Maker Pro
Maker Pro

So where's the rest of the 74HC family?

M

Michael

Jan 1, 1970
0
I need to replace a 3.6 MHz VCO which drives a uC. Looking at xtals, TCO's and
OCXO's in DigiKey, you'd think that nobody ever used - or has use for - a 3.6
MHz clock. There's some high speed, crazy freq. stuff but no 3.6 MHz.

OK, I'll bag accurate and stable and settle for a vanilla osc. Can't find a 3.6
MHz. Try a multiple of 3.6 How about 7.2 (3.6 x 2)? No got, but there is
7.3728. Or 10.8 (3.6 x 3) or 18.0 (3.6 x 5)? (Odd multiples suck but I'm
getting desperate.)? Plenty of 10.24. But I DO find 18.0 and - wonder of
wonders! - 36.0! (3.6 x 10)

I'll get a 36 MHz and run the output thru a decade counter. Lessee ... vanilla
74 family can't hack it. We'll go with 74HC. (sound of shuffling pages here)
Here 'tis, but only basic functions - OR, AND, etc - and some others ... except
counters. What happened to all the 74HC stuff over the past 10 years or so,
while I wasn't looking?!
 
H

Hal Murray

Jan 1, 1970
0
I'll get a 36 MHz and run the output thru a decade counter. Lessee ... vanilla
74 family can't hack it. We'll go with 74HC. (sound of shuffling pages here)
Here 'tis, but only basic functions - OR, AND, etc - and some others ... except
counters. What happened to all the 74HC stuff over the past 10 years or so,
while I wasn't looking?!

They are still available, maybe not as easily as they were a while
ago. Try the HC192 or HC390. Digikey has them.

But I don't think HC will solve your problem. Even the simple
HC74 only goes at 25 MHz at 4.5V over temp. You might squeek
through if you keep the voltage up keep it cool.

Have you considered one of the cut-to-order osc places?
In the old days they would actually make crystals to any
frequency reasonably quickly. Now you can get osc packages
that have a PLL and are programmed to order. The downside
is jitter from the PLL which might matter if you are running
an A/D.
 
1) Why does a uC need a VCO?
2) 3.6MHz sounds close to the NTSC colorburst of 3.579545 MHz
3) No one uses discrete logic much anymore. If they do, it's in a
hand-unusable package.
 
M

Michael

Jan 1, 1970
0
1) Why does a uC need a VCO?
2) 3.6MHz sounds close to the NTSC colorburst of 3.579545 MHz
3) No one uses discrete logic much anymore. If they do, it's in a
hand-unusable package.

Answers:
1) Ever hear of "disciplined oscillator"? This uC keeps itself slaved to a
remote system.
2) The numbers you cite are indeed close to each other. Surely you don't mean
to imply that a 3.6 MHz crystal is a direct replacement for a color burst
crystal.
3) On the contrary, I use discrete logic, and this uC is in a handheld package.

School's over now.
 
1) You could have said so. Why where you going on and on about crystals
and dividing down frequencies if the voltage CONTROL part is important?
2) No, I meant the reverse. That is what you were asking for right? You
want to replace a 3.6 MHz VCO with a crystal? Why would you think I
implied 3.6MHz is a replacement for a color burst? I said the reverse.
The other way around. Reverse.
3) I meant packages are so tiny you can't see them. You know, the
actual part itself, the package. Like DIP or SOIC. Package. Tiny.
Assemble by hand you do not.

Recess is over now.
 
T

Tim Shoppa

Jan 1, 1970
0
Michael said:
Answers:
1) Ever hear of "disciplined oscillator"? This uC keeps itself slaved to a
remote system.

In which case it probably does timekeeping by tweaking a drift rate and
counters inside the microcontroller.

I'm still confused as to what you're writing about, "Crystal" and "VCO"
are not necessarily exclusionary concepts but now you're throwing in
microcontrollers too.

Do you actually need (maybe) a 3.6MHz VCXO module? Or a little LC
oscillator with a varactor?
2) The numbers you cite are indeed close to each other. Surely you don't mean
to imply that a 3.6 MHz crystal is a direct replacement for a color burst
crystal.

No, it isn't, but I've seen 3.579545 MHz crystals labeled "3.6"
especially in surface mount where you don't have a lot of room to stick
all those digits :).

Digikey does stock 7.2 MHz crystals BTW. And if you're looking for
4-bit counters you will probably end up with slightly-higher
integration HC393's, HC390's in async or HC161/HC163 sync counters
today. A couple counters and a 74HC4046 PLL/VCO/phase detector can let
you synthesize almost any frequency out of any other if worse comes to
worse.

Jameco has a pretty good selection of older 7400 families etc.

Tim.
 
T

Tim Shoppa

Jan 1, 1970
0
Michael wrote
[email protected] wrote

1) Why does a uC need a VCO
2) 3.6MHz sounds close to the NTSC colorburst of 3.579545 MH
3) No one uses discrete logic much anymore. If they do, it's in
hand-unusable package

Answers
1) Ever hear of "disciplined oscillator"? This uC keeps itsel slaved to
remote system
In which case it probably does timekeeping by tweaking a drift rat
an
counters inside the microcontroller

I'm still confused as to what you're writing about, "Crystal" an
"VCO
are not necessarily exclusionary concepts but now you're throwing i
microcontrollers too

Do you actually need (maybe) a 3.6MHz VCXO module? Or a little L
oscillator with a varactor
2) The numbers you cite are indeed close to each other. Surely yo don't mea
to imply that a 3.6 MHz crystal is a direct replacement for a colo burs
crystal
No, it isn't, but I've seen 3.579545 MHz crystals labeled "3.6
especially in surface mount where you don't have a lot of room t
stic
all those digits :)

Digikey does stock 7.2 MHz crystals BTW. And if you're looking fo
4-bit counters you will probably end up with slightly-highe
integration HC393's, HC390's in async or HC161/HC163 sync counter
today. A couple counters and a 74HC4046 PLL/VCO/phase detector can le
you synthesize almost any frequency out of any other if worse comes t
worse

Jameco has a pretty good selection of older 7400 families etc

Tim
 
M

Michael

Jan 1, 1970
0
Tim said:
In which case it probably does timekeeping by tweaking a drift rate
and
counters inside the microcontroller.

I'm still confused as to what you're writing about, "Crystal" and
"VCO"
are not necessarily exclusionary concepts but now you're throwing in
microcontrollers too.

Do you actually need (maybe) a 3.6MHz VCXO module? Or a little LC
oscillator with a varactor?

No, it isn't, but I've seen 3.579545 MHz crystals labeled "3.6"
especially in surface mount where you don't have a lot of room to
stick
all those digits :).

Digikey does stock 7.2 MHz crystals BTW. And if you're looking for
4-bit counters you will probably end up with slightly-higher
integration HC393's, HC390's in async or HC161/HC163 sync counters
today. A couple counters and a 74HC4046 PLL/VCO/phase detector can let
you synthesize almost any frequency out of any other if worse comes to
worse.

Jameco has a pretty good selection of older 7400 families etc.

Tim.


Thank-you for the suggestions, Tim.

The vintage-early-80's broken box has a VCXO made out of discretes, including a
varactor and a 3.6 MHz, leaded XTAL. This osc. goes completely tits-up - stops
- when temperature rises above about 55F. Swapping out the various components
(except XTAL) showed that they are OK; after replacing the 3.6 XTAL with a
3.6864, the osc. worked fine overnight (though the rest of the box was quite
confused).

Of course a 3.6 substituted for a color burst would give unsatisfactory
results. My point was that substituting a color burst XTAL for my 3.6MHz -
implied by the othe
r poster's "3.6MHz sounds close to the NTSC colorburst" - would give similarly
unsatisfactory service.

Mea culpa: I read through the DigiKey pages - for the umpteenth time - yesterday
afternoon and did find one, just one!, 7.2MHz XTAL. Packaged like the little
watch XTALs, says it can't handle very much power, so I dunno if it will work in
my VCXO. It's dirt cheap though, so worth a try.

The flaky XTAL was made by Fox. I surfed their site today and saw that, sadly,
they have almost completely transitioned to surface mount. Many other kinds of
components are also going SM, of course, and so I despair for today's budding
electrical and electronics engineers and hobbyests.

I will keep your 74HC4046 PLL/VCO/phase detector idea in mind, just in case.
 
T

Tim Shoppa

Jan 1, 1970
0
Michael said:
The vintage-early-80's broken box has a VCXO made out of discretes, including a
varactor and a 3.6 MHz, leaded XTAL.

I would be very critical of just being able to plop in "any old" 3.6MHz
crystal in this place, much less one at a multiple of the frequency.
The rubber (pullability) you get with a crystal is very dependent on
multiple crystal parameters, not just frequency.

If you can characterize the VCXO part of the flaky crystal in-circuit
(e.g. 3V change in control gives 20ppm shift) you'll have better luck
at least testing your replacement.

Just because the crystal quits oscillating above some temperature
doesn't necessarily point the finger at the crystal - any part of the
oscillator might have gain depend on temperature. I would finger the
semiconductors before the crystal. Any polypropylenes in that
oscillator? If they've ever been above 85C they can start giving very
odd symptoms.
Mea culpa: I read through the DigiKey pages - for the umpteenth time - yesterday
afternoon and did find one, just one!, 7.2MHz XTAL.

Their website is very nice for such searches.
The flaky XTAL was made by Fox.

Is it vaguely possible you can find a "parts" unit for stripping out a
crystal?

Tim.
 
Top