8 Aug
2008
8 Aug
'08
1:03 p.m.
I don't mean to repost, but I had not seen this posted as yet:
http://www.computerworld.com/action/article.do?command=viewArticleBasic&…
--
Jim Brain, Brain Innovations (X)
brain at jbrain.com
Dabbling in WWW, Embedded Systems, Old CBM computers, and Good Times!
Home: http://www.jbrain.com
8 Aug
8 Aug
3:47 p.m.
On 8 Aug 2008 at 12:03, Jim Brain wrote:
I don't mean to repost, but I had not seen this
posted as yet:
http://www.computerworld.com/action/article.do?command=viewArticleBasic&…
Doesn't the "home computer" (not the Honeywell "kitchen computer")
go
back at least as far as 1968? Wasn't there an outfit that marketed a
box with two tape drives and a display in it for home use? Not a
computer, per se, but an information storage and retrieval device. A
name like "Viacon"? My brain has too many dustbunnies to remember
much more, but I recall reading about it a long time ago.
Perhaps one of the digital archaeologists on the list remembers.
Cheers,
Chuck
9:21 p.m.
On Friday 08 August 2008 13:03, Jim Brain wrote:
I don't mean to repost, but I had not seen this
posted as yet:
http://www.computerworld.com/action/article.do?command=viewArticleBasic&…
icleId=9111341&source=NLT_AM&nlid=1
An interesting read. One thing that jumped out at me when I got to the second
page of the article was how wide that screen was compared to its height. Do
you know offhand what the screen was, in terms of characters?
The other thing that I took some notice of was mention of an 18-pin chip
package as "recently developed" at one point in the story, and later on the
mention of the 40-pin package as being then available. When I first started
to take notice of chips at all, most of them were 14 or 16 pins, and those
were mostly op amps. This was around 1970, I think. And it was only later
on that I started seeing different sizes, but I never really knew all that
much about which sizes came out when, or how limiting some of the earlier
package restrictions were. What was up with that, anyway? Why were the
earlier chips so limited in terms of packaging?
--
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
9:44 p.m.
Roy J. Tellason wrote:
The other thing that I took some notice of was mention
of an 18-pin chip
package as "recently developed" at one point in the story, and later on the
mention of the 40-pin package as being then available.
The 18 pin-per-package limit claims regarding the 4004 and 8008 have
been around for a long time, and I think they came from interviews with
Intel's founders and/or early employees, but I think they're factually
incorrect, at least as commonly stated.
The 24-pin DIP was very well established by 1968, and was already used
by TI at that time. There were certainly higher pin-count packages at
that time also. I'm not sure about the 40-pin DIP, but in 1969
Fairchild was shipping at least one memory chip in a 36-pin DIP, though
that particular package never became popular.
Possibly whatever specific company Intel was contracting with to supply
lead frames and ceramic packages didn't yet offer higher pin count
packages, but they obviously were available from some vendors since
other semiconductor companies like Fairchild and TI were using them.
Eric
9:58 p.m.
On Friday 08 August 2008 21:44, Eric Smith wrote:
Roy J. Tellason wrote:
My first TTL databook was from TI, and I believe there was some small number
of 24-pin devices in there, like the 74154 and similar. (Hope I'm
remembering right here. :-)
The other thing that I took some notice of was
mention of an 18-pin chip
package as "recently developed" at one point in the story, and later on
the mention of the 40-pin package as being then available.
The 18 pin-per-package limit claims regarding the 4004 and 8008 have
been around for a long time, and I think they came from interviews with
Intel's founders and/or early employees, but I think they're factually
incorrect, at least as commonly stated.
The 24-pin DIP was very well established by 1968, and was already used
by TI at that time. There were certainly higher pin-count packages at that
time also. I'm not
sure about the 40-pin DIP, but in 1969 Fairchild was shipping at least one
memory chip in a 36-pin DIP, though that particular package never became
popular.
There were also those "flatpacks" which I never could figure out. A precursor
to surface mount? Something else?
Possibly whatever specific company Intel was
contracting with to supply
lead frames and ceramic packages didn't yet offer higher pin count
packages, but they obviously were available from some vendors since
other semiconductor companies like Fairchild and TI were using them.
I guess a large part of it was the machinery to handle that stuff too,
besides the lead frames themselves. I don't know much about that stuff so I
don't know if any given production machinery would be adaptable to many
different sizes or if you'd need a different bigger machine to handle them.
If the latter was the case then I can see where they might be reluctant if
there weren't a lot of demand.
--
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
11:16 p.m.
On Friday 08 August 2008 23:04, bfranchuk at jetnet.ab.ca wrote:
Roy J. Tellason wrote:
Aside from one of my databooks that has pictures on the cover, I can't say I
have either, though it's possible that I did and forgot. There were times
early on when I was starting to get to know that stuff that I momentarily
considered the possibility of using some of it, and I decided pretty quick
that it just flat-out wasn't worth the hassle.
--
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
There were also those "flatpacks" which
I never could figure out. A
precursor to surface mount? Something else?
I would guess so, but I've never seen a flat pack in the flesh myself so
I can't say. 
11:55 p.m.
From: bfranchuk at jetnet.ab.ca
Roy J. Tellason wrote:
Hi
The only use I saw for the flat packs was while I was in the
service. The boards had multiple layers and the leads went
directly into the layers. In otherwords they were not surface
mounted. The boards didn't have vias to connect layers. It
was all chip to chip. There was a hole through the board layers for
each flat pack.
Reworking failed boards was a challenge because one had to
unsolder and peel the layers back from the IC with conformal
coat on every thing.
Dwight
_________________________________________________________________
Get Windows Live and get whatever you need, wherever you are. Start here.
http://www.windowslive.com/default.html?ocid=TXT_TAGLM_WL_Home_082008
There were also those "flatpacks" which
I never could figure out. A precursor
to surface mount? Something else?
I would guess so, but I've never seen a flat pack in the flesh myself so
I can't
say.
11:55 p.m.
On 8 Aug 2008 at 21:58, Roy J. Tellason wrote:
There were also those "flatpacks" which I
never could figure out. A precursor
to surface mount? Something else?
My first exposure to ICs was largely TO-99 style "cans", followed by
flatpacks. The flatpack stuff that I had was mostly RTL. I didn't
see many DIPs until TTL. I mounted the flatpacks as surface-mount
and did a similar thing with the TO packages, just bending the leads
in an L-shape and soldering them down to the PCB pads.
That's the way I did my first IC boards--single-sided PCB, laid out
with tape and etched. Hooked the pads together with No. 26 magnet
wire.
I didn't know any better, but it worked.
Cheers,
Chuck
9 Aug
9 Aug
12:39 a.m.
On Friday 08 August 2008 23:55, Chuck Guzis wrote:
On 8 Aug 2008 at 21:58, Roy J. Tellason wrote:
I remember a couple or three hobby kits of one sort or another that were RTL
in cans. They came with a board but I'm not remembering the orientation of
the board holes at this point, might've been DIP but we're talking 30+ years
ago so I'm not sure. I do remember that I had to bend the leads,
though. :-)
There were also those "flatpacks" which
I never could figure out. A
precursor to surface mount? Something else?
My first exposure to ICs was largely TO-99 style "cans", followed by
flatpacks. The flatpack stuff that I had was mostly RTL. I didn't
see many DIPs until TTL. I mounted the flatpacks as surface-mount
and did a similar thing with the TO packages, just bending the leads
in an L-shape and soldering them down to the PCB pads. That's the way I did my first IC
boards--single-sided PCB, laid out
with tape and etched. Hooked the pads together with No. 26 magnet
wire.
I didn't know any better, but it worked.
I guess that's what counts, eh?
--
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
1:12 a.m.
On 9 Aug 2008 at 0:39, Roy J. Tellason wrote:
I remember a couple or three hobby kits of one sort or
another that were RTL
in cans. They came with a board but I'm not remembering the orientation of
the board holes at this point, might've been DIP but we're talking 30+ years
ago so I'm not sure. I do remember that I had to bend the leads,
though. :-)
I remember the Moto mWRTL experimenter's pack. Cans, but no boards--
and a fair number of analog suggested applications. RTL was kind of
neat that way--it could swing both ways. If you wanted to make a
microphone mixer using a 4-input NAND gate, you could without much
trouble at all. Probably getting close to 40 years, no?
Cheers,
Chuck
That's the way I did my first IC
boards--single-sided PCB, laid out
with tape and etched. Hooked the pads together with No. 26 magnet
wire.
I didn't know any better, but it worked.
I guess that's what counts, eh?
--
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
1:29 a.m.
On Saturday 09 August 2008 01:12, Chuck Guzis wrote:
On 9 Aug 2008 at 0:39, Roy J. Tellason wrote:
Probably, yeah. I used to wander around lower Manhattan and other places
around NYC where I might find interesting places selling assorted odd bits of
electronic junk (an obsession I still have to some extent, collecting that
stuff...).
One time I went into this place that wasn't in the usual areas, not on Canal
St. nor down around Cortlandt/Chambers, that area further down, but I'm
thinking West Broadway maybe. I can't recall the name of the company, but I
can still picture the guy's face.
The kit that I remember well was this "electronic music maker" thingy that had
basically an LFO driving a 4-bit counter and on each counter output you had a
pot and a switch with the common point of those driving an oscillator, so
you could select all sorts of patterns, and sequences. I put a lot of care
into building that for one of my kids back then. Dunno whatever became of
it, either. But it was all definitely RTL in there, it ran off a set of 3
or 4 C cells if I'm remembering right, because those were what would fit in
the box.
In recent email conversation talking about all those old Don Lancaster
cookbooks I mentioned that the RTL one was the only one out of the set that I
didn't have yet, and it was pointed out to me that there were copies *real*
cheap on Amazon, which surprised me. I may yet get a hold of one of those,
though I have little hope of running across any of those chips any more,
even though they used to be common.
And I don't suppose you can have quite as much fun with TTL, either. CMOS
does have some possibilities, maybe. :-)
I remember a couple or three hobby kits of one
sort or another that were
RTL in cans. They came with a board but I'm not remembering the
orientation of the board holes at this point, might've been DIP but
we're talking 30+ years ago so I'm not sure. I do remember that I had to
bend the leads, though. :-)
I remember the Moto mWRTL experimenter's pack. Cans, but no boards--
and a fair number of analog suggested applications. RTL was kind of
neat that way--it could swing both ways. If you wanted to make a
microphone mixer using a 4-input NAND gate, you could without much
trouble at all. Probably getting close to 40 years, no?
Cheers,
Chuck > > That's the way I did my first IC
boards--single-sided PCB, laid out
> > with tape and etched. Hooked the pads together with No. 26 magnet
> > wire.
> >
> > I didn't know any better, but it worked.
>
> I guess that's what counts, eh?
That reminds me of my first wire-wrap board, a CPU (only) build around an
8080 chip. Three regulators on the board, power supply connections by way
of some strips of Vector bus stips cut in thirds, and I never did nail down
any particular niterconnect scheme before I got majorly sidetracked away from
it. But it did work, when I tried it...
--
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
5:32 a.m.
When I worked at Odetics Anaheim,CA in the mid-70s we used tons of the
"flatpacks" in our Spacebourne black boxes. They came in TTL, CMOS and I
think even some ECL. The parts were spot welded with the legs straight out
to gold posts that protruded slightly off of the PCBs. Expensive stuff, a
RAM chip cost about $600 at the time. The parts were real low profile and
weighed less than DIPs ( important in spacecraft, weight / space is at a
premium ). The parts were all MIL-STD and some projects even RAD-hardened
parts. Fun stuff.
Best regards, Steven
My first TTL databook was from TI, and I believe there was some small number
precursor
On Friday 08 August 2008 21:44, Eric Smith wrote:
> Roy J. Tellason wrote:
> > The other thing that I took some notice of was mention of an 18-pin
chip
> > package as "recently developed" at
one point in the story, and later
on
the mention of the 40-pin package as being then
available.
The 18 pin-per-package limit claims regarding the 4004 and 8008 have
been around for a long time, and I think they came from interviews with
Intel's founders and/or early employees, but I think they're factually
incorrect, at least as commonly stated.
The 24-pin DIP was very well established by 1968, and was already used
by TI at that time. of 24-pin devices in there, like the 74154 and
similar. (Hope I'm
remembering right here. :-)
> There were certainly higher pin-count packages at that time also. I'm
not
> sure about the 40-pin DIP, but in 1969 Fairchild
was shipping at least
one
memory chip in
a 36-pin DIP, though that particular package never became
popular.
There were also those "flatpacks" which I never could figure out. A
to surface mount? Something else?
I
Possibly whatever specific company Intel was
contracting with to supply
lead frames and ceramic packages didn't yet offer higher pin count
packages, but they obviously were available from some vendors since
other semiconductor companies like Fairchild and TI were using them.
I guess a large part of it was the machinery to handle that stuff too,
besides the lead frames themselves. I don't know much about that stuff so don't know if any given production machinery would
be adaptable to many
different sizes or if you'd need a different bigger machine to handle
them.
If the latter was the case then I can see where they
might be reluctant if
there weren't a lot of demand.
--
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
12:06 p.m.
New subject: Flat Packs (Forgotten PC History)
Scanning ????????:
When I worked at Odetics Anaheim,CA in the mid-70s we
used tons of the
"flatpacks" in our Spacebourne black boxes. They came in TTL, CMOS and I
think even some ECL. The parts were spot welded with the legs straight out
to gold posts that protruded slightly off of the PCBs. Expensive stuff, a
RAM chip cost about $600 at the time. The parts were real low profile and
weighed less than DIPs ( important in spacecraft, weight / space is at a
premium ). The parts were all MIL-STD and some projects even RAD-hardened
parts. Fun stuff.
I've seen a few things with flat packs and have had a few of the
little critters myself. Most were from Fairchild. Some didn't have the
maker's logo and some weren't marked at all.
THey were typically white ceramic packages with gold lids and gold
leads. They are either square, rectangular or round, with 2, 4, 6,
8--up to 20 leads.
I'm pretty sure that most of the stuff was space grade and came from
satellite and missile projects.
I found them in analog, digital and analog/digital gear, often in UHF
and microwave RF stuff, sometimes mixed with tubes, either soldered to
2oz gold plated copper PCB's, or spot welded onto gold plated posts or
small blocks, or in one case, trimmed and welded in an unusual way to
very heavy copper on a Teflon PCB. Wherever they were PCB mounted,
whether soldered or welded, there usually was a hole cut out for them.
Sometimes the solder was some gold alloy. THe leads were typically not
bent as SMT components are now. I've also seen large arrays of them
attached to leads that came out of PCB layers. These PCB's were
covered with solid gold-plated copper on both sides, with the
circuitry on the layers inside.
They still exist, in some form, at least. The most common form seems
to be as medium powered microwave (above 500MHz) RF transistors,
arrays, amplifiers or LSI/ASIC radios. Some leads are modified to
serve as heatsinks. It seems that they are caught up in the trend
toward leadless packages as well, so they will probably go away too.
They are almost effective as shuriken...
--
jd
5:39 p.m.
New subject: Flat Packs (Forgotten PC History)
On 9 Aug 2008 at 9:06, jd wrote:
I've seen a few things with flat packs and have
had a few of the
little critters myself. Most were from Fairchild. Some didn't have the
maker's logo and some weren't marked at all.
THey were typically white ceramic packages with gold lids and gold
leads. They are either square, rectangular or round, with 2, 4, 6,
8--up to 20 leads.
Here's a nice exhibit:
http://www.vintchip.com/flatpack.html
The ones I used were ceramic, but black/gray not white. No lids,
gold leads. Typical of these was the 723 JK flip-flop--probably
because of defense spending, easier to get and cheaper than the
plastic DIP or TO-99 package, if you could find them.
The original flatpacks, I believe, came from TI. DIPs came later.
Cheers,
Chuck
8:26 p.m.
New subject: Flat Packs (Forgotten PC History)
Mine were like Chuck's and always ceramic but could be black or gray and
always had gold leads. Mostly TTL and some 4000 Series CMOS. They conformed
to MIL-STD-1835C and cost a lot. If you want to see the progression in
packaging of ICs from 1960 forward check out;
http://homepages.nildram.co.uk/~wylie/ICs/monolith.htm
Best regards, Steven
On 9 Aug 2008 at 9:06, jd wrote:
I've seen a few things with flat packs and
have had a few of the
little critters myself. Most were from Fairchild. Some didn't have the
maker's logo and some weren't marked at all.
THey were typically white ceramic packages with gold lids and gold
leads. They are either square, rectangular or round, with 2, 4, 6,
8--up to 20 leads.
Here's a nice exhibit:
http://www.vintchip.com/flatpack.html
The ones I used were ceramic, but black/gray not white. No lids,
gold leads. Typical of these was the 723 JK flip-flop--probably
because of defense spending, easier to get and cheaper than the
plastic DIP or TO-99 package, if you could find them.
The original flatpacks, I believe, came from TI. DIPs came later.
Cheers,
Chuck 
9:07 p.m.
New subject: Flat Packs
I've only come across two electronic calculators that use flatpack IC's.
In 1965, Victor Comptometer made an attempt to build a fully-electronic
desktop electronic calculator, called the Victor 3900, that proved to
push the state of the art just a little too far. The 3900 used
round-case flat-pack IC's, with 22 leads, with spot-weld type technology
used to connect the leads to the circuit traces on the circuit board.
Unfortunately, the yields of the ICs were not very high, and there were
lots of infant mortality problems with the chips dying once the machines
were in the customer's hands. The machine, had it been successful,
would have put Victor in the enviable position of making a massive jump
in technology over the electronics used by anyone else at the time, and
would have kept them ahead of the game for quite a long time. There's
some information on the Old Calculator Museum website exhibit on the
Victor 14-322 (a machine made significantly later using very
conventional technology) at
http://oldcalculatormuseum.com/vic14-322.html.
In around 1970, the Casio AL-2000
(http://oldcalculatormuseum.com/casal2k.html), an early small-scale MOS
IC-based programmable (in very limited form) calculator, used two 24-pin
flat-pack integrated circuits made by Philco, in ceramic bases with gold
lids and leads in the arithmetic logic unit. These are medium-scale
devices, one of which (part number SC-1770) is a bit-serial adder adder,
and the other chip (SC1771) is a complimenter IC, date coded in late
1968 to early 1969. The chips are surface mounted, again using some
kind of spot-welding to connect the leads of the device to the traces on
the circuit board.
Rick Bensene
The Old Calculator Museum
http://oldcalculatormuseum.com
11 Aug
11 Aug
3 a.m.
New subject: Flat Packs
TI's TTL Data Book (1976) offered flat packages. I don't remember
ever using such.
bs
On Sat, 9 Aug 2008, Rick Bensene wrote:
I've only come across two electronic calculators
that use flatpack IC's.
In 1965, Victor Comptometer made an attempt to build a fully-electronic
desktop electronic calculator, called the Victor 3900, that proved to
push the state of the art just a little too far. The 3900 used
round-case flat-pack IC's, with 22 leads, with spot-weld type technology
used to connect the leads to the circuit traces on the circuit board.
Unfortunately, the yields of the ICs were not very high, and there were
lots of infant mortality problems with the chips dying once the machines
were in the customer's hands. The machine, had it been successful,
would have put Victor in the enviable position of making a massive jump
in technology over the electronics used by anyone else at the time, and
would have kept them ahead of the game for quite a long time. There's
some information on the Old Calculator Museum website exhibit on the
Victor 14-322 (a machine made significantly later using very
conventional technology) at
http://oldcalculatormuseum.com/vic14-322.html.
In around 1970, the Casio AL-2000
(http://oldcalculatormuseum.com/casal2k.html), an early small-scale MOS
IC-based programmable (in very limited form) calculator, used two 24-pin
flat-pack integrated circuits made by Philco, in ceramic bases with gold
lids and leads in the arithmetic logic unit. These are medium-scale
devices, one of which (part number SC-1770) is a bit-serial adder adder,
and the other chip (SC1771) is a complimenter IC, date coded in late
1968 to early 1969. The chips are surface mounted, again using some
kind of spot-welding to connect the leads of the device to the traces on
the circuit board.
Rick Bensene
The Old Calculator Museum
http://oldcalculatormuseum.com
9 Aug
9 Aug
9:59 p.m.
Roy J. Tellason wrote:
There were also those "flatpacks" which I
never could figure out. A precursor
to surface mount? Something else?
Yes, for military applications. For instance, the Apollo Guidance
Computer is full of chips in flatpacks.
I'm not an expert on early IC packaging, but AFAIK the flatpack predated
the DIP. Prior to the DIP, most ICs had 10 pins or fewer, and were
offered in flatpacks or round metal cans.
10 Aug
10 Aug
5:05 p.m.
New subject: IC packages / was Re: Forgotten PC History
Eric Smith wrote:
Roy J. Tellason wrote:
Not definitive as to when the DIP was first introduced, but a couple of datapoints:
- 1965 TI catalog shows only flatpacks
- 1966 Fairchild catalog shows flatpacks, round cans (metal and epoxy),
and 14 and 16-pin DIP packages.
The DIP packages are also referred to as being "plug-in packages".
Quote from the 1966 Fairchild catalog:
The Dual In-Line - the most significant contribution to
micro-circuit packaging in recent years - is a Fairchild concept.
Its mechanical design incorporates the ultimate in design balance:
thermal resistance is optimum, packaging density can be maximized
without economic burden, and the cost of interconnection and attachment
typically runs 10% of the cost of similar assembly with flat packs.
The Dual In-Line package is fully hermetic, and although designed
primarily to meet the economic objective of most industrial
applications, it meets all the environmental requirements imposed
by military systems.
There were also those "flatpacks" which
I never could figure out. A precursor
to surface mount? Something else?
Yes, for military applications. For instance, the Apollo Guidance
Computer is full of chips in flatpacks.
I'm not an expert on early IC packaging, but AFAIK the flatpack predated
the DIP. Prior to the DIP, most ICs had 10 pins or fewer, and were
offered in flatpacks or round metal cans.
8:11 p.m.
New subject: IC packages / was Re: Forgotten PC History
On 10 Aug 2008 at 13:05, Brent Hilpert wrote:
Not definitive as to when the DIP was first
introduced, but a couple of datapoints:
- 1965 TI catalog shows only flatpacks
- 1966 Fairchild catalog shows flatpacks, round cans (metal and epoxy),
and 14 and 16-pin DIP packages.
The DIP packages are also referred to as being "plug-in packages".
From:
http://www.hbci.com/~wenonah/history/edpart6.htm
"For both linear and digital ICs, packaging problems had to be
overcome. Transistor packages were found to lack sufficient
heat-sinking capability and an adequate number of interconnections.
One solution was the flatpack, created by Yung Tao while at Texas
Instruments. The original flatpack had 10 leads, and measured 1/4 x
1/8 inch. In 1964 Fairchild's Bryant ("Buck") Rogers fostered. the
invention of the dual-inline package. The original DIP had 14 leads,
and looked just as it does today. The same year, Martin LepSelter of
Bell Labs invented the beam lead as a mechanical and electrical
interconnection between the IC and its case."
Cheers,
Chuck
8 Aug
8 Aug
11:44 p.m.
From: eric at brouhaha.com
---snip--->
Possibly whatever specific company Intel was
contracting with to supply
lead frames and ceramic packages didn't yet offer higher pin count
packages, but they obviously were available from some vendors since
other semiconductor companies like Fairchild and TI were using them.
Eric
Hi
My understanding was that Intel wasn't using them and they
didn't expect to sell enough uP chips to make it practical to use
a larger size. Later designs did use more pins.
Dwight
_________________________________________________________________
Reveal your inner athlete and share it with friends on Windows Live.
http://revealyourinnerathlete.windowslive.com?locale=en-us&ocid=TXT_TAG…
11 Aug
11 Aug
2:37 a.m.
On Fri, 8 Aug 2008, dwight elvey wrote:
The 8008 was only introduced a year after the 4004 which was
introduced in 1971. I can't find a pin out for either of them
but I suspect that they might be pin compatable so that the
upper four bits would be simply not used if the 8008 was put
in a system designed for a 4004. They were used mostly in
industrial control systems according to my college text of
only slightly later.
Unlike the 8080 and later, they used PMOS logic and multiplexed
addressing. I cannot find a pin out or instruction set.
Oh, the 8008 had 48 instructions as opposed to 78 for the 8080
family.
bs
From: eric at brouhaha.com
---snip--->
Possibly whatever specific company Intel was
contracting with to supply
lead frames and ceramic packages didn't yet offer higher pin count
packages, but they obviously were available from some vendors since
other semiconductor companies like Fairchild and TI were using them.
Eric
Hi
My understanding was that Intel wasn't using them and they
didn't expect to sell enough uP chips to make it practical to use
a larger size. Later designs did use more pins.
Dwight
____________________________________________________________ 
3:29 a.m.
On 11/08/2008 07:37, schwepes at moog.netaxs.com wrote:
The 8008 was only introduced a year after the 4004
which was
introduced in 1971. I can't find a pin out for either of them
but I suspect that they might be pin compatable so that the
upper four bits would be simply not used if the 8008 was put
in a system designed for a 4004.
Not even close, I'm afraid. The 4004 is a 16-pin device with Vdd (-10V)
and Vss (+5V) on the centre pins of the two sides, the 4 data bits on
pins 1-4, and the two clock phases and SYNC on pins 6-8. The 8008 is an
18-pin device with Vcc (+5V) and Vss (-9V) on diagonally opposite
corners (pins 1 and 10), D0-D7 on pins 2-9, and the clock phases and
SYNC on the other side (pins 14-16):
4004 8008
__ __ __ __
D0 | U | CMRAM0 (-9V) Vss | U | INT
D1 | | CMRAM1 D0 | | READY
D2 | | CMRAM2 D1 | | phi1
D3 | | CMRAM3 D2 | | phi2
(+5V) Vss | | Vdd (-10V) D3 | | SYNC
phi1 | | CMROM D4 | | S0
phi2 | | TEST D5 | | S1
SYNC |_____| RESET D6 | | S2
D7 |_____| Vcc (+5V)
--
Pete Peter Turnbull
Network Manager
University of York
4:56 a.m.
New subject: 4004/8008 (was Re: Forgotten PC History)
On Mon, Aug 11, 2008 at 08:29:28AM +0100, Pete Turnbull wrote:
On 11/08/2008 07:37, schwepes at moog.netaxs.com
wrote:
Hmm... I happen to have a couple of 4004 chips (one pulled from a
non-UPC barcode scanner, the other from a commercial kitchen scale),
and I've long fantasized about building some showcase 4004 circuit,
probably a digital clock.
I don't happen to have the 4004 datasheet handy (and can't get to it
for a number of hours), but the -10V Vdd requirement makes me wonder
if it's possible/feasible to use a pair of ganged ICL7660s (still made
by Dallas/Maxim) to generate it. This chip came up recently as a
possibility for generating -5V Vdd for 4116 DRAMs (and I think the
concensus was that it would probably be adequate for up to 16 chips),
so I do recall that it can only source a few mA (each 4116 needs a
few dozen uA)...
I don't see a GND pin on the 4004, so I'm at a loss to estimate
what it would require for Vdd. Anyone know or can check easily?
Also, does anyone know about any sort of 4004 emulator or simulator?
I wouldn't mind trying my hand at some code *before* trying to build
out the hardware.
-ethan
--
Ethan Dicks, A-333-S Current South Pole Weather at 11-Aug-2008 at 08:50 Z
South Pole Station
PSC 468 Box 400 Temp -68.4 F (-55.8 C) Windchill -111.4 F (-79.7 C)
APO AP 96598 Wind 17.0 kts Grid 23 Barometer 671.5 mb (10950 ft)
Ethan.Dicks at usap.gov http://penguincentral.com/penguincentral.html
The 8008 was only introduced a year after the 4004
which was
introduced in 1971. I can't find a pin out for either of them
but I suspect that they might be pin compatable...
Not even close, I'm afraid. The 4004 is a 16-pin device with Vdd (-10V)
and Vss (+5V) on the centre pins of the two sides, the 4 data bits on
pins 1-4, and the two clock phases and SYNC on pins 6-8. The 8008 is an
18-pin device with Vcc (+5V) and Vss (-9V) on diagonally opposite
corners (pins 1 and 10), D0-D7 on pins 2-9, and the clock phases and
SYNC on the other side (pins 14-16):
4004 8008
__ __ __ __
D0 | U | CMRAM0 (-9V) Vss | U | INT
D1 | | CMRAM1 D0 | | READY
D2 | | CMRAM2 D1 | | phi1
D3 | | CMRAM3 D2 | | phi2
(+5V) Vss | | Vdd (-10V) D3 | | SYNC
phi1 | | CMROM D4 | | S0
phi2 | | TEST D5 | | S1
SYNC |_____| RESET D6 | | S2
D7 |_____| Vcc (+5V)
8:36 a.m.
New subject: 4004/8008 (was Re: Forgotten PC History)
On 11/08/2008 09:56, Ethan Dicks wrote:
On Mon, Aug 11, 2008 at 08:29:28AM +0100, Pete
Turnbull wrote:
> 4004 8008
> __ __ __ __
> D0 | U | CMRAM0 (-9V) Vss | U | INT
> D1 | | CMRAM1 D0 | | READY
> D2 | | CMRAM2 D1 | | phi1
> D3 | | CMRAM3 D2 | | phi2
> (+5V) Vss | | Vdd (-10V) D3 | | SYNC
> phi1 | | CMROM D4 | | S0
> phi2 | | TEST D5 | | S1
> SYNC |_____| RESET D6 | | S2
> D7 |_____| Vcc (+5V)
>
I don't happen to have the 4004 datasheet handy
(and can't get to it
for a number of hours), but the -10V Vdd requirement makes me wonder
if it's possible/feasible to use a pair of ganged ICL7660s (still made
by Dallas/Maxim) to generate it.
It takes about 30mA to run a 4004.
I don't see a GND pin on the 4004, so I'm at a
loss to estimate
what it would require for Vdd. Anyone know or can check easily?
Some data sheets list Vdd = -15V and Vss = 0V, so the -10/+5 is just a
convenience. Logic levels are relative to Vss.
Also, does anyone know about any sort of 4004 emulator
or simulator?
I wouldn't mind trying my hand at some code *before* trying to build
out the hardware.
The whole manual is available at
http://www.intel.com/museum/archives/4004ip.htm
I came across this while looking for something else:
http://tams-www.informatik.uni-hamburg.de/applets/hades/webdemos/80-mcs4/in…
Never tried it, but it might be a place to start.
--
Pete Peter Turnbull
Network Manager
University of York
3:43 p.m.
New subject: 4004/8008 (was Re: Forgotten PC History)
On Mon, Aug 11, 2008 at 01:36:45PM +0100, Pete Turnbull wrote:
Ah. OK. I've worked with DEC R-series logic (+10V/-15V power, +3V/GND
signals), so I didn't assume anything about the logic levels.
Nice. Thanks. I'll grab that when the sats rise.
On 11/08/2008 09:56, Ethan Dicks wrote:
OK. That might be a bit much for a pair of ICL7660s. If not, it's
right on the edge.
I don't happen to have the 4004 datasheet
handy (and can't get to it
for a number of hours), but the -10V Vdd requirement makes me wonder
if it's possible/feasible to use a pair of ganged ICL7660s (still made
by Dallas/Maxim) to generate it.
It takes about 30mA to run a 4004. I don't see
a GND pin on the 4004, so I'm at a loss to estimate
what it would require for Vdd. Anyone know or can check easily?
Some data sheets list Vdd = -15V and Vss = 0V, so the -10/+5 is just a
convenience. Logic levels are relative to Vss. Also, does
anyone know about any sort of 4004 emulator or simulator?
I wouldn't mind trying my hand at some code *before* trying to build
out the hardware.
The whole manual is available at
http://www.intel.com/museum/archives/4004ip.htm I came across this while looking for something else:
http://tams-www.informatik.uni-hamburg.de/applets/hades/webdemos/80-mcs4/in…
Never tried it, but it might be a place to start.
Yes... I'll check that out, too.
Thanks,
-ethan
--
Ethan Dicks, A-333-S Current South Pole Weather at 11-Aug-2008 at 19:40 Z
South Pole Station
PSC 468 Box 400 Temp -78.7 F (-61.5 C) Windchill -123.7 F (-86.5 C)
APO AP 96598 Wind 15.9 kts Grid 13 Barometer 671.6 mb (10947 ft)
Ethan.Dicks at usap.gov http://penguincentral.com/penguincentral.html
10:08 a.m.
New subject: 4004/8008 (was Re: Forgotten PC History)
From: ethan.dicks at usap.gov
Hmm... I happen to have a couple of 4004
chips (one pulled from a
non-UPC barcode scanner, the other from a commercial kitchen scale),
and I've long fantasized about building some showcase 4004 circuit,
probably a digital clock.
I don't happen to have the 4004 datasheet handy (and can't get to it
for a number of hours), but the -10V Vdd requirement makes me wonder
if it's possible/feasible to use a pair of ganged ICL7660s (still made
by Dallas/Maxim) to generate it. This chip came up recently as a
possibility for generating -5V Vdd for 4116 DRAMs (and I think the
concensus was that it would probably be adequate for up to 16 chips),
so I do recall that it can only source a few mA (each 4116 needs a
few dozen uA)...
I don't see a GND pin on the 4004, so I'm at a loss to estimate
what it would require for Vdd. Anyone know or can check easily?
Also, does anyone know about any sort of 4004 emulator or simulator?
I wouldn't mind trying my hand at some code *before* trying to build
out the hardware.
-ethan
Hi Ethan
The chip is P-mos and it doesn't directly connect to ttl.
The down is very week and only pulls up with any
strength.
There were a couple chip interfaces ( 4008 and 4009 )
that created TTL level signals to run standard RAM/ROM.
The only issue here is that one lost the instruction stream
so that one couldn' t take advantage of any RAM ot ROM
specific commands.
The later came out with a single chip version of the
interface( I think it was 4269 ).
These come up on ebay every now and then.
The manual on Al's site shows a circuit to convert the Pmos
levels to a TTL eprom ( used on the SIM4-1 ). The circuit does
use 4002-1 and 4002-2 RAMs.
The 4001/4002 had built in I/O. They would watch the instuction
steam and execute an I/O operation when the 4004 executed
the fetched instruction.
It is easier to modify a -12V supply to power the -10v than
to make a converter. Many floppy disk supplies are fine for
this.
Dwight
_________________________________________________________________
Your PC, mobile phone, and online services work together like never before.
http://clk.atdmt.com/MRT/go/108587394/direct/01/
3:08 p.m.
New subject: 4004/8008 (was Re: Forgotten PC History)
dwight elvey wrote:
There were a couple chip interfaces ( 4008 and 4009 )
[...]
The later came out with a single chip version of the
interface( I think it was 4269 ).
4289
The 4269 was a keyboard/display interface, conceptually similar to the 8279.
12 Aug
12 Aug
1:31 a.m.
New subject: 4004/8008 (was Re: Forgotten PC History)
From: eric at brouhaha.com
dwight elvey wrote:
Hi Eric
Thanks, you are correct. I should have looked it up. I have seen
these on ebay at times.
Dwight
_________________________________________________________________
Your PC, mobile phone, and online services work together like never before.
http://clk.atdmt.com/MRT/go/108587394/direct/01/
There were a couple chip interfaces ( 4008 and
4009 )
[...]
The later came out with a single chip version of
the
interface( I think it was 4269 ).
4289
The 4269 was a keyboard/display interface, conceptually similar to the 8279.
11 Aug
11 Aug
3:51 p.m.
New subject: 4004/8008 (was Re: Forgotten PC History)
On Mon, Aug 11, 2008 at 07:08:43AM -0700, dwight elvey wrote:
I do remember that it is P-MOS, but that's a bit out of my
experience - I got my start in the CMOS and TTL days of the
late 1970s. What little I know about older circuit types (RTL,
DTL, etc.) is from playing with DEC equipment.
I am not surprised to hear that it won't connect directly
to TTL, but I was thinking a 4007 or 4050 might do the trick.
From: ethan.dicks at usap.gov
Hmm... I happen to have a couple of 4004
chips (one pulled from a
non-UPC barcode scanner, the other from a commercial kitchen scale),
and I've long fantasized about building some showcase 4004 circuit,
probably a digital clock.
Hi Ethan
The chip is P-mos and it doesn't directly connect to ttl. The down is very week and only pulls up with any
strength.
OK.
There were a couple chip interfaces ( 4008 and 4009
)
that created TTL level signals to run standard RAM/ROM.
The only issue here is that one lost the instruction stream
so that one couldn' t take advantage of any RAM ot ROM
specific commands.
The later came out with a single chip version of the
interface( I think it was 4269 ).
These come up on ebay every now and then.
OK. I don't know if I have _those_ exact chips, but I do
have some support chips along with the processors.
The manual on Al's site shows a circuit to convert
the Pmos
levels to a TTL eprom ( used on the SIM4-1 ). The circuit does
use 4002-1 and 4002-2 RAMs.
Those I have, at least one or two of.
The EPROM interface sounds quite useful - that was really my
worry since it's not possible to order genine MCS4-family
ROMs anymore.
The 4001/4002 had built in I/O. They would watch the
instuction
steam and execute an I/O operation when the 4004 executed
the fetched instruction.
Right... I remember something about built-in I/O ports, but not
the details.
It is easier to modify a -12V supply to power the
-10v than
to make a converter. Many floppy disk supplies are fine for
this.
Sure. I was just hoping to be able to dedicate a square inch or
two of the design and have a single-supply board. I also have
some +5V/-12V potted chargepump blocks at home, a remnant of
a particular COMBOARD design for a DEC chassis that lacked -12V
or -15V to run the EIA level shifters. Those might be able to
put out more than a few mA at -12V.
Thanks for the more detailed info,
-ethan
--
Ethan Dicks, A-333-S Current South Pole Weather at 11-Aug-2008 at 19:40 Z
South Pole Station
PSC 468 Box 400 Temp -78.7 F (-61.5 C) Windchill -123.7 F (-86.5 C)
APO AP 96598 Wind 15.9 kts Grid 13 Barometer 671.6 mb (10947 ft)
Ethan.Dicks at usap.gov http://penguincentral.com/penguincentral.html
12 Aug
12 Aug
1:48 a.m.
New subject: 4004/8008 (was Re: Forgotten PC History)
From: ethan.dicks at usap.gov
---snip--->
> Thanks for the more detailed info,> Hi Ethan
I should mention I do have a working SIM4-1 and programmer card,
as described in the manual.
I also have a hand written assembler and disassembler.
The disassembler creates nice Intel code but I wrote the
code for the assembler at a different time so they are not
compatible( no good excuse ).
Both are written in Forth and run under F-PC ( a easy to get
freeware version ). You don't need to know much Forth to
use either.
I also have an emulator, also written in Forth. It executes code
but I never finished connecting I/O functions. It is still useful
but one needs to manually load input values and then execute
until the next input. I'd intended to have it read input from a file
and use the cycle count to determine each I/O data change.
I just got side tracked and never finished it. It does still
read out instructions as it runs and allows breakpoints and
single stepping.
If you think you'd be interested, I can dig it up.
Dwight
_________________________________________________________________
Your PC, mobile phone, and online services work together like never before.
http://clk.atdmt.com/MRT/go/108587394/direct/01/
11 Aug
11 Aug
10:23 a.m.
New subject: 4004/8008 (was Re: Forgotten PC History)
>>>> "Ethan" == Ethan Dicks
<ethan.dicks at usap.gov> writes:
Ethan> Hmm... I happen to have a couple of 4004 chips (one pulled
Ethan> from a non-UPC barcode scanner, the other from a commercial
Ethan> kitchen scale), and I've long fantasized about building some
Ethan> showcase 4004 circuit, probably a digital clock.
Or you could do like a classmate of mine did in 1973-1974 and build a
PC. Not easy, though; I remember a large wire-wrap board with over
100 chips on it.
paul
3:52 p.m.
New subject: 4004/8008 (was Re: Forgotten PC History)
On Mon, Aug 11, 2008 at 10:23:55AM -0400, Paul Koning wrote:
I think that's much too ambitious for me. If I can't make a simple digital
clock, I doubt I could make a full-on computer.
-ethan
--
Ethan Dicks, A-333-S Current South Pole Weather at 11-Aug-2008 at 19:50 Z
South Pole Station
PSC 468 Box 400 Temp -78.9 F (-61.6 C) Windchill -124.1 F (-86.7 C)
APO AP 96598 Wind 16.1 kts Grid 11 Barometer 671.6 mb (10947 ft)
Ethan.Dicks at usap.gov http://penguincentral.com/penguincentral.html
>>>> "Ethan" == Ethan Dicks <ethan.dicks at usap.gov> writes:
Ethan> Hmm... I happen to have a couple of 4004 chips (one pulled
Ethan> from a non-UPC barcode scanner, the other from a commercial
Ethan> kitchen scale), and I've long fantasized about building some
Ethan> showcase 4004 circuit, probably a digital clock.
Or you could do like a classmate of mine did in 1973-1974 and build a
PC. Not easy, though; I remember a large wire-wrap board with over
100 chips on it. 
12:23 p.m.
New subject: 4004/8008 (was Re: Forgotten PC History)
In article <20080811085639.GB11609 at usap.gov>,
Ethan Dicks <ethan.dicks at usap.gov> writes:
Hmm... I happen to have a couple of 4004 chips (one
pulled from a
non-UPC barcode scanner, the other from a commercial kitchen scale),
and I've long fantasized about building some showcase 4004 circuit,
probably a digital clock.
Wouldn't a terminal or calculator be the canonical circuit for which
the 4004 was designed?
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://www.xmission.com/~legalize/book/download/index.html>
Legalize Adulthood! <http://blogs.xmission.com/legalize/>
3:54 p.m.
New subject: 4004/8008 (was Re: Forgotten PC History)
On Mon, Aug 11, 2008 at 10:23:58AM -0600, Richard wrote:
In article <20080811085639.GB11609 at usap.gov>,
Ethan Dicks <ethan.dicks at usap.gov> writes:
Probably, but I was thinking small to begin with. A 4004-based terminal
would be neat, but far from simple, especially if it was meant to be
anything more than a glass TTY.
-ethan
--
Ethan Dicks, A-333-S Current South Pole Weather at 11-Aug-2008 at 19:50 Z
South Pole Station
PSC 468 Box 400 Temp -78.9 F (-61.6 C) Windchill -124.1 F (-86.7 C)
APO AP 96598 Wind 16.1 kts Grid 11 Barometer 671.6 mb (10947 ft)
Ethan.Dicks at usap.gov http://penguincentral.com/penguincentral.html
Hmm... I happen to have a couple of 4004 chips
(one pulled from a
non-UPC barcode scanner, the other from a commercial kitchen scale),
and I've long fantasized about building some showcase 4004 circuit,
probably a digital clock.
Wouldn't a terminal or calculator be the canonical circuit for which
the 4004 was designed?
6:09 a.m.
And then the 1974 8080 seems to be even more different.
It doesn't even look like the same design philosophies on
any of the chips.
bs
On Mon, 11 Aug 2008, Pete Turnbull wrote:
On 11/08/2008 07:37, schwepes at moog.netaxs.com
wrote:
The 8008 was only introduced a year after the
4004 which was
introduced in 1971. I can't find a pin out for either of them
but I suspect that they might be pin compatable so that the
upper four bits would be simply not used if the 8008 was put
in a system designed for a 4004.
Not even close, I'm afraid. The 4004 is a 16-pin device with Vdd (-10V)
and Vss (+5V) on the centre pins of the two sides, the 4 data bits on
pins 1-4, and the two clock phases and SYNC on pins 6-8. The 8008 is an
18-pin device with Vcc (+5V) and Vss (-9V) on diagonally opposite
corners (pins 1 and 10), D0-D7 on pins 2-9, and the clock phases and
SYNC on the other side (pins 14-16):
4004 8008
__ __ __ __
D0 | U | CMRAM0 (-9V) Vss | U | INT
D1 | | CMRAM1 D0 | | READY
D2 | | CMRAM2 D1 | | phi1
D3 | | CMRAM3 D2 | | phi2
(+5V) Vss | | Vdd (-10V) D3 | | SYNC
phi1 | | CMROM D4 | | S0
phi2 | | TEST D5 | | S1
SYNC |_____| RESET D6 | | S2
D7 |_____| Vcc (+5V)
--
Pete Peter Turnbull
Network Manager
University of York
7:46 a.m.
Pete Turnbull wrote:
On 11/08/2008 07:37, schwepes at moog.netaxs.com
wrote:
<nitpick>
This doesn't look correct. My 8008 8 bit parallel CPU manual gives the
following order on the data bits. (Also at bitsavers at
http://bitsavers.org/pdf/intel/MCS8/8008usersManualRev4_Nov73.pdf )
8008
__ __
(-9V) Vss | U | INT
D7 | | READY
D6 | | phi1
D5 | | phi2
D4 | | SYNC
D3 | | S0
D2 | | S1
D1 | | S2
D0 |_____| Vcc (+5V)
</nitpick>
And to answer another question about the screen on the Datapoint 2200,
it was 12 rows of 80 characters.
Btw, anyone have an Intellec 8 computer they don't need? It's one of my
"wish list" machines. :-)
/G?ran
The 8008 was only introduced a year after the
4004 which was
introduced in 1971. I can't find a pin out for either of them
but I suspect that they might be pin compatable so that the
upper four bits would be simply not used if the 8008 was put
in a system designed for a 4004.
Not even close, I'm afraid. The 4004 is a 16-pin device with Vdd
(-10V) and Vss (+5V) on the centre pins of the two sides, the 4 data
bits on pins 1-4, and the two clock phases and SYNC on pins 6-8. The
8008 is an 18-pin device with Vcc (+5V) and Vss (-9V) on diagonally
opposite corners (pins 1 and 10), D0-D7 on pins 2-9, and the clock
phases and SYNC on the other side (pins 14-16):
4004 8008
__ __ __ __
D0 | U | CMRAM0 (-9V) Vss | U | INT
D1 | | CMRAM1 D0 | | READY
D2 | | CMRAM2 D1 | | phi1
D3 | | CMRAM3 D2 | | phi2
(+5V) Vss | | Vdd (-10V) D3 | | SYNC
phi1 | | CMROM D4 | | S0
phi2 | | TEST D5 | | S1
SYNC |_____| RESET D6 | | S2
D7 |_____| Vcc (+5V)
12 Aug
12 Aug
3:17 a.m.
On 11/08/2008 12:46, G?ran Axelsson wrote:
Pete Turnbull wrote:
{ pinouts snipped ]
<nitpick>
This doesn't look correct. My 8008 8 bit parallel CPU manual gives the
following order on the data bits. (Also at bitsavers at
http://bitsavers.org/pdf/intel/MCS8/8008usersManualRev4_Nov73.pdf )
8008
__ __
(-9V) Vss | U | INT
D7 | | READY
D6 | | phi1
D5 | | phi2
D4 | | SYNC
D3 | | S0
D2 | | S1
D1 | | S2
D0 |_____| Vcc (+5V)
</nitpick>
Oops! You're right, and that's what I get for being too hasty with my
typing. Well spotted.
--
Pete Peter Turnbull
Network Manager
University of York
11 Aug
11 Aug
2:37 p.m.
schwepes at moog.netaxs.com wrote:
The 8008 was only introduced a year after the 4004
which was
introduced in 1971. I can't find a pin out for either of them
but I suspect that they might be pin compatable so that the
upper four bits would be simply not used if the 8008 was put
in a system designed for a 4004.
Not even close. The bus works significantly differently.
9 Aug
9 Aug
3:49 p.m.
The 18 pin-per-package limit claims regarding the 4004
and 8008 have
been around for a long time, and I think they came from interviews with
Intel's founders and/or early employees, but I think they're factually
incorrect, at least as commonly stated.
The 24-pin DIP was very well established by 1968, and was already used
by TI at that time. There were certainly higher pin-count packages at
that time also. I'm not sure about the 40-pin DIP, but in 1969
Fairchild was shipping at least one memory chip in a 36-pin DIP, though
that particular package never became popular.
I would find it hard to believe that 24-pin and higher dips were
unavailable.
I have a couple of Sharp calculators from about that date. I just
grabbed the nearest (an ELSI-160) and opened it up, to find that the
four chips have date codes in early 1971 and forty-two (yes, 42) pins
each in a dual zigzag arrangement. The pins are spaced at 0.05 inch,
and alternate pins stick out 0.1 inch further than the rest. The whole
is a gold-top ceramic package a little smaller than a 24-pin DIP.
(FWIW at least one of the calculators has edge connectors with a pin
spacing of 1.25mm. But the chips are still 0.05", i.e. 1.27mm. And
yes, you can tell the difference. For the early '70s, this is high
precision stuff!)
Possibly whatever specific company Intel was
contracting with to supply
lead frames and ceramic packages didn't yet offer higher pin count
packages, but they obviously were available from some vendors since
other semiconductor companies like Fairchild and TI were using them.
That makes sense. I think someone said it already: if the chip was a
custom job for a lowish-volume contract, Intel wouldn't want to go and
find a new packaging contractor for it.
Philip.
10 Aug
10 Aug
4:53 a.m.
Philip Belben wrote:
I have a couple of Sharp calculators from about that
date. I just
grabbed the nearest (an ELSI-160) and opened it up, to find that the
four chips have date codes in early 1971 and forty-two (yes, 42) pins
each in a dual zigzag arrangement.
42-pin DIP packages were common among Japanese semiconductor vendors but
I don't recall seeing any from US vendors.
For example, the early NEC cassette/cartridge tape controller chip,
uPD371 was in a 42-pin DIP. I think their single-density floppy
controller chip, uPD372, might also have been in a 42-pin DIP.
Eric
11 Aug
11 Aug
11:13 p.m.
On Sunday 10 August 2008 04:53, Eric Smith wrote:
Philip Belben wrote:
I think I may even have a NEC databook that features some of those, and
remember how odd it was that those seemed to me at the time. Probably
somewhere around 1978, maybe?
--
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
I have a couple of Sharp calculators from about
that date. I just
grabbed the nearest (an ELSI-160) and opened it up, to find that the
four chips have date codes in early 1971 and forty-two (yes, 42) pins
each in a dual zigzag arrangement.
42-pin DIP packages were common among Japanese semiconductor vendors but
I don't recall seeing any from US vendors.
For example, the early NEC cassette/cartridge tape controller chip,
uPD371 was in a 42-pin DIP. I think their single-density floppy
controller chip, uPD372, might also have been in a 42-pin DIP. 
10 Aug
10 Aug
12:24 p.m.
On Aug 9, 2008, at 3:49 PM, Philip Belben wrote:
That makes sense. I think someone said it already: if
the chip was
a custom job for a lowish-volume contract, Intel wouldn't want to
go and find a new packaging contractor for it.
Did they actually outsource their packaging?
-Dave
--
Dave McGuire
Port Charlotte, FL
1:10 p.m.
Dave McGuire wrote:
On Aug 9, 2008, at 3:49 PM, Philip Belben wrote:
That's actually a fairly common occurrence.
Peace... Sridhar
That makes sense. I think someone said it
already: if the chip was a
custom job for a lowish-volume contract, Intel wouldn't want to go and
find a new packaging contractor for it.
Did they actually outsource their packaging?
12:38 p.m.
Moto had flatpacks with 32 pins (8 per side on 4 sides) for their
semicustom gate arrays in the late 60's.
Cheers,
Chuck
2:13 p.m.
bfranchuk at jetnet.ab.ca wrote:
I am not shure of the packaging, but Motorola's
ECL alu
came in a weird package too.
MC10800 series. Quad-In-Line Package (QUIP): pins on two opposite sides
of the body like a DIP, but with the lengths protruding from the package
(before the downward bend) alternating between short and long, so that
the pins are in two rows on each side.
Rockwell was one of the biggest users of QUIP packages for
microcontrollers and such. I'm not sure whether their QUIPs were
dimensioned the same as the Motorola ones.
Note that these are different than the QUIP package used by Intel for
the iAPX-432 components, and by both Intel and Zilog for emulation
bond-out chips, prior to the introduction of the JEDEC 68 position
ceramic leadless chip carrier. This QUIP was designed by Intel and AMP,
and was a leadless ceramic package with contacts along the two long
sides, the contacts being staggered in a similar manner to the leaded QUIPs.
2:44 p.m.
On 10 Aug 2008 at 11:13, Eric Smith wrote:
Rockwell was one of the biggest users of QUIP packages
for
microcontrollers and such. I'm not sure whether their QUIPs were
dimensioned the same as the Motorola ones.
The PPS-4 microprocessor 1972 in QUIP being particularly noteworthy.
I found a photo of one here:
http://www.antiquetech.com/pictures%20full-size/Rockwell12660EA.jpg
Cheers,
Chuck
5:05 p.m.
Philip Belben wrote:
This issue is another of those recurring topics on the list (see URL):
Subject: Re: 4004 and IC history / was Re: Vintage computer photogallery
Date: Sat, 13 Oct 2007 14:14:25 -0700 Date: Sat, 13 Oct 2007 14:14:25 -0700
Brent Hilpert wrote:
Possibly
whatever specific company Intel was contracting with to supply
lead frames and ceramic packages didn't yet offer higher pin count
packages, but they obviously were available from some vendors since
other semiconductor companies like Fairchild and TI were using them.
That makes sense. I think someone said it already: if the chip was a
custom job for a lowish-volume contract, Intel wouldn't want to go and
find a new packaging contractor for it. dwight elvey wrote:
> This was what Federico Faggin had stated at a talk he gave at the CHM.
> Packaging was expensive and they'd made volume deals on 16 pin
> packages. Faggin stated that his design could have been a few times
> faster had he not been restricted to the small bus.
Here it is, straight from Faggin:
http://www.intel4004.com/qa4004.htm, under the
question "Q.: Was there anything unique about the 4004?" or search for
"16-pin".
9 Aug
9 Aug
3:51 p.m.
An interesting read. One thing that jumped out at me
when I got to the second
page of the article was how wide that screen was compared to its height. Do
you know offhand what the screen was, in terms of characters?
No, but the article mentioned that the screen was reduced in height to
make it look like an IBM punch card, so I'd guess 12 lines of 80 characters.
Philip
7:24 p.m.
I'm surprised... well, not that much, I have lurked on this list for a
while now.
When I first read this article and saw how many responses it got I
thought that now I would finally learn some more about my old DP2200.
But the discussion immediately changed into a discussion about various
chip carriers.
In a vane attempt to steer the discussion back into the original topic,
the Datapoint 2200, I wonder if there is anyone else that is restoring
these machines and if there are any schematics out there.
I added a small teaser with some pictures of a 4 kbyte memory card, the
only thing I have of it in my apartment.
http://www.home.neab.net/DP2200/
Some day I will try to restore it.
/G?ran
Jim Brain wrote:
I don't mean to repost, but I had not seen this
posted as yet:
http://www.computerworld.com/action/article.do?command=viewArticleBasic&…
10 Aug
10 Aug
5:05 a.m.
G?ran Axelsson wrote:
I'm surprised... well, not that much, I have
lurked on this list for a
while now.
When I first read this article and saw how many responses it got I
thought that now I would finally learn some more about my old DP2200.
But the discussion immediately changed into a discussion about various
chip carriers.
In a vane attempt to steer the discussion back into the original topic,
the Datapoint 2200, I wonder if there is anyone else that is restoring
these machines and if there are any schematics out there.
I added a small teaser with some pictures of a 4 kbyte memory card, the
only thing I have of it in my apartment.
http://www.home.neab.net/DP2200/
Neat!
I had a Datapoint 2200 for a while. It didn't work either. I see yours
is a first generation (bit serial). Mine was the 2nd generation CPU
which was still discreet TTL, but was data parallel.
Eventually I sold my 2200 due to (1) lack of time and (2) lack of any
documents. I had contacted a couple people who had schematics and such,
but neither was willing to make copies or loan their docs to me so I
could copy them.
I even went to a reunion of datapoint employees and found one guy who
had just about all datapoint internal engineering documents and change
notices in his barn. However, 6 months prior he moved to a smaller
place, so he incinerated everything!
The person I sold my machine to, David Monroe, was one of the engineers
who originally designed the machine. He has schematics too. I never
found out if he got it working again.
5835
days inactive
5839
days old
53 comments
21 participants
participants (21)
-
alexandre-listas＠e-secure.com.br -
axelsson＠acc.umu.se -
bfranchuk＠jetnet.ab.ca -
brain＠jbrain.com -
cclist＠sydex.com -
dkelvey＠hotmail.com -
eric＠brouhaha.com -
ethan.dicks＠usap.gov -
frustum＠pacbell.net -
hilpert＠cs.ubc.ca -
legalize＠xmission.com -
mcguire＠neurotica.com -
onymouse＠garlic.com -
Paul_Koning＠Dell.com -
pete＠dunnington.plus.com -
philip＠axeside.co.uk -
ploopster＠gmail.com -
rickb＠bensene.com -
rtellason＠verizon.net -
schwepes＠moog.netaxs.com -
steven.alan.canning＠verizon.net