19 Aug
1998
19 Aug
'98
3:56 a.m.
< ObCC: Supposedly, MITS came out with a hobbyist machine a couple of year
< before the Altair called the MITS 816. I haven't found much info about
< it. Is it real?
The only prior products to the altair were a hand calc (several) and also
some hobbiest telemetry bits. If there was another computer it was a
maybe to compete in the MARK-8 realm (8008 based).
In the 8008 realm there were several companies that predate MITS. There
used to be a company here in MA called Control Logic I think that had
their "L series" modules that were 8008 based.
Personally I view collecting Altairs to wine... altair was the first lot
produced by a particular vintor. It was as drinkable as hog urine but,
it was their first and the vintor would become famous and fail making
that first lot scarce and "collectable".
Most of the early altairs had to be hacked (required) to make them work.
They are now shelf addornments and should be held up as an example of
how not to do it. I'm being serious. Anyone that wants a S100 front
pannel box to actually run would do far better with an IMSAI, Ithica
Intersystems, or several others. Most however learn the first thing
with these is that once you have it running the first code you want is
a terminal keyboard monitor as replacement switches are a few bucks each
and the originals are getting poor. Also toggling in 25 or 50 bytes every
power up or crash gets tiring. I say this as someone that lived with the
MITS ACR (sill have the analog board for one) where you toggle
in a 33byte loader, then load MITS basic crash half way through (10
minutes), toggle in the loader again, load the tape... I think that
lasted three months until I made my first PROM card using 8223 proms.
A few weeks later the ACR tape interface would be removed in favor of
a digital tape of my own design. The load, crash, reload cycles were
a great detriment to the goal of doing some serious programming. I
still feel the flakey hardware and really awful audio cassette kept me
from doing anything useful for the first 6-7 months of
ownership.
Hacks applied to mine to make it work with some reliablity before it was
replaced with a NS* system in early 78.
* Moved AC power from front pannel switch as I felt it was dangerous
having an unprotected traces with 110v on them.
* 8V DC line was 7.8v with cpu and two 4k dram cards. Added 12 turns to
the winding. Upgraded filter cap to larger value, replaced undersized
rectifers. Mits offered a new transformer at users cost.
* +-16V DC line marginal, rectifers fried, upgraded.
* Backplane was in four slot segments and was a single sided affair.
I had 8 slots and had to add bussbar to the power and ground lines
as the power distribution from one end of the bus tot he otehr was
poor.
* modifed memory cards (88MCD) to improve timing and signal quality.
* Replaced backplane with a improved two sided one from third party.
* Replaced 2mhz crystal and oneshots used for 8080 clock with 8224 and
18mhz crystal.
* modified memory cards to replace 4060s with uPD410 static devices of
the same pinout.
After all this the system would load MITS basic and run for a week solid
without crashing.
By the middle of '76 I'd be up to 4 memory cards of both the 88mcd design
and a vastly better 88S4k design. a then whopping 16k of ram. That would
be eclipsed by the addition of a Seals 8k static card that was 1/3 the
cost of the first 16k and still works!
If I'd waited a year and bought an IMSAI I'd have saved all that time,
about $3000 and had a machine that would not have required replacement.
The upside is by that time I'd learned how it should NOT be done.
Allison