U1 (74S288) and U2 (74S174) together constitute a
"finite state machine" (so
they told to me cause I'm not so expert) whose purpose is to provide a read
A finite state machine (sometiems shortend to a 'state machine') is a way
of designing and building sequential logic circuits -- that is circuits
where the outputs(s) depend not only on the inputs now, but also on
previous sets of inputs.
The basic idea is tha tthe circuit stores internally a 'state', normally
as an n-bit binary number in n D-tyoe flip-flopes (that's the 74LS174 in
your circuit, a hex D-type flip-flop). All flip-flops are clocked at the
same time to change to a new state, Andthen there is a block of
combinatorial logic (a cricuit where the outputs depend only o nthe
inouts now, not what they were) which has as inputs the external inptus
ot the state machine and the outputs of the D type flip-flops and which
produces the inputs to the D-type flip-flops to give the new state.
Sievn a ROM is a gnneral combinatorial circuit, you can make a stte
machine from a ROM and a set of D-type flip-flops, aas here.
Who wrote that article decided to build itself a FDC
controller card for his
xerox 820-II extracting a block of its schematic from a Kaypro-II's
schematic really making very few changes and using a FDC WD1793 + WD9216 for
external separator data.
Due to its small number of components, and having already a WD1797 I'd just
need of a WD9216 (i hope to find it on the Internet yet) and wiring it would
be maybe more simple.
The 9216 was a very common data separator about 25 eyars ago. It was used
on a lot of PC/XT cloen FDC cards (not on the original IBM one though)
and in other machines too. I am pretty sure it was made by several
manufacturers.
Last suggestion (always from the so expert Chuck that
I thanks so much)
would be to use a WD2797 instead of WD1797, but I'm not a designer and while
i can understand something of all those things it's very hard for me to
change schematics or adjust components.
The 2979 is essential a 1797 with an intenral data separator. It works
well, but IIRC there are 3 analogue adjustments associated with it (a
couple of puse widths and the free-running frequency of a VCO). It's not
hard to set these up, but you do need a 'socpe or logic anlyser to do it.
The 9216 data separator IIRC needs no external adjustments.
-tony