Maker Pro
Maker Pro

Why don't all CD4000 chips have Schmitt inputs?

J

Joerg

Jan 1, 1970
0
Hi All,

Just out of curiosity: Why is it that flip-flops and other chips in the
CD4000 series are "schmitt-less"?

So far I design at least 80% of the stuff around 40106 chips, simply
because they interface to analog very well because of their Schmitt
trigger inputs. But making a flip flop always costs a third of the chip
plus two resistors, meaning valuable space. If the 4013, 40175 and all
those had Schmitts this would open a whole lot of applications. After
all, this stuff is never used on fast circuitry anyways. I'd be happy if
they left maybe one non-inverting buffer sans Schmitt so it could be
(ab)used as an analog amp.

Regards, Joerg
 
J

John Larkin

Jan 1, 1970
0
Hi All,

Just out of curiosity: Why is it that flip-flops and other chips in the
CD4000 series are "schmitt-less"?

So far I design at least 80% of the stuff around 40106 chips, simply
because they interface to analog very well because of their Schmitt
trigger inputs. But making a flip flop always costs a third of the chip
plus two resistors, meaning valuable space. If the 4013, 40175 and all
those had Schmitts this would open a whole lot of applications. After
all, this stuff is never used on fast circuitry anyways. I'd be happy if
they left maybe one non-inverting buffer sans Schmitt so it could be
(ab)used as an analog amp.

Regards, Joerg


Schmitts are usually slow. They probably didn't want to give up the
speed on an entire logic family. 4000's are pokey enough already.

John
 
T

Tim Wescott

Jan 1, 1970
0
Joerg said:
Hi All,

Just out of curiosity: Why is it that flip-flops and other chips in the
CD4000 series are "schmitt-less"?

So far I design at least 80% of the stuff around 40106 chips, simply
because they interface to analog very well because of their Schmitt
trigger inputs. But making a flip flop always costs a third of the chip
plus two resistors, meaning valuable space. If the 4013, 40175 and all
those had Schmitts this would open a whole lot of applications. After
all, this stuff is never used on fast circuitry anyways. I'd be happy if
they left maybe one non-inverting buffer sans Schmitt so it could be
(ab)used as an analog amp.

Regards, Joerg

If they did that, how would we make cheezy crystal oscillators?

Besides, when they first came out they were the only CMOS in town, and
probably needed to be as fast as possible.
 
J

Joerg

Jan 1, 1970
0
Hi John,
Schmitts are usually slow. They probably didn't want to give up the
speed on an entire logic family.
The hex Schmitt CD40106 has a prop delay around 50-150nsec depending on
VCC and similar transition times. Not bad at all for ordinary logic
applications. The non-Schmitt devices aren't much different. I have used
them to make switch mode supplies and all kinds of other stuff.
4000's are pokey enough already.
I think they are still great chips. Nearly always multi-sourced, no
leadtimes, next to nothing in quiescent current, VCC can be all over the
place, no problem making timers with several megohms and a tiny ceramic
cap. Most important to me is that they are only pennies per section.

How many times have we seen articles about the demise of the CD series?
Then all of a sudden they came out with them in TSSOP. To me that is a
hard indicator that CD4000 is here to stay.

When I was a kid and everybody used TTL people scoffed when I used the
4000. In those days they were kind of fickle if you weren't careful with
ESD and some folks called them Sissy-logic. Thing is, my stuff usually
ran for more than a year on a 4.5V flashlight battery (remember those?)
where the others needed chunky power supplies.

Now I just wish they made a version that can operate between 1V and 3V.
Got to have dreams...

Regards, Joerg
 
J

Joerg

Jan 1, 1970
0
Hi Tim,
If they did that, how would we make cheezy crystal oscillators?

Well, back to the roots I guess: With a transistor. "A what?... yecch!"
was the answer the last time I did that. Then when I proposed the BFS17
in the clock distribution the digital guys almost wanted to wrap me in
duct tape.

Otherwise there is always the CD4000UB. You can make all kinds of neat
analog circuits with that for about 15-20 cents a pop.
Besides, when they first came out they were the only CMOS in town, and
probably needed to be as fast as possible.

Yes, but that was when I still had hair on my head and listened to the
Stones with the volume on ten ;-)

I thought that some progress could have been made by now.

Regards, Joerg
 
T

Tim Wescott

Jan 1, 1970
0
Joerg said:
Hi Tim,



Well, back to the roots I guess: With a transistor. "A what?... yecch!"
was the answer the last time I did that.

A transistor, two resistors, another transistor or a CMOS gate to bring
it up to logic levels -- why not just use the crystal on the one CMOS
gate & call it good?
Then when I proposed the BFS17
in the clock distribution the digital guys almost wanted to wrap me in
duct tape.

I want to work on a project with you, just for the entertainment value.
Otherwise there is always the CD4000UB. You can make all kinds of neat
analog circuits with that for about 15-20 cents a pop.
If you don't mind sloooow.
Yes, but that was when I still had hair on my head and listened to the
Stones with the volume on ten ;-)
What? Not 11?
I thought that some progress could have been made by now.
Well, yea, but anybody who tried to change things without changing part
numbers would have all these customers screaming, and anybody who made
new versions would want to charge way more than pennies per gate.
 
F

Frank Bemelman

Jan 1, 1970
0
Joerg said:
Hi All,

Just out of curiosity: Why is it that flip-flops and other chips in the
CD4000 series are "schmitt-less"?

So far I design at least 80% of the stuff around 40106 chips, simply
because they interface to analog very well because of their Schmitt
trigger inputs. But making a flip flop always costs a third of the chip
plus two resistors, meaning valuable space. If the 4013, 40175 and all
those had Schmitts this would open a whole lot of applications. After
all, this stuff is never used on fast circuitry anyways. I'd be happy if
they left maybe one non-inverting buffer sans Schmitt so it could be
(ab)used as an analog amp.

HEF4013 has a schmitt trigger clock input. I believe there were some
other HEF4xxx chips that also use schmitt trigger inputs where you don't
expect them, but I am not sure which ones.
 
J

Joerg

Jan 1, 1970
0
Hi Tim,
A transistor, two resistors, another transistor or a CMOS gate to
bring it up to logic levels -- why not just use the crystal on the one
CMOS gate & call it good?

That is the way I do it, too. But not in very noise critical apps such
as ultrasound with Doppler or communications.
Then when I proposed the BFS17 in the clock distribution the digital
guys almost wanted to wrap me in duct tape.

I want to work on a project with you, just for the entertainment value.

Would be fun. I guess our background isn't that different and includes
lots of analog. BTW, that BFS17 case was where they had hung all kinds
of FPGA onto a clock line and the clock looked like the roller coaster
at Disneyland. Too late for a major relayout for a home-run style clock
distribution but we still could do "local" mods. Luckily they had some
other voltages on the board. Oh, and then we had a nice argument about
AC termination. That's voodoo, they said. Until we tried it.
What? Not 11?

10 was where my stereo pegged. Later we made an amp with a few huge
tubes and that, together with E-guitars, really rocked when used
outdoors. Until the laws came...
Well, yea, but anybody who tried to change things without changing
part numbers would have all these customers screaming, and anybody who
made new versions would want to charge way more than pennies per gate.

That is certainly a problem. Especially in view of all the folks that
have used CD chips in analog designs. You'd have to call the something
else, like CD5000.

Regards, Joerg
 
J

Joerg

Jan 1, 1970
0
Hi Frank,
HEF4013 has a schmitt trigger clock input. I believe there were some
other HEF4xxx chips that also use schmitt trigger inputs where you don't
expect them, but I am not sure which ones.
Thanks, I didn't know that. Have to check it out. I just wish Philips
would learn how to create a practical web site.

Regards, Joerg
 
N

nospam

Jan 1, 1970
0
Joerg said:
How many times have we seen articles about the demise of the CD series?

I haven't but then I don't read many. Probably written by someone who
thinks thing go obsolete because they are old - they don't.

Things go obsolete because the market disappears or because something
better replaces them. A limited market for some 4000 series parts will be
around for ages. There is little scope to improve the power consumption and
anyone wanting to go faster has had alternative logic families for years.
 
S

Spehro Pefhany

Jan 1, 1970
0
HEF4013 has a schmitt trigger clock input. I believe there were some
other HEF4xxx chips that also use schmitt trigger inputs where you don't
expect them, but I am not sure which ones.

For example, the clock on 4017, 4020, 4022, 4024, 4040 (but not the
clock gate input).


Best regards,
Spehro Pefhany
 
T

Tim Wescott

Jan 1, 1970
0
Joerg said:
Hi Tim,



That is the way I do it, too. But not in very noise critical apps such
as ultrasound with Doppler or communications.

Well, then you don't need a cheezy oscillator, you need a good one. I'd
certainly not use a logic gate for the oscillator in _that_ case.
Would be fun. I guess our background isn't that different and includes
lots of analog.

Actually I'm in the odd position of being an analog circuit designer who
writes lots of signal processing code - it comes from being interested
in making the signal behave, and using what works. This causes
consternation with circuit designers who don't know me well,
particularly when they then conclude that I must know digital design as
well (I can put a microcontroller on a board, but that's as far as I go).

-- snip --
 
J

Joerg

Jan 1, 1970
0
Hi Tim,
Actually I'm in the odd position of being an analog circuit designer
who writes lots of signal processing code - it comes from being
interested in making the signal behave, and using what works. This
causes consternation with circuit designers who don't know me well,
particularly when they then conclude that I must know digital design
as well (I can put a microcontroller on a board, but that's as far as
I go).

Same here but I don't write code. Except maybe some C and a few 8051
routines. I want to dive into controllers some more because nowadays
they become so powerful that they can replace analog circuits at times.
The only one that is truly capable of matching the usual power
constraints is the MSP430 so when I have free time I'll check that out.
Wish they made a 5V version though so it could drive FETs better.

The only downside of uCs is that the versions with sufficient on-board
capabilities are still above $1 and that precludes many applications.

Regards, Joerg
 
J

Joerg

Jan 1, 1970
0
Hi John,
40106. Must be the most handy GP chip out there.
Finished last week a lo-cost, hi-volume, timer using a 4060 + 40106. One
inverter for a trigger. 2 inverters for an RS latch. Two inverters for
internal monostables. One inverter for power up hold-off.
I've yet to find a simpler, cheaper method of generating delays than a
C+R+schmitt inverter.
If a customers priority is 'cheap' then it's odds on that 40106's will turn
up in the design.
Can you imagine that some engineers of the newer generation don't even
know the 4000 series? It blew my mind. Did they ever have to be within a
budget in their life?

Regards, Joerg
 
F

Frank Bemelman

Jan 1, 1970
0
Spehro Pefhany said:
For example, the clock on 4017, 4020, 4022, 4024, 4040 (but not the
clock gate input).

You really have an impressive memory, Speff ;-) I checked the 40175 but
that one is 'normal'. You would expect some consistency within a family.
My favorite 4000 chips were 4013, 14557 (shiftreg) and 14538 (dual timer).
 
J

Joerg

Jan 1, 1970
0
Hi,

Sorry but I don't know your name.
I haven't but then I don't read many. Probably written by someone who
thinks thing go obsolete because they are old - they don't.
The ones I have read were written by the same authors that touted the
latest and greatest FPGA in other articles. The arrays where you had to
inform your electricity provider before flipping the power switch. We
used to call them "coffee warmers".
Things go obsolete because the market disappears or because something
better replaces them. A limited market for some 4000 series parts will be
around for ages.
The market actually seems to be growing, considering the newly offered
smaller packages.
There is little scope to improve the power consumption and
anyone wanting to go faster has had alternative logic families for years.
Well, power consumption on most CD4000 is pretty close to zilch. You can
have half a dozen of these chips, leave them on for a year and the
battery will still be as good as new.

The new very low voltage stuff from TI is really impressive. 74AUP and
others. Some can operate on a single cell and swing the output under
1nsec. I am waitning for 2nd sources and prices to drop, then I'll
design with these.

Regards, Joerg
 
J

john jardine

Jan 1, 1970
0
Joerg said:
Hi All,

Just out of curiosity: Why is it that flip-flops and other chips in the
CD4000 series are "schmitt-less"?

So far I design at least 80% of the stuff around 40106 chips, simply
because they interface to analog very well because of their Schmitt
trigger inputs. But making a flip flop always costs a third of the chip
plus two resistors, meaning valuable space. If the 4013, 40175 and all
those had Schmitts this would open a whole lot of applications. After
all, this stuff is never used on fast circuitry anyways. I'd be happy if
they left maybe one non-inverting buffer sans Schmitt so it could be
(ab)used as an analog amp.

Regards, Joerg

40106. Must be the most handy GP chip out there.
Finished last week a lo-cost, hi-volume, timer using a 4060 + 40106. One
inverter for a trigger. 2 inverters for an RS latch. Two inverters for
internal monostables. One inverter for power up hold-off.
I've yet to find a simpler, cheaper method of generating delays than a
C+R+schmitt inverter.
If a customers priority is 'cheap' then it's odds on that 40106's will turn
up in the design.
regards
john
 
J

John Woodgate

Jan 1, 1970
0
I read in sci.electronics.design that Joerg <notthisjoergsch@removethisp
acbell.net> wrote (in <[email protected]>)
about 'Why don't all CD4000 chips have Schmitt inputs?', on Sat, 2 Oct
2004:
Can you imagine that some engineers of the newer generation don't even
know the 4000 series? It blew my mind. Did they ever have to be within a
budget in their life?

Not only that but they are low-emission devices, mostly, which is a
great help in meeting EMC requirements.
 
F

Fred Bartoli

Jan 1, 1970
0
John Woodgate said:
I read in sci.electronics.design that Joerg <notthisjoergsch@removethisp
acbell.net> wrote (in <[email protected]>)
about 'Why don't all CD4000 chips have Schmitt inputs?', on Sat, 2 Oct
2004:

Not only that but they are low-emission devices, mostly,


You mean those engineers ?


which is a
great help in meeting EMC requirements.

Enginnering-Marketing Compatibility ?
 
J

John Woodgate

Jan 1, 1970
0
I read in sci.electronics.design that Fred Bartoli <fred._canxxxel_this_
bartoli@RemoveThatAlso_free.fr_AndThisToo> wrote (in <415e9d81$0$24028$6
[email protected]>) about 'Why don't all CD4000 chips have Schmitt
inputs?', on Sat, 2 Oct 2004:
You mean those engineers ?

Oops! Interspersion frolic. But yes, I suspect that many who despise the
4000 series are low-emission. Of effective and economical designs.
which is a

Enginnering-Marketing Compatibility ?

No, that requires meeting the marketing version of the Olympic motto;
'Faster, Smaller, Cheaper'. If 'faster' applies to 'time to market'
rather than processing speed, then the 4000 series meets it, especially
with the new packages.

I meant 'Electro-Magnetic Compatibility'; you may have vaguely heard of
it. (;-)
 
Top