer Corp used a lot of wire wrap, and had
two machines. We started out making mostly disk controllers for SMD
in WW, but also did cpu's and some other parts of our system.
Our cpu was an enhanced Microdata 1600, and ran with a 20mhz clock
module, divided down to a 10mhz micro machine rate. There were a
lot of naysayers who said that the clock rate was too high, but we never
had any major problems with the clock for that reason.
The 1600 design originally used a poly phase clocking (not my term
but what I recall it being described as) which used an offset derived
pair of clocks (of by 90 degrees) that were 10mhz. Then the two
clocks were run all over the place, and different things happened
at different times depending on what had to be done, with each
micro instruction being completed every 200 ns.
The implementation we had was clocked every 100ns and all
logic ran off the same clock. where the 1600 was a 4 tick
cpu, we had a single tick.
any other observations on WW as far as clocking? I know that
the more expensive stuff ran mini coax around the backplane
to get much better signals, but we just used regular old 32 or
whatever guage WW wires, sometimes 26 guage for high
current.
We also had a large amount of decoupleing built into our basic
WW design boards that was put in before any wires were
wrapped, so there was very little but IC pins with wire wrap
signals going to them. The power and ground grid was all
done in the PC carrier board, and we ran 5 and ground
everywhere that way. I know some logic probably
had other voltages that made that harder to do as far
as power.
jim