13 Sep
1999
13 Sep
'99
8:50 a.m.
Almost all of the early HP PMOS calculator chips were
fabbed by
either Mostek and AMI. They are custom for HP, so don't bother
looking in catalogs.
:)
> From first look the CPU seams to hafe a serial
architecture.
Yes. There is a two-phase clock and synchronization
signal to
denote that start of a 56-clock cycle. There is a data line which
is used to transfer a 56-bit word to or from memory and peripherals
each cycle, although most cycles it is not actually used. There
is a line that the CPU uses to shift out an 8-bit instruction address
to ROM (addressing beyond 256 words is by bank-switching), and a line
that the ROM uses to send back a 10-bit instruction.
For more details, check out my web site:
http://www.brouhaha.com/~eric/hpcalc/
Bookmarked
I've written a complete microinstruction-level
simulator of the HP-45
which is available there.
Well, interesting.
So if I get your decription right (last mail), the 1818 0060 10 Pin
Mostek MK6068P is a 1 kBit ROM and the 1818 0058/9 (MK6107/8) are RAM
chips (Which I assumed, since the board layout looks _way_ like it.
But why two different Chips ? Already including an preset address
comperator ? Each chip can keep 8 'words' (I havn't 'decoded' the
simulator right now - Maybe you can recover your description pages ?)
Assuming that there are only X,Y,Z,T, STO 0-9 and 2 calcualtion registers ?
Also, following the Board layout, there are 4 more locations for RAM - could
it be that THe calculator can be extended to 32 more registers ? Just a flow
of free ideas: the design was already redy to become a programmable
calculator ? The 65 was just in planing I asume (BTW, my unit seams to be
from early '74 - since all time stamps are between
73/31 and 74/04)
I already love this beast.
And Yes, comparing to an HP45 picture I found
http://www.hpmuseum.org/45intern.jpg
The chips are basicly the same (I'm missing the round ROM).
Gruss
H.
P.S.: Has anybody a HP16C for sale ?
--
Der Kopf ist auch nur ein Auswuchs wie der kleine Zeh.
H.Achternbusch
