Doctor Dark’s Diary – Colour Supplement Edition.
The MAP 80 RAM board.
It was getting to be about time someone brought out a new 80-BUS board
that would astound the plastic box brigade, and suddenly there are two around,
both containing massive amounts of memory. The MAP 80 board is first on the
agenda, simply to make you wonder what the other one is until you get to the
appropriate page, although I have put a really subtle clue in the heading. You
have seen the advertisements, so you know that the MAP 80 board can hold up to
256K of memory. You may even have seen a hardware review somewhere, that went on
about the sheer amazing blueness of the board. So I am going to write about how I
went about putting mine together, and the software that is supplied for it. Mind
you it is very, very blue...
First of all, being the adventurous type, I ordered the bare printed
circuit board, which costs £42-50. The day after I sent the order, I was
telephoned by MAP 80, who wanted to make sure that I was not trying to use the
board with a Nascom 1. Apparently, this is a combination that does not work! So,
I reassured them that I was now using a Nascom 2, and the board arrived very soon
after. The board is of a high standard, nice solid glass fibre, with the blue
solder resist, and the usual silk screening to show where things should be put.
Things are pretty closely packed, as you would expect, since there are fifty four
DIL sockets to fit on the board. With the board were the parts list and
construction instructions. I ordered all the sockets, resistors, capacitors and
most of the TIL chips from Maplin, who managed their usual rapid response. A
couple of the chips needed are not mentioned in the Maplin catalogue, but can be
obtained from other firms, such as Watford Electronics, or Technomatic. When I
unpacked the sockets, I was surprised to find that they were made from a rather
nice shade of pale blue plastic, instead of the usual black. When these had been
soldered to the blue pcb, the result looked like something from another
dimension, after years of looking at green boards with black sockets! Cost of
components (excluding PCB and RAMs) was under £20, just! The next step, which I
am still saving up for, is to buy the thirty two RAM chips the board can carry.
As these are about £4-50 each, the 256K board can be built for just over £200,
which is a lot cheaper than people with S100 based computers seem to pay for
their boards. The chips have now begun the gradual slide down to a more sensible
price, so it is nice to know that the board will work with just a single row to
start with, as long as you don’t connect up the chip select lines for rows that
are not there.
There may well be murmurs of “What on earth can you use it for?” in the
ranks. At first, I used to find a 1K Nascom 1 big enough. Now my programs are
much bigger, but there is as yet little probability that I will overflow the 64K
mark! No, the magic phrase that made me so keen to get the board was “virtual
disk”. [Ed. – see R. Beal’s article elsewhere on VDs (oops!).] And you don’t have
to mess around modifying CP/M, because MAP 80 will do it for you. They have done
this for me, and the disk has been returned with two new versions of MOVCPM on
it, one for the usual display and one for use if you have an IVC card on your
system. The source code for the revised BIOS is supplied, which is a very
sensible approach, enabling even further customisation of the software without
any need to guess what their bit is doing. [Ed. – I wonder how Gemini feel about
other manufacturers modifying their CP/M BIOSs and sending out source listings!
Naughty!] No doubt, sooner or later, I will have time to have a look how it
works! The CP/M works in the usual way when there is 64K or less of memory
connected, but when there is more the sign on message will announce that the
virtual disk is available, and give its size, according to the documentation.
Must order some RAM chips... As a bonus, my “old” 64K, GM802, card is not made
redundant by all this. The instructions supplied with the MAP 256 tell you how to
fix it so that the software will use it as well. Presumably, the effect of the
modifications is, in part, to change the GM802’s paging control port to £FE, from