"Thirty years ago, I was just starting to
develop an interest in computers, an interest that ultimately led to a
very enjoyable career. I subscribed to all the magazines that
had just started publication -- Byte, Kilobaud, Dr Dobbs Journal of
Computer Calisthenics and Orthodontia (running light without
overbyte). I was determined to get computer of my very own,
but I had never built even the simplest of electronic circuits, so
doubted my ability to build one of the kits that were available, and I
eventually settled on a brand new little beige box with the unlikely
name of Apple ][. I forked over the princely sum of $1249.92 to the
Byte Shop in Miami and had to wait about three weeks for them to get
one in, but eventually I had my own little miracle machine, serial #97,
probably the first Apple delivered in Florida.
I spent many glorious hours with that machine
over the next few years learning how to program and how computers work.
I never once regretted buying it, but there was always a faint idea in
the back of my mind that I should have skipped this little plastic
"toy" machine and invested in a "real" one. Something with an
S-100 bus. Something with lights and switches. An IMSAI. Or
an Altair. Now those were computers.
That idea never left me, and recently I have
finally been able to acquire some of those computers that were out of
my reach back then. I have several S-100 and CP/M machines:
SOL-20s, IMSAIs, an Altair 8800b and others, but I haven't been able to
find the crown jewel, the Altair 8800, at least not one that I was
willing to buy, and not at a price that I was willing to pay.
And even though it is satisfying to have these machines in my
collection, it's not the same as having built one of my own.
Then, around the end of 2006, I came across www.altairkit.com,
a site run by Grant Stockly. Grant had disassembled an Altair
8800, created exact replicas of its boards and other components and was
selling what amounted to an original Altair in kit form. The
web
site makes interesting reading, especially where he describes how he
went about reproducing the boards, even resorting to x-rays to
determine the layout of the board traces. This was not just
an
Altair-like kit, but an exact replica, complete to the smallest
details. I missed my chance to order one of the first three
kits
that Grant produced, but when they were gone he offered a second run of
20 more and I immediately put my name down for one of them.
My machine (serial #9) arrived on Feb 24, 2007,
and I
immediately tore into it and started checking out all the parts. Grant
claimed on his web site that the kit contained only high quality boards
and components, and he wasn't exaggerating. There are differences
between the original boards and those in the Altairkit, but those
differences are superficial and serve only to improve the
boards.
For example, the boards have better silkscreens and solder masks and
use more modern and reliable capacitors in place of the original
ceramic disc caps. But the layout of the boards is identical
to
the originals, and rather than using easier to find modern chip
substitutes, Grant went to great lengths to track down the exact same
chips that came with the original machines.
To enhance the accuracy of the Altairkit, he
even had Optima, the original manufacturer of the Altair cases, create
exact duplicates of the original cases. To my eye, the color,
texture, and design of the new cases are indistinguishable from the
original ones. The only substantial difference from the
original Altair is in the power supply. Grant wisely chose to
omit the massive transformer and analog power supply and replace them
with two modern switching power supplies. As a result, the machine runs
much cooler and you can move it around without risking major back
injury.
The manual that comes with the kit is a
beautifully
printed copy of the original with a well written and illustrated
section added by Grant to help guide the owner through the assembly
process. The main section of the manual is not just a cheap
photocopy of the original MITS manual, but is every bit as readable as
the original and appears to be printed on better quality paper.
Assembly is best started by checking all the
parts
against the packing lists provided in the manual. I found
everything correct except for a .01uf capacitor which should have been
.001uf, and a 7408 chip which should have been a 7400. Considering the
number of parts involved, all hand packed, I'm surprised that those
were the only incorrect components, and Grant immediately offered to
replace them with the correct items, but I already had replacements on
hand so I just plowed ahead.
I found the best approach was to follow Grant's
assembly
instructions which diverge from MITS's instructions occasionally, and
when they do he points you to the appropriate page in the MITS manual
so that you can decide which sequence to follow. Assembly went very
smoothly, and it took me a little over a week to assemble the case,
backplane, front panel, memory board, and CPU. I found the
hardest
part was the myriad of individual wires between the front panel and the
backplane, but if you take your time and approach it logically it
shouldn't be a problem. Once I had everything assembled I
visually
checked everything for solder bridges and missed joints, and then
checked for short circuits between the power and ground lines on the
boards and backplane using a multimeter. At this stage I
hadn't yet
installed any of the chips. Grant provides high quality
sockets for
all chips, another welcome way that the kit differs from the original,
which requires you to solder most of the chips directly to the boards.
I inserted the chips into their sockets and
still didn't quite have the guts to just plug the machine in and power
it up. Instead, I used a current-limited power supply and monitored the
current draw. No magic smoke escaped, and the front panel lit
up as it was supposed to. I finally hooked up the power
supplies, switched the machine on and ... wow! It may have taken thirty
years, but I finally had a true Altair 8800 built with my own hands
happily twinkling its little LEDs at me.
It's hard to fully express how impressed I am
with the
Altairkit 8800. The fact that it worked first time is a
tribute
more to the quality of the kit and its instructions, than to my
assembly skills. On his web site, Grant says his intent was
to
produce a kit as true as possible to the spirit of the original that
provided the experience of building one of those ground breaking
machines, and he has succeeded admirably. The chances are
that I
will still buy an original Altair 8800 when I find the right one, but
when I do, I suspect I will value the Altairkit more than the antique.
I missed the opportunity to build an Altair all those years ago, but
now I've finally done it, and I've finally scratched that little itch
that's been bothering me all this time.
Grant is also in the middle of producing several
expansion boards for the Altair: replicas of the MITS 88-ACR cassette
tape interface, the 88-2SIO serial port card, and others.
After seeing the quality of the Altairkit, I had no qualms at all about
placing a standing order for one of each of any boards he makes for the
machine. All I have to do now is to find the patience to wait
for them to arrive."
Geoff Harrison
sanddune at solivant.com
|