13 Oct
2007
13 Oct
'07
7:42 p.m.
Ethan Dicks wrote:
I think one essentially has two choices:
- find a 4008 & 9 to break out the bus and then you can use standard
8-bit EPROMS (4004 instructions are 8-bits actually). You may still
be screwed for RAM without a 4002 though, as the 4004 is Harvard arch.
and I'm not sure that the 4008/9 bring out the data address space.
- (re)design some bus muxing logic (4000 series CMOS should interwork)
to deal with the 4004 cycle and use 4 or 8-bit (E)PROMs and RAM.
In other words, recreate the 4001 and 4002 in SSI (or 4008/9 depending
on your perspective).
(For 4004 stuff it's nice to find equipment with at least the entire digital
portion intact, precisely to avoid having to do the above.)
(And I forgot to mention the 4201 clock generator (still looking for one)).
From time to
time, I consider building a small 4004 board - nothing
more complicated than a
digital clock, but my lack of examples to
study in detail always slows me down. I've read over various
datasheets, but not recently enough to really recall any specifics. I
happen to have a couple of 4004s already - one plastic, one ceramic
package, but they came from devices that I didn't have schematics for
(one digital kitchen scale, one early barcode reader), and don't have
the original PCBs from to study.
I can't say that a modern 4004 project would be practical, but it sure
would be neat to watch. I know there were some older bi-polar 4-bit
PROMs that should still be programmable with 25-year-old programmers,
but I suppose that it's just easy enough to use modern 8-bit devices
and ignore 4 of the bits.