7 Dec
2007
7 Dec
'07
9:20 p.m.
A few days ago, someone asked about TRADIC. Tonight I stumbled upon a mention of this
computer on page 204 of the book "Crystal Fire" (Michael Riordan and Lillian
Hoddeson, 1997). The authors wrote, "In January 1954 Whippany engineers built a fully
transistorized computer for the Air Force. Called TRADIC (for TRAnsistorized DIgital
Computer), it used 700 point-contact transistors and more than 10,000 germanium crystal
rectifiers in its circuits. Capable of performing a million logical operations every
second, TRADIC was the first completely solid-state computer; it approached the speed of
computers based on vacuum tubes."
By "Whippany" they meant the Whippany, N.J. location of Bell Labs.
7 Dec
7 Dec
11:03 p.m.
Evan wrote:
A few days ago, someone asked about TRADIC. Tonight I stumbled upon a mention of this
computer on page 204 of the book "Crystal Fire" (Michael Riordan and Lillian
Hoddeson, 1997). The authors wrote, "In January 1954 Whippany engineers built a fully
transistorized computer for the Air Force. Called TRADIC (for TRAnsistorized DIgital
Computer), it used 700 point-contact transistors and more than 10,000 germanium crystal
rectifiers in its circuits. Capable of performing a million logical operations every
second, TRADIC was the first completely solid-state computer; it approached the speed of
computers based on vacuum tubes."
By "Whippany" they meant the Whippany, N.J. location of Bell Labs.
TRADIC is a well-known machine in the history of computing.
A little googling does bring up a lot of refs.
(I think there's another experimental/proof-of-concept machine commonly
mentioned as being one of the first transistorised machines (something from
Philco perhaps?).)
Another anecdote I ran across a few years ago in the IEEE AotHoC (which I wish
had kept a ref to), was of an IBM engineer working on one of the first
transistorised designs in the mid-50s (ECL IIRC), telling the story of how he
chose 5 Volts for the logic supply.
8 Dec
8 Dec
1:11 p.m.
On Saturday 08 December 2007 02:03, Brent Hilpert wrote:
Another anecdote I ran across a few years ago in the
IEEE AotHoC (which I
wish had kept a ref to), was of an IBM engineer working on one of the first
transistorised designs in the mid-50s (ECL IIRC), telling the story of how
he chose 5 Volts for the logic supply.
I've wondered often how the various supply voltages I run into get chosen...
Care to elaborate on that a bit?
--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, ?a critter that can
be killed but can't be tamed. ?--Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
3 p.m.
New subject: 5V and early ICs / was Re: TRADIC
"Roy J. Tellason" wrote:
On Saturday 08 December 2007 02:03, Brent Hilpert wrote:
I wish I could, but without the ref ...
For what I do recall, the suggestion in the article was that this was the first
use or choice of 5V for the logic supply, although I can't remember whether
the 5V was a choice of whim from which other design parameters followed, or
whether the 5V was an engineered outcome.
Even though I found it an interesting anecdote however, there were, or would still
be (IMHO) a couple of historical links necessary to show that there was a causal
connection/influence to the 5V supply standard of the later DTL & TTL integrated
circuit families.
The earliest reference for real ICs I have is a TI product catalog from 1965.
The first products mentioned are the "NEW! Series 54 TTL"
(SN5400,5410,5420..5470)
(4.5 to 5.5V of course). The commercial 74xx versions apparently followed a little
later.
Five other digital IC families are mentioned:
- Series 53 Modified DTL (SN53x) (different from 'standard' DTL) (3 to 4V)
- Series 51 RCTL (SN51x) (3 to 6V)
- Series 51R (severe environment versions of 51)
- Minuteman Series DTL (SN337A..) (produced for the Minuteman II missile)
(+6V & -3V)
- Low-power RTL (SN7xx..) (would become one of the 'standard' RTL families)
(3V)
A Fairchild catalog from 1966 presents standard 900-series DTL and some 9000 series
TTL with 5V supplies.
.. and I may be repeating myself on the list, or this may be more well-known
than I realise, but another somewhat novel point that is apparent from the
TI catalog is that the "SN" prefix on Texas Instruments ICs is short for
"Semiconductor Network", an alternative phrase before "Integrated
Circuit" became
the more accepted phrase.
Another anecdote I ran across a few years ago in
the IEEE AotHoC (which I
wish had kept a ref to), was of an IBM engineer working on one of the first
transistorised designs in the mid-50s (ECL IIRC), telling the story of how
he chose 5 Volts for the logic supply.
I've wondered often how the various supply voltages I run into get chosen...
Care to elaborate on that a bit?
9:36 p.m.
New subject: 5V and early ICs / was Re: TRADIC
On Saturday 08 December 2007 18:00, Brent Hilpert wrote:
"Roy J. Tellason" wrote:
Hm, I had no idea that stuff was around that early on. My first encounter
with those parts was around 1970, and I still have that particular databook,
which had the standard parts and the H and L variants.
On Saturday 08 December 2007 02:03, Brent Hilpert
wrote:
I wish I could, but without the ref ...
For what I do recall, the suggestion in the article was that this was the
first use or choice of 5V for the logic supply, although I can't remember
whether the 5V was a choice of whim from which other design parameters
followed, or whether the 5V was an engineered outcome.
Even though I found it an interesting anecdote however, there were, or
would still be (IMHO) a couple of historical links necessary to show that
there was a causal connection/influence to the 5V supply standard of the
later DTL & TTL integrated circuit families.
The earliest reference for real ICs I have is a TI product catalog from
1965. The first products mentioned are the "NEW! Series 54 TTL"
(SN5400,5410,5420..5470) (4.5 to 5.5V of course). The commercial 74xx
versions apparently followed a little later. Another anecdote I ran across a few years ago in
the IEEE AotHoC (which
I wish had kept a ref to), was of an IBM engineer working on one of the
first transistorised designs in the mid-50s (ECL IIRC), telling the
story of how he chose 5 Volts for the logic supply.
I've wondered often how the various supply voltages I run into get
chosen...
Care to elaborate on that a bit? Five other digital IC families are mentioned:
- Series 53 Modified DTL (SN53x) (different from 'standard' DTL) (3 to
4V) - Series 51 RCTL (SN51x) (3 to 6V)
- Series 51R (severe environment versions of 51)
- Minuteman Series DTL (SN337A..) (produced for the Minuteman II missile)
(+6V & -3V)
- Low-power RTL (SN7xx..) (would become one of the 'standard' RTL
families) (3V)
A Fairchild catalog from 1966 presents standard 900-series DTL and some
9000 series TTL with 5V supplies.
I've never heard of most of those, though I think I do have some mention of
those 9xx/9xxxx numbers in my parts pages. I also have some more tech info
on hand here that I've downloaded but haven't plowed through yet or installed
into the charts...
.. and I may be repeating myself on the list, or this
may be more
well-known than I realise, but another somewhat novel point that is
apparent from the TI catalog is that the "SN" prefix on Texas Instruments
ICs is short for "Semiconductor Network", an alternative phrase before
"Integrated Circuit" became the more accepted phrase.
Well *I* never heard about that one before, and I've wondered for some time
what the heck it stood for. :-)
--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, ?a critter that can
be killed but can't be tamed. ?--Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
9 Dec
9 Dec
2:58 a.m.
New subject: 5V and early ICs / was Re: TRADIC
On Sat, 8 Dec 2007, Brent Hilpert wrote:
The earliest reference for real ICs I have is a TI
product catalog from
1965. The first products mentioned are the "NEW! Series 54 TTL"
(SN5400,5410,5420..5470) (4.5 to 5.5V of course). The commercial 74xx
versions apparently followed a little later.
There are earlier ones. As you might know, the SN74xx family started with
the SN740, SN741 etc. (only three digits!). After TI expanded this family
they went over to four digit part numbers and just appended a zero to the
already available parts. Thus the SN740 became the SN7400, the SN741 the
SN7410 and so on. I have a TI databook somewhere that has the old part
numbers.
Christian
11:48 a.m.
New subject: 5V and early ICs / was Re: TRADIC
On Sunday 09 December 2007 05:58, Christian Corti wrote:
There are earlier ones. As you might know, the SN74xx
family started with
the SN740, SN741 etc. (only three digits!). After TI expanded this family
they went over to four digit part numbers and just appended a zero to the
already available parts. Thus the SN740 became the SN7400, the SN741 the
SN7410 and so on. I have a TI databook somewhere that has the old part
numbers.
That's the first I've ever heard that one, which explains the somewhat
puzzling number sequences, and how they got that way...
--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, ?a critter that can
be killed but can't be tamed. ?--Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
12:35 p.m.
New subject: 5V and early ICs / was Re: TRADIC
Christian Corti wrote:
On Sat, 8 Dec 2007, Brent Hilpert wrote:
Well, that makes sense to some degree, the 740 numbers conceivably following
on the 730 RTL numbers, but I have never seen or heard as you suggest.
It sounds a little more like a convolution of TI's developments. The
1965 catalog (which appears to be a full-line catalog) declares the 5400
series as new, it includes earlier/older IC series, and makes no mention of
either 740 or 7400 numbers or series. The implication from this catalog is
that TI went to the 4-digit numbers with the original and initial release
of the 5400 series. All the mentioned 5400 numbers end in 0.
A 1969 TI 54/74 series TTL databook makes no mention of 3-digit 740 numbers
in the cross-refs, although it does mention another TTL series with some 3-digit
numbers (SNG1xx,SNG2xx,SNF1xx,SNF2xx). If you find your databook I'd be
interested in hearing details.
The earliest reference for real ICs I have is a
TI product catalog from
1965. The first products mentioned are the "NEW! Series 54 TTL"
(SN5400,5410,5420..5470) (4.5 to 5.5V of course). The commercial 74xx
versions apparently followed a little later.
There are earlier ones. As you might know, the SN74xx family started with
the SN740, SN741 etc. (only three digits!). After TI expanded this family
they went over to four digit part numbers and just appended a zero to the
already available parts. Thus the SN740 became the SN7400, the SN741 the
SN7410 and so on. I have a TI databook somewhere that has the old part
numbers. 
11 Dec
11 Dec
2:48 a.m.
New subject: 5V and early ICs
On Sun, 9 Dec 2007, Brent Hilpert wrote:
Well, that makes sense to some degree, the 740 numbers
conceivably following
on the 730 RTL numbers, but I have never seen or heard as you suggest.
It sounds a little more like a convolution of TI's developments. The
1965 catalog (which appears to be a full-line catalog) declares the 5400
series as new, it includes earlier/older IC series, and makes no mention of
either 740 or 7400 numbers or series. The implication from this catalog is
that TI went to the 4-digit numbers with the original and initial release
of the 5400 series. All the mentioned 5400 numbers end in 0.
It seems that I was wrong with the three-digit 74 series. Ok, I will
simply list all TI digital IC series that I have found in two old data
books:
1. American Miniature and Microminiature Electronic Assemblies Data Annual
(1963-64)
Series 51 RCTL logic: SN510, SN511, SN512 and SN513. (and that's it...)
2. Semiconductor Integrated Circuits (Texas Instruments, August 1965)
Series 51 and 51R RCTL:
Three-digit numbers, newer parts have four-digits, e.g.
SN510, SN5101
SN511, SN5111, SN5112, SN5113
Series 53 and 73 Modified-DTL:
This is were TI seems to have introduced the four-digit numbers, e.g.
SN530 / SN7300
SN531 / SN7310
Apparently there was no three-digit 73 series, but a SN530 and SN7300
are identical parts (except for the temperature range)
Minuteman series DTL (e.g. SN337A, SN341A, SN359A)
Series 54 and 74 TTL:
Has always had four-digit numbers (SN5400, SN7400)
Series 70 ECL (SN7000, SN7001)
Series 1500 and 1580 DTL
Series 1700 RTL
Christian
13 Dec
13 Dec
12:38 a.m.
New subject: 5V and early ICs
Below I have interspersed the correlations from my 1965 TI catalog
(TI "1965 Semiconductor and Components Catalog") with further
comment following:
Christian Corti wrote:
2. Semiconductor Integrated Circuits (Texas
Instruments, August 1965)
Series 51 and 51R RCTL:
Three-digit numbers, newer parts have four-digits, e.g.
SN510, SN5101
SN511, SN5111, SN5112, SN5113
- yes
Series 53 and 73 Modified-DTL:
This is were TI seems to have introduced the four-digit numbers, e.g.
SN530 / SN7300
SN531 / SN7310
Apparently there was no three-digit 73 series, but a SN530 and SN7300
are identical parts (except for the temperature range)
- SN53x are
mentioned
- Series 73 industrial modified-DTL are mentioned in a special
late-additions fold out attached to the back.
The SN72x/73x RTL series is apparently distinct from these.
Minuteman series DTL (e.g. SN337A, SN341A, SN359A)
- yes
Series 54 and 74 TTL:
Has always had four-digit numbers (SN5400, SN7400)
-
5400,5410,5420,..5470 listed
- 74xx not mentioned
Series 70 ECL (SN7000, SN7001)
- not
mentioned
Series 1500 and 1580 DTL
- not
mentioned
Series 1700 RTL
- not mentioned
My 1965 catalog does not specify a month, although it lists several
other TI publications with the latest publication date being Mar 65.
It would appear the catalog I have was but a few months before the one
you have. Between the two of them it would seem reasonable to conclude
the 7400 series was introduced in the summer of 1965, a few months after
the 5400 series.
1. American Miniature and Microminiature Electronic
Assemblies Data Annual
If you are ever inclined to produce scans of these, I'd
be interested; someday
I'd like to start compiling a list of the numerous early IC series.
--
On another note, the catalog I have also lists "gallium-arsenide infrared
sources" and "solid-state light emitters" (for the latter: "Spectral
emission
matched to silicon light-sensors", no mention is made of use as a visible
indicator).
5993
days inactive
5998
days old
9 comments
5 participants
participants (5)
-
cc＠informatik.uni-stuttgart.de -
corticn＠linuxserv.home -
evan＠snarc.net -
hilpert＠cs.ubc.ca -
rtellason＠verizon.net