19 Jan
2004
19 Jan
'04
2:31 a.m.
Then it's high time somebody produced a _real_
computer kit. As in a pile
of standard logic chips (no microoprocessors, no programmed devices, no
FPGAs, although I will allow RAM :-))....
I would really, really like to see something like this...it'd help me
getting started with learining electronics, which is something I've really
wanted to do and not had any practical time to do.
19 Jan
19 Jan
4:31 a.m.
On Mon, Jan 19, 2004 at 07:31:07AM -0600, Christopher Cureau wrote:
^^^^^^^^^^^^^^^^^^^^^
Tssss. Not to discourage you, but a TTL computer with bazillions of ICs
is not really suited for newbies, although there is plenty to learn -
namely on those nasty aspects like circuit timing, delays, glitches,
noise on power and signal lines. TTL is to a large degree analog, not digital,
circuit development ;-)
You might not directly want to start with a monster like Computer 74 or
EGO, but perhaps with a simple ALU/register device: get a 2901 circuit
Then it's high time somebody produced a _real_
computer kit. As in a pile
of standard logic chips (no microoprocessors, no programmed devices, no
FPGAs, although I will allow RAM :-))....
I would really, really like to see something like this...it'd help me
getting started with learining electronics, which is something I've really from BG-Micro and play a bit with it (a breadboard with
lots of switches
and LEDs is sufficient). And then extend it with a 2909/-10/-11
from the same
source and build the micro control for that toy. This is already the
basic step towards an own homegrown system.
wanted to do and not had any practical time to do.
:-) If time is your limiting factor, it will be your limiting factor
here as well.
Holger
9:30 a.m.
[...learning
electronics...]
Tssss. Not to discourage you, but a TTL computer with bazillions
of
ICs is not really suited for newbies, although there is plenty to
learn - namely on those nasty aspects like circuit timing, delays,
glitches, noise on power and signal lines. TTL is to a large degree
analog, not digital, circuit development ;-) You might not directly want to start with a monster
like Computer 74
or EGO, but perhaps with a simple ALU/register device: get a 2901
circuit from BG-Micro and play a bit with it (a breadboard with lots
of switches and LEDs is sufficient). And then extend it with a
2909/-10/-11 from the same source and build the micro control for
that toy.
Good advice. Start simple. My first major piece of TTL hackery was a
little toy 4-bit microprocessor; the most complicated single chip used
was a 4-bit ALU whose 74xxx number I forget by now. (Oh, and a chip
which was basically a 16-word static RAM with 4-bit words.) I was
fortunate enough to have a _really_ _nice_ breadboard to work with.
None of this wish-board stuff; it had individual IC sockets with very
nice machined contacts, and the wires had little stackable plugs (a bit
like banana plugs done tiny) on the ends of them. All contacts were
gold-plated. I shudder to think what they cost the University, but
they were _so_ pleasant to use.
/~\ The ASCII der Mouse
\ / Ribbon Campaign
X Against HTML mouse(a)rodents.montreal.qc.ca
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
10:22 p.m.
Somewhere in my list of (too many) projects to work on some day is to implement one of the
first (tube) computers in SSI. One I had done some design on was the SSEM/Manchester
"Baby" ("the first machine to run a stored program"/1948). I did
enough of the logic design to implement it in a home-brew logic simulator (executable in
browser via shockwave at http://www.cs.ubc.ca/~hilpert/tih/SSEM), but haven't got
around to doing a physical implementation. Figured CMOS SSI would be the way to go. For
the "Baby", a scope/CRT with Z-axis control could be used to emulate the
original 32-bit-by-32-word storage-tube-memory monitor. I have an HP digital signal
analyser from 1969 (basically an early digital oscilloscope) which has all the CRT support
and power supplies and figured the new/old processor could be built into the HP DSA with a
new front panel, making an interesting amalgam of hardware and architecture spanning
several decades of computing. The same idea could apply to some o!
ther 'scope equipment of course.
Another interesting machine to implement in SSI might be the IAS machine, or perhaps
something like the IBM 709.
One could get carried away making scale-model front-panels with lots of LEDs and miniature
toggle switches.
20 Jan
20 Jan
9:36 a.m.
Well my equally crazy (but smarter) friend David Forbes and I talked
about making Williams memory, and the problems (modern) thereof. Well it
turns out he makes oscilloscope clocks (quadrature character generators
and nice things) and we talked about mocking one up -- using the 68HC08
on board, simulating a 1, 4, 8, ... bit Williams memory cell, what with
all the circle/dot/slash/focus/defocus modes, we could tightly
characterize readaround issues, very repidly reinvent those wheels of
the past. He buys CRTs from the manufacturer and knows a lot about this
stuff. http://www.cathodecorner.com.
The pedant in me says a CPU bigger than all combined planteary computing
resources of 1955 driving a single memory element might be cheating.
Those old guys did this with tubes! It's quite terrifying.
tomj
On Tue, 2004-01-20 at 01:22, Brent Hilpert wrote:
Somewhere in my list of (too many) projects to work on
some day is to implement one of the first (tube) computers in SSI. One I had done some
design on was the SSEM/Manchester "Baby" ("the first machine to run a
stored program"/1948). I did enough of the logic design to implement it in a
home-brew logic simulator (executable in browser via shockwave at
http://www.cs.ubc.ca/~hilpert/tih/SSEM), but haven't got around to doing a physical
implementation. Figured CMOS SSI would be the way to go. For the "Baby", a
scope/CRT with Z-axis control could be used to emulate the original 32-bit-by-32-word
storage-tube-memory monitor. I have an HP digital signal analyser from 1969 (basically an
early digital oscilloscope) which has all the CRT support and power supplies and figured
the new/old processor could be built into the HP DSA with a new front panel, making an
interesting amalgam of hardware and architecture spanning several decades of computing.
The same idea could apply to some
o!
ther 'scope equipment of course.
Another interesting machine to implement in SSI might be the IAS machine, or perhaps
something like the IBM 709.
One could get carried away making scale-model front-panels with lots of LEDs and
miniature toggle switches.
9:55 a.m.
New subject: Modern Tube computing
Have we forgotten John Zabolitzky's MUNIAC? While I do not think he has
completed it yet... nonetheless parts of it function and have been
displayed at various VCFs... it is constructed on standardized Eurocards,
which are them populated with the correct components to realize a
particular function - gates, flip-flops, buffers, etc.
Hans Franke might know if any documentation for this is available
someplace...
And no one has had the utter madness to suggest a replica of the Mark I
or II relay machine..... or the SSEC.... or any of the Bell System relay
computers....
Then of course - there's Stonehenge.
Cheers
John
10:02 a.m.
New subject: Modern Tube computing
John Lawson wrote:
Have we forgotten John Zabolitzky's MUNIAC? While I do not think he has
completed it yet... nonetheless parts of it function and have been
displayed at various VCFs... it is constructed on standardized Eurocards,
which are them populated with the correct components to realize a
particular function - gates, flip-flops, buffers, etc.
I would like more details too myself.
And no one has had the utter madness to suggest a replica of the Mark I
or II relay machine..... or the SSEC.... or any of the Bell System relay
computers....
Yet!
Then of course - there's Stonehenge.
They have that done already. ( But the replica don't work since it is on
the wrong latitude )
Cheers
John
1:03 p.m.
New subject: Modern Tube computing
Have we forgotten John Zabolitzky's MUNIAC?
No, but I haven't heard anything about it since the initial flurry of
activity.
I would also like to fool around with making a tube based computer, but
frankly, I have enough projects to keep me busy. I have the tubes -
probably about 2000 computer rated dual triodes of various flavors, but I
am holding them until someone rolls a tubeless 709 up my driveway.
One thing I would find very interesting is making a tube computer (or
rather just the building blocks) using technology from 1930. No modern
stuff. Vintage tubes (24s, 27s, and the like), vintage sockets, vintage
caps and resistors, and so forth. It might be fun to see if even basic
logic functions could be made reasonably. Of course, a full computer
(imagine racks and racks of black wrinkle chassis with glowing tubes
behind) is silly for many reasons - size (even a serial machine would have
a big footprint), and parts availability (the tubes are not so bad, it is
everthing else, like *good* *reliable* vintage caps, as well as the iron).
And the radio people would probably hang you for buying up all these
parts.
William Donzelli
aw288(a)osfn.org
4:21 p.m.
New subject: Modern Tube computing
William Donzelli wrote:
Great more computers with WHEELS.
Have we
forgotten John Zabolitzky's MUNIAC?
No, but I haven't heard anything about it since the initial flurry of
activity.
I would also like to fool around with making a tube based computer, but
frankly, I have enough projects to keep me busy. I have the tubes -
probably about 2000 computer rated dual triodes of various flavors, but I
am holding them until someone rolls a tubeless 709 up my driveway. One thing I would find very interesting is making a
tube computer (or
rather just the building blocks) using technology from 1930. No modern
stuff. Vintage tubes (24s, 27s, and the like), vintage sockets, vintage
caps and resistors, and so forth. It might be fun to see if even basic
logic functions could be made reasonably. Of course, a full computer
(imagine racks and racks of black wrinkle chassis with glowing tubes
behind) is silly for many reasons - size (even a serial machine would have
a big footprint), and parts availability (the tubes are not so bad, it is
everthing else, like *good* *reliable* vintage caps, as well as the iron).
And the radio people would probably hang you for buying up all these
parts.
Hey, I saw just the computer you were thinking about on a old STAR TREK
re-run: CITY ON THE EDGE OF FOREVER.
William Donzelli
aw288(a)osfn.org
Vintage TUBE stuff is the latest fad of PRO-AUDIO, so that stuff is all
dried up. You can't even get tubes from the 1940's for example: New Old
Stock 6SN7 $39 ... New sov-tech 6SN7 $9.
Ben.
21 Jan
21 Jan
12:16 a.m.
New subject: Modern Tube computing
On Wed, 2004-01-21 at 03:21, ben franchuk wrote:
Vintage TUBE stuff is the latest fad of PRO-AUDIO, so that stuff is all
dried up. You can't even get tubes from the 1940's for example: New Old
Stock 6SN7 $39 ... New sov-tech 6SN7 $9.
Ben.
The problem with NOS is that the glass is slightly porous and the valves
will have gone a bit soft by now. Why would you pay four times as much
for a valve that will be below standard when you get it, when you can
have a brand new valve, build to a higher standard than the original?
Gordon.
4:43 a.m.
New subject: Modern Tube computing
The problem with NOS is that the glass is slightly
porous and the valves
will have gone a bit soft by now.
This is basically false.
Just like the old devices, some NOS tubes are awful (various Chinese
brands). The same holds true for the oldies (Arcturus, Penn, Rad Tel,
etc.). Good stuff is good stuff, crap is crap.
This is why I really can't stand the tube audio world - there is too much
junk science, snake oil, prejudice, and false rumors. Anything to make the
rare old stock "sound" better.
William Donzelli
aw288(a)osfn.org
5:04 a.m.
New subject: Modern Tube computing
On Wed, 2004-01-21 at 15:43, William Donzelli wrote:
Well, I call it as I see it. I have literally dozens of new old stock
valves, in their original cardboard sleeves. Some are useable, some are
absolutely shagged. It's not all just cheap brands, too.
The problem
with NOS is that the glass is slightly porous and the valves
will have gone a bit soft by now.
This is basically false. Just like the old devices, some NOS tubes are awful
(various Chinese
brands). The same holds true for the oldies (Arcturus, Penn, Rad Tel,
etc.). Good stuff is good stuff, crap is crap.
Which is why I only really use Sovtek stuff now. It's real genuine
Mullard parts, just like the handbook for my amp specifies, but with a
different screen-print on the glass...
This is why I really can't stand the tube audio
world - there is too much
junk science, snake oil, prejudice, and false rumors. Anything to make the
rare old stock "sound" better.
I don't know why anyone would use a valve hi-fi amp, that just seems
silly. But they look good...
In guitar amps, it's a very different story of course.
Gordon.
11:35 a.m.
New subject: Modern Tube computing
Gordon JC Pearce wrote:
From what I have read on the internet, speakers have got better
in the last 20 years, Solid State Amps have not gotten better,
since they are often not tested on real speakers. Tube amps with
less negitive feedback handle modern speakers better.
Ben.
On Wed, 2004-01-21 at 15:43, William Donzelli wrote:
This is why I
really can't stand the tube audio world - there is too much
junk science, snake oil, prejudice, and false rumors. Anything to make the
rare old stock "sound" better.
I don't know why anyone would use a valve hi-fi amp, that just seems
silly. But they look good...
In guitar amps, it's a very different story of course. 
11:53 a.m.
New subject: Modern Tube computing
Quothe Gordon JC Pearce, from writings of Wed, Jan 21, 2004 at
03:55:40PM +0000:
On Wed, 2004-01-21 at 15:43, William Donzelli wrote:
Well, I call it as I see it. I have literally dozens of new old stock
valves, in their original cardboard sleeves. Some are useable, some are
absolutely shagged. It's not all just cheap brands, too.
That makes me wonder if some of those tubes were actually used tubes
put in the cases of new tubes when tubes were replaced. Of course,
tubes can develop leaks and become gassy. Also, some of those boxes
may have been dropped over the course a few decades or so.
I don't know why anyone would use a valve hi-fi
amp, that just seems
silly. But they look good...
Tube amps can sound good as well. Hint: use bi-amping or tri-amping.
For the low-end, use a big solid-state amps with huge filter
capacitors, and use tube amps for the higher frequencies.
--
Copyright (C) 2003 R. D. Davis The difference between humans & other animals:
All Rights Reserved an unnatural belief that we're above Nature &
rdd(a)rddavis.org 410-744-4900 her other creatures, using dogma to justify such
http://www.rddavis.org beliefs and to justify much human cruelty.
7:16 a.m.
New subject: Modern Tube computing
Quothe William Donzelli, from writings of Wed, Jan 21, 2004 at
10:43:38AM -0500:
This is why I really can't stand the tube audio
world - there is too much
junk science, snake oil, prejudice, and false rumors. Anything to make the
rare old stock "sound" better.
Something tells me that when Western Electric resumes the manufacture
of the 300B (perhaps some time this year), there will be people eager
to pay much higher prices for NOS 300B tubes because they will believe
them to sound better, more "musical", "sweeter", "less
nasal" or
whatever other ridiculous term they might come up with for how they
believe the tubes to sound.
Now then, just imagine how one could market a tube-based computer to
those computer aficionados who share certain traits with the
golden-eared audiophiles. Perhaps even the addition of a tube-based
keyboard interface of some sort, for PeeCees, would demand high prices
for its ability to make the characters enter the computer in a more
"flowing" and "ethereal" way. (ROFL!) Perhaps even using tubes as
analog components in network interfaces, to improve the sound of
digital audio, to make it sound "less harsh" would be an idea that
certain people would drool over. Perhaps one could also sell them
replacement gold-plated sockets and anode connectors for the picture
tubes in their monitors... any investors out there who want to invest
in some of my ideas? I've got even more related ideas, so, don't
delay, $$$s are being lost by the second! :-)
--
Copyright (C) 2003 R. D. Davis The difference between humans & other animals:
All Rights Reserved an unnatural belief that we're above Nature &
rdd(a)rddavis.org 410-744-4900 her other creatures, using dogma to justify such
http://www.rddavis.org beliefs and to justify much human cruelty.
7:47 a.m.
New subject: Modern Tube computing
--- "R. D. Davis" <rdd(a)rddavis.org> wrote:
Perhaps even
using tubes as
analog components in network interfaces, to improve
the sound of
digital audio, to make it sound "less harsh" would
be an idea that
certain people would drool over.
Already done, to see a tubed motherboard click on
http://www20.tomshardware.com/motherboard/20021017/hamm
er-02.html
__________________________________
Do you Yahoo!?
Yahoo! Hotjobs: Enter the "Signing Bonus" Sweepstakes
http://hotjobs.sweepstakes.yahoo.com/signingbonus
10:18 a.m.
New subject: Modern Tube computing
Quothe steve, from writings of Wed, Jan 21, 2004 at 10:47:10AM -0800:
No, it hasn't been done. The tubes on the motherboard to which you're
referring are in the audio circuitry. That's too practical. My idea
is impractical enough to satisfy even the most discriminating
aficionados of the impractical, whose apparent sensibility factor
ratings are inversely proportional to their net worth. I was
referring to the use of tubes as analog components in the network
interfaces, which would have the HYPothetical effect, approved by the
spirit of P.T. Barnum himself, of purifying the electrons entering and
leaving the systems by boiling them and also making the bits
transporting digitized audio more musical by smoothing the bit
transfer process by thermionic means, which also gives the processed
bits transporting digitized audio a much warmer sound. Without that,
the tube audio circuitry on such a motherboard, which is being fed
inadequately processed, formerly digitized, audio, still sounds far
too harsh. This is why my exclusive valved network interface, which
uses specially processed psychedelic-colored, capacitors, made from
recycled and reformed NOS capacitors, costing over US$1,000 each, from
the 1930's, with gold plated wires, is needed. Without it, the sound
using tubes
as
analog components in network interfaces, to improve
the sound of
digital audio, to make it sound "less harsh" would
be an idea that
certain people would drool over.
Already done, to see a tubed motherboard click on from a computer system can not sound pure enough to
satisfy even the
most discriminating golden ears. :-) :-) :-)
--
Copyright (C) 2003 R. D. Davis The difference between humans & other animals:
All Rights Reserved an unnatural belief that we're above Nature &
rdd(a)rddavis.org 410-744-4900 her other creatures, using dogma to justify such
http://www.rddavis.org beliefs and to justify much human cruelty.
8:19 a.m.
New subject: Modern Tube computing
On 21 Jan 2004, R. D. Davis wrote:
Now then, just imagine how one could market a
tube-based computer to
those computer aficionados who share certain traits with the
golden-eared audiophiles.
Want to know why your computer crashes all the time? Because the one
you've got is not using clean electronics! That's right, digital
computers made with Integrated Circuits (ICs) produce inferior
calculations, leading to subtle shifts and fluctuations that can cause
rounding errors and ultimately cause Windows(tm) to crash.
Instead, buy our nifty ENIAC (Entirely New Incredibly Accurate Computer)
3000 Computer system! It's based on vacuum tube electronics. That's
right, vacuum tubes!
Audiophiles have known for years that tube-based audio gear produces a
richer, more vibrant sound than that offered by digital audio electronics.
That's because tubes prouce a more clean signal. And the same
effect applies to computing. If you're using a digital computer made
with ICs, their resolution is not fine enough to handle calculations to
the precision that modern computing requires, thus causing regular
computer crashes. In fact, over 80% of computer failures in the United
States [1] are attributed by Complex Random Attenuation Problems. CRAP
is the leading cause of computer problems, beating out software bugs and
even shoddy programming. Our all new, state-of-the-art ENIAC 3000 uses
cleaner, more precise, vacuum tube logic elements that makes our computer
hum. It's like music (coming from a tube-based amplifier) to your ears!
Beat the CRAP out of your computer and upgrade to an ENIAC 3000 today!
[1] Statistics from the Bureau of Universal Numerics & Computing (BUNC)
(Ad paid for by Microsoft)
--
Sellam Ismail Vintage Computer Festival
------------------------------------------------------------------------------
International Man of Intrigue and Danger http://www.vintage.org
[ Old computing resources for business || Buy/Sell/Trade Vintage Computers ]
[ and academia at www.VintageTech.com || at http://marketplace.vintage.org ]
9:11 a.m.
New subject: Modern Tube computing
Wow, thats pretty good, and you did that on-the-fly
while sitting at a terminal too! You should supplement
your income doing some ad copy or at least an
infomercial!
--- Vintage Computer Festival <vcf(a)siconic.com> wrote:
Want to know why your computer crashes all the time?
Because the one
you've got is not using clean electronics! That's
right, digital
computers made with Integrated Circuits (ICs)
produce inferior
calculations, leading to subtle shifts and
fluctuations that can cause
rounding errors and ultimately cause Windows(tm) to
crash.
Instead, buy our nifty ENIAC (Entirely New
Incredibly Accurate Computer)
3000 Computer system! It's based on vacuum tube
electronics. That's
right, vacuum tubes!
Audiophiles have known for years that tube-based
audio gear produces a
richer, more vibrant sound than that offered by
digital audio electronics.
That's because tubes prouce a more clean signal.
And the same
effect applies to computing. If you're using a
digital computer made
with ICs, their resolution is not fine enough to
handle calculations to
the precision that modern computing requires, thus
causing regular
computer crashes. In fact, over 80% of computer
failures in the United
States [1] are attributed by Complex Random
Attenuation Problems. CRAP
is the leading cause of computer problems, beating
out software bugs and
even shoddy programming. Our all new,
state-of-the-art ENIAC 3000 uses
cleaner, more precise, vacuum tube logic elements
that makes our computer
hum. It's like music (coming from a tube-based
amplifier) to your ears!
Beat the CRAP out of your computer and upgrade to an
ENIAC 3000 today!
[1] Statistics from the Bureau of Universal Numerics
& Computing (BUNC)
(Ad paid for by Microsoft)
--
Sellam Ismail
Vintage Computer Festival
------------------------------------------------------------------------------
International Man of Intrigue and Danger
http://www.vintage.org
[ Old computing resources for business ||
Buy/Sell/Trade Vintage Computers ]
[ and academia at www.VintageTech.com || at
http://marketplace.vintage.org ]
__________________________________
Do you Yahoo!?
Yahoo! Hotjobs: Enter the "Signing Bonus" Sweepstakes
http://hotjobs.sweepstakes.yahoo.com/signingbonus
12:56 p.m.
New subject: Modern Tube computing
On Wed, 21 Jan 2004, steve wrote:
Wow, thats pretty good, and you did that on-the-fly
while sitting at a
terminal too! You should supplement your income doing some ad copy or at
least an infomercial!
Not my best work (I put about 4 minutes into it) but thanks :)
Anyone need an ad copyist? :)
--
Sellam Ismail Vintage Computer Festival
------------------------------------------------------------------------------
International Man of Intrigue and Danger http://www.vintage.org
[ Old computing resources for business || Buy/Sell/Trade Vintage Computers ]
[ and academia at www.VintageTech.com || at http://marketplace.vintage.org ]
10:47 a.m.
New subject: Modern Tube computing
I am rolling outta my chair! good one!
----- Original Message -----
From: "Vintage Computer Festival" <vcf(a)siconic.com>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk(a)classiccmp.org>
Sent: Wednesday, January 21, 2004 12:19 PM
Subject: Re: Modern Tube computing
On 21 Jan 2004, R. D. Davis wrote:
Festival
Now then, just imagine how one could market a
tube-based computer to
those computer aficionados who share certain traits with the
golden-eared audiophiles.
Want to know why your computer crashes all the time? Because the one
you've got is not using clean electronics! That's right, digital
computers made with Integrated Circuits (ICs) produce inferior
calculations, leading to subtle shifts and fluctuations that can cause
rounding errors and ultimately cause Windows(tm) to crash.
Instead, buy our nifty ENIAC (Entirely New Incredibly Accurate Computer)
3000 Computer system! It's based on vacuum tube electronics. That's
right, vacuum tubes!
Audiophiles have known for years that tube-based audio gear produces a
richer, more vibrant sound than that offered by digital audio electronics.
That's because tubes prouce a more clean signal. And the same
effect applies to computing. If you're using a digital computer made
with ICs, their resolution is not fine enough to handle calculations to
the precision that modern computing requires, thus causing regular
computer crashes. In fact, over 80% of computer failures in the United
States [1] are attributed by Complex Random Attenuation Problems. CRAP
is the leading cause of computer problems, beating out software bugs and
even shoddy programming. Our all new, state-of-the-art ENIAC 3000 uses
cleaner, more precise, vacuum tube logic elements that makes our computer
hum. It's like music (coming from a tube-based amplifier) to your ears!
Beat the CRAP out of your computer and upgrade to an ENIAC 3000 today!
[1] Statistics from the Bureau of Universal Numerics & Computing (BUNC)
(Ad paid for by Microsoft)
--
Sellam Ismail Vintage Computer
--------------------------------------------------------------------------
----
International Man of Intrigue and Danger
http://www.vintage.org
[ Old computing resources for business || Buy/Sell/Trade Vintage
mputers ]
[ and academia at www.VintageTech.com || at
http://marketplace.vintage.org ]
7:35 a.m.
New subject: Modern Tube computing
--- William Donzelli <aw288(a)osfn.org> wrote:
This is why I really can't stand the tube audio
world - there is too much
junk science, snake oil, prejudice, and false
rumors. Anything to make the
rare old stock "sound" better.
Yea, it’s quite hilarious, it seemed when modern audio
amplifiers reached a state of perfection (for all
practical purposes) all of a sudden imperfect
reproduction became high-end audio. Keeps people
buying I guess.
__________________________________
Do you Yahoo!?
Yahoo! Hotjobs: Enter the "Signing Bonus" Sweepstakes
http://hotjobs.sweepstakes.yahoo.com/signingbonus
6:27 a.m.
New subject: Modern Tube computing
Quothe ben franchuk, from writings of Tue, Jan 20, 2004 at 08:21:56PM -0700:
Vintage TUBE stuff is the latest fad of PRO-AUDIO, so
that stuff is all
dried up. You can't even get tubes from the 1940's for example: New Old
Stock 6SN7 $39 ... New sov-tech 6SN7 $9.
Ben.
Whoa! Don't confuse those involved in pro-audio (e.g., those who work
with recording, live sound reinforcement, etc.) with the golden-eared
audiophiles (typically hobbyists who don't know a balanced line from a
dBu, but who will pay small fortunes for audiophile grade cables). In
some cases, some people are involved in both. However, for the most
part, there exists a big difference.
Some of those golden-eared audiophiles are even paying unreasonably
large sums of money for hi-fi and guitar amplifiers with brand-names
such as Eico and Silvertone, which some hucksters are actually
promoting as being so-called "high-end" products. Well, we all know
what P.T. Barnum said...
--
Copyright (C) 2003 R. D. Davis The difference between humans & other animals:
All Rights Reserved an unnatural belief that we're above Nature &
rdd(a)rddavis.org 410-744-4900 her other creatures, using dogma to justify such
http://www.rddavis.org beliefs and to justify much human cruelty.
22 Jan
22 Jan
11:59 a.m.
New subject: Modern Tube computing
On Tue, 2004-01-20 at 16:03, William Donzelli wrote:
I too will have to goolg this, never hoida dat.
Have we
forgotten John Zabolitzky's MUNIAC?
No, but I haven't heard anything about it since the initial flurry of
activity. One thing I would find very interesting is making a
tube computer (or
rather just the building blocks) using technology from 1930.
Relays! Computing with 201s? EEEEaaagh...
I have a digital clock built entirely from "sensitive" relays. T-type
flip flops cascaded, decimal decoder built from steppers (yes, you could
do the whole clock w/4 steppers but what's the point). You can actually
see and hear ripple carry ripple!
It's all bakelite, nickel, redwood, fabric wire, brass thumbscrews, etc.
THe display is stacked glass sheets with numbers etched, with lamps
illuminating the edges (sound familiar). Not built yet, but I have an
old NLS display as standin.
Timebase is of course ... a pendulum! (No devices w/gain aloowed other
than relays). CdS cell for pickoff. Not built yet.
12:38 p.m.
New subject: Modern Tube computing
On 22 Jan 2004, Tom Jennings wrote:
On Tue, 2004-01-20 at 16:03, William Donzelli wrote:
I too will have to goolg this, never hoida dat.
Here's a brief write-up of VCF Europa 1.0 where John exhibited the MUNIAC:
http://www.gaby.de/helvinte.htm
John also exhibited part of the system at VCF 3.0 in 1999:
http://www.vintage.org/exhibit99.html
(4th photo down on the right)
Here's a picture of the entire system as exhibited at VCF Europa in 2000:
http://www.gaby.de/vcfpics/04.jpg
I think John put the project away for now. Hopefully he'll have time to
pick it back up and finish it.
--
Sellam Ismail Vintage Computer Festival
------------------------------------------------------------------------------
International Man of Intrigue and Danger http://www.vintage.org
[ Old computing resources for business || Buy/Sell/Trade Vintage Computers ]
[ and academia at www.VintageTech.com || at http://marketplace.vintage.org ]
Have we
forgotten John Zabolitzky's MUNIAC?
No, but I haven't heard anything about it since the initial flurry of
activity.
12:51 p.m.
New subject: Modern Tube computing
More pictures of the MUNIAC:
http://www.travelnotes.de/computer/vcfe/muniac.htm
--
Sellam Ismail Vintage Computer Festival
------------------------------------------------------------------------------
International Man of Intrigue and Danger http://www.vintage.org
[ Old computing resources for business || Buy/Sell/Trade Vintage Computers ]
[ and academia at www.VintageTech.com || at http://marketplace.vintage.org ]
20 Jan
20 Jan
2:17 p.m.
New subject: Modern Tube computing
Quothe John Lawson, from writings of Tue, Jan 20, 2004 at 03:55:27PM -0500:
And no one has had the utter madness to suggest a
replica of the Mark I
or II relay machine..... or the SSEC.... or any of the Bell System relay
computers....
Relays don't normally glow, alas, but they do make nice sounds.
My preference is for a digital computer using water instead of
electricity, but no one's mentioned one of those yet.
--
Copyright (C) 2003 R. D. Davis The difference between humans & other animals:
All Rights Reserved an unnatural belief that we're above Nature &
rdd(a)rddavis.org 410-744-4900 her other creatures, using dogma to justify such
http://www.rddavis.org beliefs and to justify much human cruelty.
2:51 p.m.
New subject: Modern Tube computing
At 20:12 20/01/2004 -0500, R. D. Davis wrote:
My preference is for a digital computer using water
instead of
electricity, but no one's mentioned one of those yet.
Even sillier...
I'd like to see someone build a replica of HEX, from the Terry Pratchet
Discworld books - from the descriptions, it's digital computer that
operates via ants running around tubes.. (I guess, get them a fine enough
bore so that the ants couldn't turn around, you could get it to work...)
Rob.
3:09 p.m.
New subject: Modern Tube computing
Quothe Rob O'Donnell, from writings of Wed, Jan 21, 2004 at 01:51:45AM +0000:
That was not a "silly" idea; it was a serious idea, and I don't see why
it isn't possible. It could have some very practical applications.
--
Copyright (C) 2003 R. D. Davis The difference between humans & other animals:
All Rights Reserved an unnatural belief that we're above Nature &
rdd(a)rddavis.org 410-744-4900 her other creatures, using dogma to justify such
http://www.rddavis.org beliefs and to justify much human cruelty.
My preference
is for a digital computer using water instead of
electricity, but no one's mentioned one of those yet.
Even sillier...
3:22 p.m.
New subject: Modern Tube computing
On Tue, 20 Jan 2004, R. D. Davis wrote:
Quothe Rob O'Donnell, from writings of Wed, Jan
21, 2004 at 01:51:45AM +0000:
That was not a "silly" idea; it was a serious idea, and I don't see why
it isn't possible. It could have some very practical applications.
One might suppose you folks have not heard of "fluidic logic"? I saw an
entire digital control system for a rocket-fuel blender/loader that used
compressed air as it's working medium.
Fluidics was a 'fad' in the late 50s, but fell out of favor due to the
precision involved in making the modules, and the fact that the tiniest
particle of dirt would cause a failure.
These devices work on aerodynamic principles - Bernoulli effects - and
boundary layer physics. For example, the flip/flop module used an internal
airfoil to guide a stream of incoming air through one of two adjacent
channels. The device was 'set' or 'reset' with a transverse stream
impingeing on the incoming 'main' stream, diverting it to the opposite
channel, where, due to capillary action and the boundary layer laws, it
stayed until it was set to the complement by another "puff" from the other
port. Bleed holes, or venturis, were then used to sense the direction of
flow.
These were generally made of Plexiglass (Perspex) and were quite
beautiful to look at... like a bunch of square prismatic blocks nestled in
a forest of black rubber tubing.
And this system would work on water as well, just the dimensions would
have to be altered to compensate for the different phsycal characteristics
of the working fluid.
Cheers
John
My
preference is for a digital computer using water instead of
electricity, but no one's mentioned one of those yet.
Even sillier... 
22 Jan
22 Jan
6:58 a.m.
New subject: Modern Tube computing
----- Original Message -----
From: "John Lawson" <jpl15(a)panix.com>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk(a)classiccmp.org>
Sent: Wednesday, January 21, 2004 2:22 AM
Subject: Re: Modern Tube computing
On Tue, 20 Jan 2004, R. D. Davis wrote:
These devices work on aerodynamic principles - Bernoulli effects - and
boundary layer physics. For example, the flip/flop module used an internal
airfoil to guide a stream of incoming air through one of two adjacent
channels. The device was 'set' or 'reset' with a transverse stream
impingeing on the incoming 'main' stream, diverting it to the opposite
channel, where, due to capillary action and the boundary layer laws,
Also utilised the Coanda effect ,where a flow of air will tend to stick to a
surface until disturbed - utilised in a range of room heaters of the same
name over here
Geoff.
it stayed until it was set to the complement by
another "puff" from the
other
port. Bleed holes, or venturis, were then used to
sense the direction of
flow.
Cheers
John
20 Jan
20 Jan
9:44 p.m.
New subject: Modern Tube computing
On 20 Jan 2004, R. D. Davis wrote:
Quothe Rob O'Donnell, from writings of Wed, Jan
21, 2004 at 01:51:45AM +0000:
That was not a "silly" idea; it was a serious idea, and I don't see why
it isn't possible. It could have some very practical applications.
It's been done before. An example (THE example?) sits in the main science
museum in England (I forgot what it's called).
--
Sellam Ismail Vintage Computer Festival
------------------------------------------------------------------------------
International Man of Intrigue and Danger http://www.vintage.org
[ Old computing resources for business || Buy/Sell/Trade Vintage Computers ]
[ and academia at www.VintageTech.com || at http://marketplace.vintage.org ]
My
preference is for a digital computer using water instead of
electricity, but no one's mentioned one of those yet.
Even sillier...
9:58 p.m.
New subject: Modern Tube computing
On Wed, 21 Jan 2004, Vintage Computer Festival wrote:
On 20 Jan 2004, R. D. Davis wrote:
BTW, it was a computer that modeled a nation's economy using water as the
analog for money. VERY cool.
--
Sellam Ismail Vintage Computer Festival
------------------------------------------------------------------------------
International Man of Intrigue and Danger http://www.vintage.org
[ Old computing resources for business || Buy/Sell/Trade Vintage Computers ]
[ and academia at www.VintageTech.com || at http://marketplace.vintage.org ]
Quothe Rob O'Donnell, from writings of Wed,
Jan 21, 2004 at 01:51:45AM +0000:
That was not a "silly" idea; it was a serious idea, and I don't see why
it isn't possible. It could have some very practical applications.
It's been done before. An example (THE example?) sits in the main science
museum in England (I forgot what it's called). My
preference is for a digital computer using water instead of
electricity, but no one's mentioned one of those yet.
Even sillier...
21 Jan
21 Jan
12:27 a.m.
New subject: Modern Tube computing
Rob O'Donnell wrote:
Even sillier...
I'd like to see someone build a replica of HEX, from the Terry Pratchet
Discworld books - from the descriptions, it's digital computer that
operates via ants running around tubes.. (I guess, get them a fine enough
bore so that the ants couldn't turn around, you could get it to work...)
Rob.
Don't forget the computer built from tinker-toy parts (wooden hubs and dowels) at some
university in, IIRC, the early eighties...
20 Jan
20 Jan
4 p.m.
New subject: Modern Tube computing
R. D. Davis wrote:
Relays don't normally glow, alas, but they do make
nice sounds.
My preference is for a digital computer using water instead of
electricity, but no one's mentioned one of those yet.
Oh, a VALVE computer ... :)
I think that digital computing using fluid logic, was air based
rather than water based.I think Scientific Amercian had something on
it as well in the 1970's.
9:42 p.m.
New subject: Modern Tube computing
On 20 Jan 2004, R. D. Davis wrote:
Quothe John Lawson, from writings of Tue, Jan 20, 2004
at 03:55:27PM -0500:
I made a rudimentary one today. I took two half glasses of water. I
needed to compute something analogous to 1 + 1, so I poured one glass into
the other and it worked! (It also turns out that whether the glass is
half full or half empty, if you combine them together then one is full and
one is empty...there's some sort of philosophical riff there).
I then kept pouring the water back and forth between each other: a
FlipFlop!
How's that? :)
--
Sellam Ismail Vintage Computer Festival
------------------------------------------------------------------------------
International Man of Intrigue and Danger http://www.vintage.org
[ Old computing resources for business || Buy/Sell/Trade Vintage Computers ]
[ and academia at www.VintageTech.com || at http://marketplace.vintage.org ]
And no one has had the utter madness to
suggest a replica of the Mark I
or II relay machine..... or the SSEC.... or any of the Bell System relay
computers....
Relays don't normally glow, alas, but they do make nice sounds.
My preference is for a digital computer using water instead of
electricity, but no one's mentioned one of those yet. 
21 Jan
21 Jan
10:33 a.m.
New subject: Modern Tube computing
I then kept pouring the water back and forth between
each other: a
FlipFlop!
Some time ago I went to a lecture on soap films, where the lecturer did
various demonstrations with such films contrained between plastic plates,
with metal pins anchoring the ends of the films. The film always forms a
local minimum of area IIRC.
Anyway, one configuration of pins had 2 stable forms of the soap film,
and could be switched from one to the other by gently blowing on it. I
pointed ou that this was essentially a flip-flop and could be used as a
memory device. I then said 'Well, I've heard of bubble memory, but this
is ridiculous' <groan>
-tony
11:09 a.m.
New subject: Modern Tube computing
[...soap films...]
Anyway, one configuration of pins had 2 stable forms of the soap
film, and could be switched from one to the other by gently blowing
on it.
Four pins at the corners of a square is a good example. You get a
structure of five pieces of film, one horizontal or veritcal through
the centre of the square, the others joining the first to the corners,
the parts meeting at 120? angles. By blowing on it horizontally or
vertically you can flip the flop into the other state.
/~\ The ASCII der Mouse
\ / Ribbon Campaign
X Against HTML mouse(a)rodents.montreal.qc.ca
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
22 Jan
22 Jan
12:03 p.m.
New subject: Modern Tube computing
On Tue, 2004-01-20 at 17:12, R. D. Davis wrote:
My preference is for a digital computer using water
instead of
electricity, but no one's mentioned one of those yet.
Umm, Frank Holmquist, who works for Ed Grothus (Los ALamos Sales Co,
505-662-5053) makes hydraulic logic elements... apparently some amazing
and beautiful water art and digital stuff. It's a weird world if you
look for it :-) Tell him I sent you.
20 Jan
20 Jan
9:44 a.m.
Holger Veit wrote:
^^^^^^^^^^^^^^^^^^^^^
Tssss. Not to discourage you, but a TTL computer with
bazillions of ICs
is not really suited for newbies, although there is plenty to learn -
namely on those nasty aspects like circuit timing, delays, glitches,
noise on power and signal lines. TTL is to a large degree analog, not digital,
circuit development ;-)
So are TUBE computers better?
You might not directly want to start with a monster
like Computer 74 or
EGO, but perhaps with a simple ALU/register device: get a 2901 circuit
from BG-Micro and play a bit with it (a breadboard with lots of switches
and LEDs is sufficient). And then extend it with a 2909/-10/-11 from the same
source and build the micro control for that toy. This is already the
basic step towards an own homegrown system.
Using LS TTL ( also from BG-MICRO ) I expect I'll need about 32 chips
for a 12 bit ALU and 32 chips for basic control. I suspect about the
same for front panel logic. Since I don't have a PROM or PAL burner
a bit slice design out of the question for me.
Ben.
2:19 p.m.
Quothe ben franchuk, from writings of Tue, Jan 20, 2004 at 01:44:52PM -0700:
So are TUBE computers better?
Theoretically, they could be faster than solid state computers.
--
Copyright (C) 2003 R. D. Davis The difference between humans & other animals:
All Rights Reserved an unnatural belief that we're above Nature &
rdd(a)rddavis.org 410-744-4900 her other creatures, using dogma to justify such
http://www.rddavis.org beliefs and to justify much human cruelty.
4:58 p.m.
So are TUBE
computers better?
Theoretically, they could be faster than solid state computers.
5:32 p.m.
How so? Are flight times between cathode and anode
shorter than
transit times between emitter and collector? Or is there some other
effect going on that I've missed?
That is basically it - flying electrons don't have other junk to bump into,
like they do in solid state.
This transit time problem pretty much killed standard tubes in the microwave
region - to get the things to work properly, the cathode-grid-plate
spacing had to be extremely small. Small, however, is not good for other
reasons, namely power and capacitance.
William Donzelli
aw288(a)osfn.org
6:37 p.m.
Okay, I'm confused. I wrote (of vacuum tube computers being, in
principle, faster than transistor)
> How so? Are flight times between cathode and
anode shorter than
> transit times between emitter and collector? Or is there some other
> effect going on that I've missed?
Then William Donzelli <aw288(a)osfn.org> replied
That is basically it - flying electrons don't have
other junk to bump
into, like they do in solid state.
This (and the original note) implies that tubes are inherently faster
in this way. But then...
This transit time problem pretty much killed standard
tubes in the
microwave region - to get the things to work properly, the
cathode-grid-plate spacing had to be extremely small.
...this sounds as though it's the converse; this sounds as though tube
transit times are inconveniently _high_.
So I don't get it.
/~\ The ASCII der Mouse
\ / Ribbon Campaign
X Against HTML mouse(a)rodents.montreal.qc.ca
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
7:04 p.m.
...this sounds as though it's the converse; this
sounds as though tube
transit times are inconveniently _high_.
So I don't get it.
In theory, computing tubes would not use much power at all, so one could
space the cathode-grid-plate structure very close, and get the super
speeds due to the very short distances involved. Imagine tubes as small
today's transistors - the speeds would be fantastic, as the electrons
would speed right thru the micron of vacuum much faster that a micron of
silicon.
However, the real world hits the tubes far before element spacing gets
this tiny. The breakdown voltage between the cathode and plate probably
hits first - too close, and the electrons will simply jump over on their
own, like a spark. Likewise, with very short distances, capacitance will
become a big issue. The best conventional tubes top out somewhere around 4
GHz, and these were only good for small amounts of power. They were a
bastard to build, as well.
So, in a perfect world, with no capacitance or power problems, tubes
would be great.
William Donzelli
aw288(a)osfn.org
10:31 p.m.
Hi,
William Donzelli <aw288(a)osfn.org> said:
See http://www.triodeel.com/area51.htm - if it still works.
--
Cheers,
Stan Barr stanb(a)dial.pipex.com
The future was never like this!
...this sounds
as though it's the converse; this sounds as though tube
transit times are inconveniently _high_.
So I don't get it.
In theory, computing tubes would not use much power at all, so one could
space the cathode-grid-plate structure very close, and get the super
speeds due to the very short distances involved. Imagine tubes as small
today's transistors - the speeds would be fantastic, as the electrons
would speed right thru the micron of vacuum much faster that a micron of
silicon.
However, the real world hits the tubes far before element spacing gets
this tiny. The breakdown voltage between the cathode and plate probably
hits first - too close, and the electrons will simply jump over on their
own, like a spark. Likewise, with very short distances, capacitance will
become a big issue. The best conventional tubes top out somewhere around 4
GHz, and these were only good for small amounts of power. They were a
bastard to build, as well.
So, in a perfect world, with no capacitance or power problems, tubes
would be great.
21 Jan
21 Jan
4:34 a.m.
See http://www.triodeel.com/area51.htm - if it still
works.
About every three years, micro tubes are announced as a huge breakthough
that will revolutionize the computer industry. I think we are due for
another announcement.
William Donzelli
aw288(a)osfn.org
6:57 a.m.
New subject: Looking for a parallel keyboard
Tony and the gang - I just spotted this thread:
> Does this mean there might be a small market for a
PS/2 to parallel
> keyboard converter? I can envision two modes, output only, where ASCII
That was my first thought -- a PC keyboard + a
microcontroller. Shouldn't
be too hard to write the firmware....
Of course, not being the business type, I would
probably have GPLed it,
not tried to sell it ;-)
-tony
fwiw I did that a few years ago, using an 87C750 (almost a classic itself by
now) and a hacked up version of Phillips AN434 "Connecting a PC keyboard to
the I2C bus". Photos here
http://www.widomaker.com/~cswiger/dscn1468s.jpg
http://www.widomaker.com/~cswiger/dscn1472s.jpg
it works for me. In the blurry photo, the leftmost bit is the keypress strobe,
goes off when a key is pressed; and a c/r was hit before the photo. If someone
needs one I can (maybe) burn an OTP chip but you'll have to build the circuit.
Modified firmware free. Has shift, lock and ctrl, etc.
--Chuck
22 Jan
22 Jan
2:27 p.m.
On Wed, 2004-01-21 at 07:34, William Donzelli wrote:
And it's a great announcement every time!
Now all we need are some sample devices :-)
See
http://www.triodeel.com/area51.htm - if it still works.
About every three years, micro tubes are announced as a huge breakthough
that will revolutionize the computer industry. I think we are due for
another announcement.
3:39 p.m.
And it's a great announcement every time!
Now all we need are some sample devices :-)
Just like the announcement that core memory is coming back in chip form.
Well, maybe. At least someone shipped samples about ten years ago
(Ramtron? I have one somewhere).
William Donzelli
aw288(a)osfn.org
23 Jan
23 Jan
10:11 a.m.
On Thu, 2004-01-22 at 18:39, William Donzelli wrote:
(I remember those things from the trade rags, never saw one... 16 bits
in a DIP package for non-volatile config storage. Not too shabby, but
gone the way of bubble, core, etc...)
And it's a
great announcement every time!
Now all we need are some sample devices :-)
Just like the announcement that core memory is coming back in chip form.
Well, maybe. At least someone shipped samples about ten years ago
(Ramtron? I have one somewhere).
21 Jan
21 Jan
4:48 a.m.
On Wed, 21 Jan 2004, Stan Barr wrote:
Too bad all the links he refers to are dead.
--
Sellam Ismail Vintage Computer Festival
------------------------------------------------------------------------------
International Man of Intrigue and Danger http://www.vintage.org
[ Old computing resources for business || Buy/Sell/Trade Vintage Computers ]
[ and academia at www.VintageTech.com || at http://marketplace.vintage.org ]
So, in a
perfect world, with no capacitance or power problems, tubes
would be great.
See http://www.triodeel.com/area51.htm - if it still works.
22 Jan
22 Jan
2:25 p.m.
I do think one possible area of unexplored power-microwave tube
development might be a space-borne sun-heated filament magnetron or
something. Big collector on one side, fins on the other.
I'm not a tube nut, I just researched them for various projects. I don't
believe in illuminati conspiracies and 100 mpg carburetors either!
tomj
On Wed, 2004-01-21 at 01:31, Stan Barr wrote:
Hi,
William Donzelli <aw288(a)osfn.org> said:
See http://www.triodeel.com/area51.htm - if it still works.
...this
sounds as though it's the converse; this sounds as though tube
transit times are inconveniently _high_.
So I don't get it.
In theory, computing tubes would not use much power at all, so one could
space the cathode-grid-plate structure very close, and get the super
speeds due to the very short distances involved. Imagine tubes as small
today's transistors - the speeds would be fantastic, as the electrons
would speed right thru the micron of vacuum much faster that a micron of
silicon.
However, the real world hits the tubes far before element spacing gets
this tiny. The breakdown voltage between the cathode and plate probably
hits first - too close, and the electrons will simply jump over on their
own, like a spark. Likewise, with very short distances, capacitance will
become a big issue. The best conventional tubes top out somewhere around 4
GHz, and these were only good for small amounts of power. They were a
bastard to build, as well.
So, in a perfect world, with no capacitance or power problems, tubes
would be great.
2:28 p.m.
Tom Jennings wrote:
I do think one possible area of unexplored
power-microwave tube
development might be a space-borne sun-heated filament magnetron or
something. Big collector on one side, fins on the other.
Now what I want is a solar powered anti-matter production
plant in solar orbit. Ben.
23 Jan
23 Jan
12:24 a.m.
On Fri, 2004-01-23 at 01:15, Tom Jennings wrote:
I do think one possible area of unexplored
power-microwave tube
development might be a space-borne sun-heated filament magnetron or
something. Big collector on one side, fins on the other.
I'm not a tube nut, I just researched them for various projects. I don't
believe in illuminati conspiracies and 100 mpg carburetors either!
tomj
100mpg carburettors do exist. I had one, and got pretty close to
100mpg. Bog standard Keihin slide carb on my Honda CD200, gentle
riding.
Gordon.
22 Jan
22 Jan
2:20 p.m.
On Tue, 2004-01-20 at 22:04, William Donzelli wrote:
In theory, computing tubes would not use much power at
all, so one could
space the cathode-grid-plate structure very close, and get the super
speeds due to the very short distances involved.
Well with gate::emitter separated by fractions of a micron, I don't know
how any theoretical tube would improve on that. I know the path an
electron or hole travels is somewhat brownian, hence longer, still, it's
pretty small!
Also, you need a source of electrons; in semiconductor stuff it's done
at the factory (more or less) by creating doped materials with low
quantum thresholds, such that mere voltage difference is enough to move
them around. With tubes, you gotta heat the miserable things up to fling
electrons loose, and to get enough electrons, yuo need a lot of metal
(though as you point out part this would scale very well).
So, in a perfect world, with no capacitance or power
problems, tubes
would be great.
Another way to think of tubes is big delicate airy capacitors with heat
in the middle. They sure look cool though, and things like big 811's,
the very idea that the carbon plate is glowing cherry red merely from
SLOW ELECTRON BOMBARDMENT! is Very Cool!
3:42 p.m.
Another way to think of tubes is big delicate airy
capacitors with heat
in the middle. They sure look cool though, and things like big 811's,
the very idea that the carbon plate is glowing cherry red merely from
SLOW ELECTRON BOMBARDMENT! is Very Cool!
If you think 811s are big, you need to get out more.
William Donzelli
aw288(a)osfn.org
21 Jan
21 Jan
3:09 a.m.
On Tue, Jan 20, 2004 at 08:14:43PM -0500, R. D. Davis wrote:
Quothe ben franchuk, from writings of Tue, Jan 20,
2004 at 01:44:52PM -0700:
No.
This would be true only if you'd manage to reach as high integration
as with semiconductors (and even then, you'd have a hard job competing
with III-V semiconductors and recent fast silicon technologies).
Surely, you can build tubed HF transmitter devices that work in the some
GHz range, and such are still in use when it comes to high power, but the
limiting factor is the size of the complete system. It is useless to
build a tube gate or flip flop that could switch at some GHz when two of
such circuits are connected by a wire that is much longer than the time
a light or EM pulse spreads. Note that an EM wave can move (in vacuum)
about 30cm/nsec. In modern semiconductor circuits, this has already been
shown to be a limiting factor for reachable speed. I don't know what size
of a building you'd need, what the power consumption would be, if you'd
consider to construct a "fast" tube-based Pentium even at some lousy 33MHz
speed (I said Pentium, not Pentium IV two magnitudes faster). Definitely,
you wouldn't be able to put it into a cube of 30 cm^3 to at least guarantee
a gate-to-gate delay of 1ns.
Nostalgy left alone, tubes may still have their application fields, such
as ripping off crazy HiFi freaks (although you could do better with gold-
plated loudspeaker cables there), but computers are not even by far among
them. In the 50s, with Ge-PNP transistors with transit frequencies of
some 100kHz, they were appropriate.
Holger
So are TUBE computers better?
Theoretically, they could be faster than solid state computers.
22 Jan
22 Jan
12:09 p.m.
On Tue, 2004-01-20 at 17:14, R. D. Davis wrote:
Quothe ben franchuk, from writings of Tue, Jan 20,
2004 at 01:44:52PM -0700:
Aww come on, tubes suck!
First off, a cloud of decidedly non-relativistic electrons has to travel
a SIGNIFICANT fraction of an inch. The elements are HUGE massive
capacitors (even nuvistors have at least a couple pF element to
element). Read about the Miller Effect. It's hard to make tubes run
(logic) at one motorcycle. The element areas are such that electrons
leave from different distances on their way to the plate that it causes
essentially multipath problems!
Plus, their physical size makes circuit-length a major problem.
Speed-of-light-in-metal is a limit speed factor in today's chips, and is
one of the things that Seymour figured out how to deal with in his Cray
machines. Even tubes consuming 1/100th the volume are still huge. Never
mind filaments. Filaments! Vff!
Tubes are fun, but they're not really any good for anything useful (and
I don't want to hear from any audio tube nuts :-)
tomj
So are TUBE computers better?
Theoretically, they could be faster than solid state computers.
12:48 p.m.
Quothe Tom Jennings, from writings of Thu, Jan 22, 2004 at 02:59:27PM -0800:
Aww come on, tubes suck!
No, they don't. They get sucked so that they contain a vacuum.
Tubes are fun, but they're not really any good for
anything useful (and
I don't want to hear from any audio tube nuts :-)
Aside from them being useful for audio equipment, like the ring
modulator that I'm building that uses tubes, they're also very useful
in electronic test equipment (e.g., VTVMs, TEK 500 series 'scopes,
etc.), etc. Tube circuits are also fun to create and work with; more
fun than solid-state.
--
Copyright (C) 2003 R. D. Davis The difference between humans & other animals:
All Rights Reserved an unnatural belief that we're above Nature &
rdd(a)rddavis.org 410-744-4900 her other creatures, using dogma to justify such
http://www.rddavis.org beliefs and to justify much human cruelty.
1:57 p.m.
Tom Jennings wrote:
Tubes are fun, but they're not really any good for
anything useful (and
I don't want to hear from any audio tube nuts :-)
You are too late for that ... LONG LIVE TUBE AUDIO! ...
PS. I find the many modern tube amps, have crappy power supplies
too as well as tube computers. The tube amp I am building has
a real regulated ( solid state ) power supply. Ben.
3:10 p.m.
ben franchuk wrote:
Tom Jennings wrote:
Amen to that. My current motherboard (AOpen AX4B-533TUBE) has a vacuum
tube audio pre-amp on it. You also can't beat tubes for high output RF
amplifiers. My Ham rig consists of a Kenwood TS520 (3 tubes) hooked to
an Ameritron 400W amplifier (4 Tubes).
Tubes are fun, but they're not really any
good for anything useful (and
I don't want to hear from any audio tube nuts :-)
You are too late for that ... LONG LIVE TUBE AUDIO! ...
10:13 p.m.
Hi,
Tom Jennings <tomj(a)wps.com> said:
Tubes are fun, but they're not really any good for
anything useful (and
I don't want to hear from any audio tube nuts :-)
They're still very useful for radio transmitters - especially really
big ones.
--
Cheers,
Stan Barr stanb(a)dial.pipex.com
The future was never like this!
20 Jan
20 Jan
9:52 p.m.
On Tue, Jan 20, 2004 at 01:44:52PM -0700, ben franchuk wrote:
And this can be had with just three 2901. Surely you could go the hard way
and assemble the ALU out of three 74181, some 74175s for the accu register,
and some glue logic as I did as well 20 years or more ago. Besides more
power consumption, more points of failure and larger boards you don't
gain very much.
Holger Veit wrote:
^^^^^^^^^^^^^^^^^^^^^
The effects occur if you run the system at "high" speed - the definition of
"high" depends on the quality of construction. Wire-Wrap and breadbaords are
susceptible to cross-talk - wire lengths are a factor unless you run
almost static (the mentioned breadboard with a 2901 controlled by switches,
debounced with RS-Nand-FFs, is such a case.
Tube computers are larger and thus effectively run at by far lower speeds,
but they are likewise analog technology as discrete flip-flops and
RTL logic.
Tssss. Not to discourage you, but a TTL computer
with bazillions of ICs
is not really suited for newbies, although there is plenty to learn -
namely on those nasty aspects like circuit timing, delays, glitches,
noise on power and signal lines. TTL is to a large degree analog, not digital,
circuit development ;-)
So are TUBE computers better? You might not
directly want to start with a monster like Computer 74 or
EGO, but perhaps with a simple ALU/register device: get a 2901 circuit
from BG-Micro and play a bit with it (a breadboard with lots of switches
and LEDs is sufficient). And then extend it with a 2909/-10/-11 from the same
source and build the micro control for that toy. This is already the
basic step towards an own homegrown system.
Using LS TTL ( also from BG-MICRO ) I expect I'll need about 32 chips
for a 12 bit ALU and 32 chips for basic control. I suspect about the same for front panel logic. Since I don't have a
PROM or PAL burner
a bit slice design out of the question for me.
You don't need any PROM or PAL burner. You might put your microcode into
a fast cache RAM (from old PCs) and have some small serial shift register
logic to load it from a PC parallel port (a similar circuit BTW also allows
to program EEPROMs and flash ROMs from the parallel port). You would not
use PALs, but CPLDs which are bit serial programmable through a JTAG port -
provided you accept the use of modern circuits to replace TTL graveyards.
The first microcode I built for such a homegrown CPU I did with a diode matrix
after finding out that the 74S188 PROMs that were available those days were
too expensive for bug fixing (they could BTW also be programmed manually with
a simple socket and a number of switches).
Holger
7487
days inactive
7491
days old
65 comments
21 participants
participants (21)
-
ard＠p850ug1.demon.co.uk -
aw288＠osfn.org -
bfranchuk＠jetnet.ab.ca -
chrisc＠addpower.com -
classiccmp.org＠irrelevant.fsnet.co.uk -
cmcnabb＠4mcnabb.net -
cswiger＠widomaker.com -
esharpe＠uswest.net -
geoffreythomas＠onetel.net.uk -
gkicomputers＠yahoo.com -
gordonjcp＠gjcp.net -
hilpert＠cs.ubc.ca -
holger.veit＠ais.fhg.de -
jpl15＠panix.com -
jwest＠classiccmp.org -
mouse＠Rodents.Montreal.QC.CA -
mtapley＠swri.edu -
rdd＠rddavis.org -
stanb＠dial.pipex.com -
tomj＠wps.com -
vcf＠siconic.com