Classic Chips

[Paul Burridge]

Hi all,

I've been following the discussion on the thread about chips that have
been around for 30 years and was just wondering if anyone would care
to nominate their favourite 'classic chip' of all time. For example
the chip that they believe deserves, for whatever reason, to be marked
out as being of exceptional merit or usefulness.

p.

 

[Sir Charles W. Shults III]

For me, it would probably be a runoff between the 723 voltage regulator and
the 741 op amp. Both have a great deal of flexibility and have survived ages.
You can make almost anything with them.

Cheers!

Chip Shults
My robotics, space and CGI web page - http://home.cfl.rr.com/aichip

 

[Wouter van Ooijen]

723, 741

add the 555



Wouter van Ooijen

-- ------------------------------------
http://www.voti.nl
PICmicro chips, programmers, consulting

 

[ELAL]

Hi all,

I've been following the discussion on the thread about chips that have
been around for 30 years and was just wondering if anyone would care
to nominate their favourite 'classic chip' of all time.

I'd nominate the 555 (and its siamese twin brother 556).
Can you name any IC that matches both its simplicty AND versatility?

 

[John Jardine]

Hi all,

I've been following the discussion on the thread about chips that have
been around for 30 years and was just wondering if anyone would care
to nominate their favourite 'classic chip' of all time. For example
the chip that they believe deserves, for whatever reason, to be marked
out as being of exceptional merit or usefulness.

p.

Undoubtably the LM393.

 

[Spehro Pefhany]

Hello Paul,
my favorite will be the LM324. It is nowadays still a good choice,
because it can be used in single supply systems.

The 741 in contrast doesn't have an input common mode range down to the
negative supply voltage and it need much higher supply voltages.

The LM324 was later on the market, but it will last longer.

I agree. I've used the LM324 (and it's half-brother the LM358) more in
new designs than the others (the most recently for an NiMH charger
current sink in a portable instrument). The 1mA it draws didn't matter
in this case. I can't think of why I'd use a uA741. I used the uA723 a
decade or so ago in a product and it would still be the right decision
if a similar problem (programmable linear power module for pressure
tranducers, process transmitters etc.) if the same problem came up
today. I still have some TO-99 uA709As around (with a different
Toshiba part number) used for repairing old temperature controllers.

The 4558 is also very popular for low-end audio (it's something like a
dual LM741). Back before good op-amps got cheap I built a
computer-controlled test rig to screen hundreds at a time for
temperature drift. I designed a circuit that used two 4558's, only one
of which was critical, and the yield was always such that there was no
waste.

Best regards,
Spehro Pefhany

 

[John Larkin]

Hi all,

I've been following the discussion on the thread about chips that have
been around for 30 years and was just wondering if anyone would care
to nominate their favourite 'classic chip' of all time. For example
the chip that they believe deserves, for whatever reason, to be marked
out as being of exceptional merit or usefulness.

p.

7805.

John

 

[Jim Thompson]

Hi all,

I've been following the discussion on the thread about chips that have
been around for 30 years and was just wondering if anyone would care
to nominate their favourite 'classic chip' of all time. For example
the chip that they believe deserves, for whatever reason, to be marked
out as being of exceptional merit or usefulness.

p.

My G-job work-horses are the LM324 and the LM339... you can build
almost anything with those two devices. (I can even build a 555
equivalent ;-)

...Jim Thompson

 

[Mark]

Hi all,

I've been following the discussion on the thread about chips that have
been around for 30 years and was just wondering if anyone would care
to nominate their favourite 'classic chip' of all time. For example
the chip that they believe deserves, for whatever reason, to be marked
out as being of exceptional merit or usefulness.

p.

For me it would be the 4046 PLL.

M.

 

[Jim Thompson]

For me it would be the 4046 PLL.

M.

My design of the MC4024/MC4044 was first ;-)

...Jim Thompson

 

[Michael Black]

7805.

John

Wait a minute.

Was it a "7805" when introduced? I remember coming across the LM309
first, which was of course a 3-terminal 5volt regulator, and maybe
specifically in a TO-3 package (and a lower current TO-5). National
had other voltage regulators in their line, with the LM340-XX designator
(where XX was the actual voltage), which I don't know if they came after
the 309 or I simply heard of them a tad later. They were available in TO-3
and TO-220 packages. And what never quite made sense, there was a 5V
version too.

A 1972 National databook does not show a "7805", but I can't remember if
anyone else would have used that designator before National.

Admittedly, there seems to be a few differences in the internal circuitry
between the 309 and the 340-XX line, but surely the usefulness of
the three terminal regulator dates from when it was first succesfully
available.

(Someone did have a 3-terminal regulator in the sixties, maybe about
1968, but it was from an obscure manufacturer and I saw only one
reference to it.)

Michael

 

[Michael Black]

My design of the MC4024/MC4044 was first ;-)

...Jim Thompson

Somewhere around, I have a clipping from Electronics magazine of
a circuit sent in by a reader for making a faster 4044 out of
discrete gates. I don't think it was ECL, more like 74AS, but
it's been a while since I last looked at it.

I wouldn't deny the importance of the 4044, but I suspect
the 4046 despite it's frequency limitations, was a more general
device. After all, with the built in VCO it could be use where
a simple PLL was required (as long as in the low frequency range)
and that VCO meant that it fit into lots of uses where people merely
wanted a simple oscillator or VCO, tossing aside the phase comparator.

Then of course, there was that Signetics line of analog PLLs, the 565,
560, 561 and such. I'm not sure how much commercial use they saw, but
they made a wave when they first came out in hobby circles, "wow, we
can now have a PLL in a single package". Never mind that the analog
phase detector wasn't the best choice for some of the circuits.

Michael

 

[George R. Gonzalez]

74HC14... Makes a swell schmitt triger, oscillator, monostable, inverter,
....

 

[Peter O. Brackett]

Paul:

I nominate:

1488/1489 RS-232 drivers and recievers. They "bridge" the analog and
digital worlds.

If you could count the number of RS-232 ports shipped over the past 30
years, each such port contains at least 2 each and often more, it probably
totals in the billions. i.e. every modem had at least one, every
multiplexer had 8 or more, often up to 256, every router, every computer,
every bridge, etc... had at least one, often many, and each generation of
modem, mux, computer, router, bridge, etc... appeared and was shipped to
replace the older ones that were junked.

I'm willing to bet a few bucks that there have been more 1488/1489 chips
[and their equivalents] than any other chip.

Thoughts, comments?

 

[Geoffrey G. Rochat]

I've been following the discussion on the thread about chips that have

been around for 30 years and was just wondering if anyone would care
to nominate their favourite 'classic chip' of all time. For example
the chip that they believe deserves, for whatever reason, to be marked
out as being of exceptional merit or usefulness.

I nominate the RCA CD4057A COS/MOS LSI 4-bit arithmetic logic unit. That
was a 4-bit slice of computer brains that not only included the ALU, but
also a 4-bit shift/storage register and an arithmetic overflow flip-flop to
boot. The interesting thing about the 4057, though, was that it had 2
control lines that allowed the chip to function as any 4-bit section of a
larger ALU, or even as a separable section of a larger ALU. With four of
these babies, for example, you could build a computer that was dynamically
changeable into a 16-bit machine, a 12-bit and a 4-bit machine, or two 8-bit
machines. Oh, the power! Oh, the flexibility! Oh, the elegance! (Oh, the
impenetrable complexity!)

My second favorite is Motorola's MC14500B Industrial Control Unit, a
single-bit computer. No, not a single-chip computer (far from it), a
single-bit computer. Ayup. They were just the thing for replacing relay
tree logic with chips - that would then drive output relays.

Please note that the given acceptance qualification for a 'classic chip' is
that it be of exceptional merit OR exceptional usefulness. Both criteria
need not apply, and there is nothing to say that a 'classic chip' can't be
long gone and forgotten. And while the magnitude of the exceptional merit
vector of the above chips is unquestionable, there is perhaps the matter of
that vector's sign...

Also, please note that the acceptance qualification says nothing about a
'classic chip' being one of those funny analogalisticous things. But if you
insist on that sort of stuff, I nominate the National Semiconductor LM373
AM/FM/SSB IF strip. Supposedly, it could do AM, CW, SSB or FM detection,
and was intended for everything in communications. Unfortunately, contrary
to National application note AN-54 (April 1972, in the 1973 "Linear
Applications Handbook 1"), or linear brief LB-13 (November 1970, same
place), or a dimly-remembered article in an old "Ham Radio" magazine
somewhere, I don't think anyone actually ever got a circuit based on the
chip to work.

 

[Jim Thompson]

Paul:

I nominate:

1488/1489 RS-232 drivers and receivers. They "bridge" the analog and
digital worlds.

[snip]

Absolutely wonderful set of chips...

Particularly since I designed them ;-)

...Jim Thompson

 

[Leon Heller]

I nominate the RCA CD4057A COS/MOS LSI 4-bit arithmetic logic unit. That
was a 4-bit slice of computer brains that not only included the ALU, but
also a 4-bit shift/storage register and an arithmetic overflow flip-flop to
boot. The interesting thing about the 4057, though, was that it had 2
control lines that allowed the chip to function as any 4-bit section of a
larger ALU, or even as a separable section of a larger ALU. With four of
these babies, for example, you could build a computer that was dynamically
changeable into a 16-bit machine, a 12-bit and a 4-bit machine, or two 8-bit
machines. Oh, the power! Oh, the flexibility! Oh, the elegance! (Oh, the
impenetrable complexity!)

My second favorite is Motorola's MC14500B Industrial Control Unit, a
single-bit computer. No, not a single-chip computer (far from it), a
single-bit computer. Ayup. They were just the thing for replacing relay
tree logic with chips - that would then drive output relays.

Please note that the given acceptance qualification for a 'classic chip' is
that it be of exceptional merit OR exceptional usefulness. Both criteria
need not apply, and there is nothing to say that a 'classic chip' can't be
long gone and forgotten. And while the magnitude of the exceptional merit
vector of the above chips is unquestionable, there is perhaps the matter of
that vector's sign...

Also, please note that the acceptance qualification says nothing about a
'classic chip' being one of those funny analogalisticous things. But if you
insist on that sort of stuff, I nominate the National Semiconductor LM373
AM/FM/SSB IF strip. Supposedly, it could do AM, CW, SSB or FM detection,
and was intended for everything in communications. Unfortunately, contrary
to National application note AN-54 (April 1972, in the 1973 "Linear
Applications Handbook 1"), or linear brief LB-13 (November 1970, same
place), or a dimly-remembered article in an old "Ham Radio" magazine
somewhere, I don't think anyone actually ever got a circuit based on the
chip to work.

I got a couple and played around with them when they first came out. IIRC I
got one working OK for AM. For SSB I think I had problems with the CIO
leaking into the input and swamping it. It was a long time ago, though.

The Plessey RF and IF amplifiers (612 etc.) were nice chips.

Leon

 

[kansas_ray]

My design of the MC4024/MC4044 was first ;-)

...Jim Thompson

Hi Jim,

Would you give us some history? Why did the 4046 become the "de-facto"? It
would give us outsiders (and youngsters) some insight.

Best regards,
Ray

 

[Mike Diack]

If imitation is the sincerest form of flattery then surely the Intel
8031 deserves honorable mention.
M

 

[Jim Thompson]

Jim:

[snip]

Absolutely wonderful set of chips...

Particularly since I designed them ;-)

...Jim Thompson

[snip]

I knew that!

But don't you think I'm right?

There are lots more RS-232 ports out there than there are networking and
computing devices sold and they all use the 1488/89 or equivalent. Gotta be
lots more of those shipped and sold than any old analog chips. Perhaps as
many 1488/89's as 7404's, eh?

BTW...

I'll betca' you wish you got a 1% royalty on everyone sold.

God Bless the "Batwings". ;-)

What's really distressing is that those were "accidental" designs...
wasn't in my department (I was A-D-A at the time). Young buck
engineer comes by one afternoon and says "I have a problem", and I
sketched them both out on the back of an envelope.

(For Win's information... the driver uses lateral PNPs.)

...Jim Thompson