Mark-8 Computer - Alive and Ticking!

From: John Lewczyk <jlewczyk_at_his.com>
Date: Thu Apr 8 23:49:36 1999

Late in Februrary I pulled out of a corner of my workshop a box containing
my old Mark-8 computer, the one whose construction was the subject of the
cover of Radio Electronics magazine in July 1974. I hadn't looked at in
over 22 years! Ah, the memories...


When I first read the RE issue, I was in no position financially to purchase
anything but the plans ($5 and a SASE, IIRC), being a poor college student
(of computer science). It was not until the fall of 1976 that I was able to
buy an already constructed Mark-8 from someone who was selling it because he
was upgrading (buying a newer microcomputer). I don't know who the original
owner was, as the computer was bought by a third party for me.

It was astounding to me that I owned my own computer! I proceeded to
experiment with it for weeks, hand assembling programs and then keying them
in through the front panel! But entering the code through the front panel
switches got old real quick, and I endevored to design and build an "octal
keyboard and display" for it. The original "backplane" made working on any
individual board a real problem: you had to "fan out" the boards to get at
the inside ones. So I attempted to make a backplane out of 100-pin wire
wrap sockets and bolting edge connectors removed from surplus boards onto
the edges of the six Mark-8 circuit boards. After all the work of rewiring
the boards with their new edge connectors and wiring the backplane, and then
building a new octal front panel board, the computer didn't work at all.
(Surprise! ;-) After some time trying to debug it with just a VOM, I
abandoned the whole project because I did not have access to the tools
(triggered 'scope, logic analyzer, etc.) that I realized I would need to
reasonably debug it. Of course there were some great new microcomputers
available that were faster and easier to use and program. I bought a KIM-1
and never looked back.

Until now.


So, I have been attempting to restore the Mark-8 to as close to original
condition as possible, which of course includes making it work! ;-). I am
pleased to announce that I've been successfull and I'm again the proud owner
of a working, .5MHZ computing colossus! I've removed the attempt at a
"backplane" and restored it to as close as possible to the origional
condition.


Originally the computer came with 1K ram, composed of 32 * 256 bit sram
chips (Intel 1101A and equivilents). I was dissapointed when I noticed that
1 of the ram chips was missing (I believe it failed, and I threw it out and
never replaced it). After getting the system to run with a minimum of 256
bytes, I used memory test routines to check out the remaining memories. I
found many more failures, which leaves me with 26 working memory chips. I'm
trying to locate more 1101A, so that I'll have enough for 1K and a few spare
ones too. Anyone out there have any 1101A or equivilent? (256 bit, pmos,
16 pin package +5V, -9V supply).


Another problem I had was a missing 8263 chip (Signetics dual 3-to-1
multiplexor) on the input mux board. I cannot remember why that chip was
gone, as the computer certainly wouldn't run at all without it. I probably
had removed it to test it, then lost it. I've found that the 8263 chip is
impossible to find and have subsequently replaced it with some 74XX chips.
I didn't have the data sheet, so I deduced its function by examining the
circuit. Only later did I realize that it had inverting outputs, requiring
a third chip (a 7404) to complete the replacement circuitry. If anybody out
there has an 8263 chip, I'd love to install it an get rid of the replacement
circuity.

During debugging, I found and replaced three other bad chips, all dealing
with the "bus" that I had tried to replace. The other problems were a
coupld of solder bridges and various broken connections, especially on the
"back plane", which gets flexed every time you fan open the boards in order
to access components on them. But the bottom line is.. IT WORKS!!!! And
I'm having a great time messing with it.

Now what?

In the short term, I'm installing the Mark-8 in an appropriate enclosure
with a modern switching power supply and fan. I plan on adding an EPROM and
UART to it. The EPROM will contain a bootstrap routine for performing a
memory check and a "monitor" routine allowing upload from the UART. I want
the UART and EPROM to be "vintage" parts (like the AY-3-1013 or COM 2052
UARTs, and 1702A EPROMs. Anyone out there capable of burning a 1702A (or
two) for me?

I also plan on using a cross-assembler or cross-compiler to write routines
for the Mark-8. I'm looking for Intel's PL/M cross compiler and
cross-assembler for the 8008. Apparently there also was an 8008 similator.
Intel has been suprisingly helpful in reproducing old manuals for the 8008,
but the source code for these old tools, THOSE are hard to find. All were
written in Fortran, and I understand that Gary Kildall wrote the CP/M
compiler. I figure that if I can get the sources, I could port them to the
PC. Anybody out there have these old software tools? I also heard that
there was a BASIC language for the 8008. Any help with that?

Anyone out there have a Mark-8?

- John Lewczyk
- IO Consulting
- 401 Queens Row Street
- Herndon, Virginia 20170-3131
- jlewczyk_at_his.com
Received on Thu Apr 08 1999 - 23:49:36 BST