> From: Jacob Ritorto
> My NASA EE buddy recommended working up an appropriate load from
> regular household lightbulbs. They won't light, naturally, but you can
> dial in your load by simply choosing certain wattages and doing the math.
Err, both tungsten and carbon filament incandescents have very different
resistance when hot (lit) and cold. To make it even more, they are of opposite
sign: with tungsten, the resistance increases as it heats up (a nice
stabilizing feedback), and with carbon, it goes the other way. So if you're
not running enough current to light the bulb, you can't use the printed watt
rating; you'll have to use an ohm-meter.
Noel
Mattis Lind from this forum sent me some failed National Semiconductor 74xx
series, plastic encapsulated. These were from an HP 98xx series calculator I
believe. He noticed than when they failed they failed open. I failed myself
to have the local National Semi people interested in looking at them, and I
just did some XRays to find out if something obvious had happened to the
wire bonds. Nothing we could see, but we didn't take enough close ups. I
want to take a look again, and if I don't find anything could send them on
to you if Mattis is OK with it.
Marc
> Darn! That sounded like a good collection. Alas, I know I haven't kept
> every bad IC either (except that one from the PDP-8/E...not really sure
why
> until now, I guess). But if you do find them, I could certainly use them.
On 14/02/15 11:02, Cory Heisterkamp wrote:
> Tom, any reason for choosing the Harvard architecture?
it may have some Modified Harvard architecture
the computer is NOT Stored-program based, it may have a small (128/256
byte) CPU styled cache
> I'm currently in the middle of a relay computer build so I'll toss in
my 2 cents.
> The most important thing I hit on early is to define the instruction set,
> work out a number of potential applications/uses on paper (or excel),
> and step through each instruction to make sure the machine is capable
enough.
> You'll probably discover a number of efficiency improvements you can
make to the hardware while doing so,
> and a few "hardware subroutines" worth adding.
> You can also baseline how long a program will take to run if you
settle on a clock rate, and determine if there's a better way to do it.
> Ultimately, relay computers are SLOW. Don't take 6 cycles to do what
could be accomplished with a couple extra relays and a single pulse.
> Well thought out HW and instructions will pay big dividends when the
soldering iron comes out. -Cory
good point, I am prototyping the computer design in software and
software in 6502 assembly
Sorry, this was accidentally sent before it was completed.
At 04:40 PM 2/13/2015, Tom Sparks wrote:
>I've been bouncing around the idea of building my own relay computer
I built a greatly scaled down version of Simon more than 50 years ago. See Chapter 3 in:
Edmund Callis Berkeley
Giant Brains, or Machines That Think
New York: John Wiley & Sons, Inc., 1949; Reprinted New York, Science Editions, Inc., 1961
Available in DejaVu format at:
https://ia600504.us.archive.org/17/items/GiantBrains/giant_brains.djvu
Dale H. Cook, GR / HP Collector, Roanoke/Lynchburg, VA
http://plymouthcolony.net/starcity/radios/index.html
On 14/02/15 09:52, Brent Hilpert wrote:
> On 2015-Feb-13, at 2:56 PM, Tom Sparks wrote:
>> On 14/02/15 08:20, Brent Hilpert wrote:
>>> On 2015-Feb-13, at 1:40 PM, Tom Sparks wrote:
<snipped>
> Just an opinion, but the Zusie / Z3 recreation ref'd by Pontus looks most interesting, if one wants to do an historical design vs. something novel.
my project is a what if?....
"1890 portable telegraph switching computer"
I am limiting myself to 1930's knowledge and ideas because they could
have be discovered early in time, but unpublished/lost/etc
I am welling to use modern versions to reduce the build time
> The Zusie site mentions the Z3 recreation in Germany.
> With a few recreations of the same machine there's the potential to benefit from community software development.
>
> The Zusie fellow sure was fortunate to obtain a large quantity of good relays at scrap prices.
> I've long been interested in building a relay computer but that's one of the stumbling blocks - relays new are so pricey today.
I noticed that with the relays
> Had in mind a physical implementation much like the Zusie, upright front-access 19" relay rack with an LED per relay to show state (and many blinkenlights).
>
I am writing my master's thesis on accelerated life testing and would like
to get a collection of parts to decap and experiment on. What prompted my
study is a failed 7474 dual flip-flop in a PDP-8/E. As I later read (on
here, I believe) was that the 7474s seem to fail in higher number than
other 7400-series ICs. Is this because of a design flaw? I'm not sure
(yet), but would like to find out.
More recently, I saw a 680k 2W carbon resistor that read 1.3 meg out of
circuit. As it turns out, it was in a high voltage portion of the circuit,
and I suspect that electromigration caused its failure over time. Many of
you are probably already aware that carbon resistors tend to increase in
value as they age, though there are a few explanations too for that
behavior, as I understand it.
I would like to tie all of this together by researching older accelerated
life testing schemes, examine how those parts actually fared, and see how
current day models might be improved.
If anyone has a collection of dead ICs (preferably those that have died
inexplicably, and not by extreme overvoltage, for instance), I would like
to talk to you about getting a few. If they are 7474s from Texas
Instruments, ca. early 1970s, even better!
Thanks in advance,
Kyle
I presume I can't be the only one watching this eBay auction in disbelief.
I'm wondering if this is a museum or not. Presumably a private collector
would not have such deep pockets, but you never know!
Over $13k USD and still a little less than 2 days to go!
http://www.ebay.com/itm/Ultra-rare-Commodore-65-C65-DX64-prototype-working-…
I have a friend with a large stash of DG boards which comes from a
collection of stuff he and his company had from supporting Novas. He has
just the stash of boards and would like to get a rig to test them with.
Ideally acquire, but from discussing what he wants to do he would
probably need a minimal set including power, and the front panel to run
tests. I wanted to locats contacts to put him in touch with that might
be able to help him, and you and he can discuss the best way to do what
he is after.
I don't know if he wants to sell / trade all of them, end up with a
working machine or what he may need to do, so he is flexible.
I'll put you in touch with him in whatever manner you feel comfortable with.
thanks
Jim
