Re: WTB imsai or altair

On Mon, 30 Jan 2006 19:21:40 GMT,
Dave.Dunfield@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx (Dave Dunfield)
wrote:

>>>Sorry, I just don't see it. Everyone I know who has ever bought a
>>>used S-100 machine has had constant problems making it run and LESS
>>>often even KEEPING it running because they never DID GET it running.
>>>I know of literally DOZENS of these albatrosses that have gotten
>>>passed around without EVER running well enough to actuallly DO
>>>anything with them!
>
>>You have met your first one. ;)
>
>And now a second - a quick count shows 15 S-100 machines residing
>in my basement - of these, 14 are currently running, and the last one is
>a big oddball Compuduct machine which was a basket case when I
>got it and I haven't really worked on it much yet (It was physically taken
>apart and many connections removed - including all the wiring for the
>internal monitor).
>
>I've run into very few "mysterious bus problems", a few of the machines
>needed cleaning of connectors and sockets, and I've had a few caps
>go bad over the years, and even a few bad bits of silicon, but nothing I
>would not expect from 30 year old equipment.
>
>
>>Personally if you are going to buy an Altair, you deserve what you
>>get and it's not pretty. It one of the few I got early (1975) and
>>retired early and have no plans to ever power again. If there was
>>one clear thing I can say about the IMSAI is it was one huge
>>improvement. However it would be a few years and getting rid of the
>>front pannel that significantly improved reliability. That and a
>>decent airflow which many didn't get even close to right.
>
>The Altair was a kit machine, and I think you get a lot of variation
>in the construction. I have two original 1975 Altair 8800's, and both
>are running perfectly.

Mine was in the earliest production lots. Most all of the ECOs were
required. However the dependence on oneshots was problematic.
Mine the CPU would go off into space and it was the timing generators
oneshots misfiring. By the summer of '75 I'd changed the oneshots to
8224 and system stability was good. However I also had the problem of
some MITS boards being solder plated edge connectors and the early
88MCD rams were gold over copper (green gold) problems and by the end
of it's first year most of the connectors were badly contaminated from
the Logisland costal climate. New backplane fixed that. I bet if I
dug it out and verified the PS it would run still.

On the other hand I have an 8800B-T chock full of all MITS cards CA
1978 including disk and it runs with no problems at all other than
occasional dusting inside. The pertec disks are even operational.

>One of these is a macine which I have owned since it was almost new
>(late 70's) and it has never let me down. It was my main computing
>platform for quite a few years (many of my early development tools
>began on this machine). Worst problem I've ever had with it is that after
>sitting in box a for quite a few years, I had to clean up the sockets on
>the NorthStar disk controller before it would boot ...

NS* was near indestructable. My first one suffered two lightining
hits. First was direct hit to house TV antenna, fried a lot of the
house electronics and the secod was a nearly pole hit. Both times I
replaced tha bad chips and kept on trucking.

>Another thing that people often forget is that the Altair was the first
>successful commercial "personal computer" - although there were
>other micros before then, this was the first one used by "the masses"
>that defined a standard bus and was popular enough that it got adopted
>by other vendors.

Indeed it was and highly copied in many variations.

>It was a very simple straightforward design, which consisted mainly of
>taking the signals available on the 8080 CPU and support chips straight
>to the bus connectors. Most of the problems I've seen where people
>complain about the Altair bus are in cases where they have changed the
>CPU to a 4Mhz Z80 - not only does it not match the original 8080 timing,
>but it is being "overclocked" to double it's design speed. Not to mention
>that many of the early Z80 cards (likely to be used in an Altair) were not
>the pinnacle of engineering either...

Two issues made it twitchy. One was the bundle of wires (8800 and A)
to the front pannel and the other was the older 4 slot ssegmented
single sided backplanes. Just replacing the backplane with later
Altair or wamco parts made that mostly go away.

One item that plagued S100 as a general thing. There were many ram
cards of the 8k of 2102 flavor were the inputs to the 2102s were not
buffered from the bus. Bad design, huge capacitive loading and
issues with ringing even with 8080 (worse if you had 6-7 of them!).

>Before anyone cries "but the S-100 bus is supposed to be able to do
>this", let me make clear that we are NOT talking about the S-100 bus.
>We are talking about the "Altair bus", also known as the "Roberts bus".
>
>Mits didn't design the Altair bus to handle S-100 specifications or 4Mhz
>Z80s because nieither of those existed at the time. The S-100 bus grew
>from the Altair implementation. The Altair bus was designed for the Altair,
>which is a 2Mhz 8080 - nothing more, nothing less. The fact that it was
>adopted by so many, and grew to encompass systems far beyond
>the original design is a testiment to the fact that Mits openly documented
>it and did not complicate it with proprietary "junk".
>
>Yes - the guys who came even a little bit later did it better - this is what
>is supposed to happen, you learn from observing the problems with
>other designs. And with microcontrollers getting cheap, and Z80s on the
>horizon, it was easy to see that dependance on the 8080 bus was not a
>good thing. In 1975 with the 8080 being the only game in town (and not a
>cheap one at that), this was not so obvious...
>
>To put things in perspective, the last crop of S-100 machines produced
>looked far more like the Altair than pretty much any later crop if "PC"s
>to the IBM 5150.
>
>Yeah, I am an Altair fan - my Altair was the first machine I put a diskette
>drive on, and got serious about writing software with - In many ways, the
>Altair launched me into my career, and I have deep attachment to it.
>But I also know what it is, which is a pioneering machine, built without
>the benefit of prior experience to draw on - Yet so many people talk
>about how good the later S-100 machines were in comparison to the
>Mits Altair, but only talk about the original 8800 - never the A or B, which
>were Mits improved designs - and I haven't found one of either yet :-(

It was my first and I'll keep it forever. However the 8800B was a
robust machine based on a lot of learning.

>>For "historical purposes" is the best answer, like old planes and
>>boats don't try to fly it or float it, the failure will destroy the
>>investment in a something worth more as a static object.
>>Those that do are usually experienced and have done it before
>>and know the pot of worms well.
>
>To buy an old Altair thats been sitting in someones basement and
>expect to turn the switch and boot it up is akin to buying a Model-T
>thats been rusting out behind someones barn and drive it away...
>But like the Model-T when properly restored, the Altair is worth
>running, and worth showing.

More like buying a ramp queen (little to unused aircraft thats sat for
far too long) and flying it without a good inspection and detailed
maintenance.

>I get lots of people who come to see my collection, and the Altairs are
>always among the most popular. I've taken one of them out to local
>schools to demonstrate the "early years" to the students and it was
>very well received. When a local TV station did a feature on my
>collection, they asked me to "shut up" during the boot sequence so
>that they could feature a closeup of the Altair front panel with all the
>sounds of the machine booting --- More and more we are becoming a
>highly digital society, and the early years of personal computing is
>an important part of that history. Please don't toss it in the trash.

My first got pretty ugly from traveling and general use. Then again
it was a work machine not a hobby toy. Despite its warts it worked,
it was first, and it held fairly well. For it's time it was no better
or worse than the infamous 5150 was for it's day though by then it was
almost 7 years old! In computers thats a lifetime.


Allison

.

Relevant Pages

  • Re: WTB imsai or altair
    ... >> Wanted dead or alive imsai or altair computers. ... >Why does everyone slobber over old S-100 machines, ... >The S-100 is a design best NOT preserved for posterity! ... The S100 bus was at best a poor design and really was ...
    (comp.os.cpm)
  • Re: WTB imsai or altair
    ... I've run into very few "mysterious bus problems", ... The Altair was a kit machine, and I think you get a lot of variation ... but it is being "overclocked" to double it's design speed. ... Mits didn't design the Altair bus to handle S-100 specifications or 4Mhz ...
    (comp.os.cpm)
  • Re: sad news...
    ... to support the view that prior to the Altair, ... MITS advertisement in the same issue lists the Altair system at $750 ... number of homebuilt cards, and every now and then I'd find a spot, stick ... not invent the bus: he was searching for a bus, ...
    (comp.os.cpm)
  • Re: sad news...
    ... this is what I wrote about it in the introduction to my Altair simulation: ... MITS created the "Roberts" bus, ... hardware for a computer was available only from the machines manufacturer. ... beginning of the personal computer industry started without question at ...
    (comp.os.cpm)
  • Re: homebrew computer - where to start?
    ... >>trying to sort out my memory here - was the Altair 8800 sold be SWTP? ... It was sold my MITS, which was a company that used to make ... > selling kits actually made some money. ... The November 1971 issue of Popular Electronics has their calculator, ...
    (sci.electronics.basics)