which explains why SHIFT is only partly regular in its behaviour and
why CTL is highly irregular – as a little time with an ACSII table and
the Nascom keyboard will show. SHIFT, CTL and GR (graphics) can be
separately identified as depressed in conjunction with any other
The input of data from the keyboard is entirely software
dependent and the computer readily can (and does) get into a state
in which no input from the keyboard is possible. Control can always
be regained by Reset which acts in a totally different way. It puts
an earth on the Z80 reset input, via PL3/12, and forces a reset to
whatever address is set up on LSW1/1 to 4.
The above table will help to localize a keyboard fault (or
a fault around IC25 on the main board) which, as in my case, affects
the whole of one column or the whole of one row of the table.
NASCOM HARDWARE MODIFICATION
A. O. ZIENKIEWICZ
MODIFICATION TO ‘TYPE A’ RAM BOARD
This enables the use of 2716 type EPROMs in the four EPROM sockets
allowing 8K of firmware to be held in memory.
|Cut tracks near:–|
|IC 24 p2|
|IC 24 p3|
|IC 27 p21|
|IC 27 p19|
|IC 24 p2||to||A11|
|IC 24 p3||to||A12 (At hole near Bus line 42)|
|IC 28 p19||to||A10|
|IC 27 p21||to||IC 27 p24|
|P5||to||Two pads as for WIRED-OR RAM select.|
(i.e. for C0000-DFFF link P5 to 9+10)
Note that if the selection is to an odd ‘000 the sockets will not be
in consecutive order.
To use 2732 type EPROMs for a total of 16K firmware, cut tracks
as above,then cut tracks near p20 for each socket,
|IC 28 p19 to A10|
|IC 28 p21 to A11|
|IC 24 p16 to IC 24 p2+p3|
The Chip-select (p20) of each socket is now wire linked directly
to an appropriate 4K decode pad.