[time-nuts] Have 10 MHz need 19.2 MHz

Perry Sandeen sandeenpa at yahoo.com
Sat Jun 15 04:29:27 UTC 2013



List,
 
My reply to Javier Herrero<jherrero at hvsistemas.es> 
 
P  Double
the 10 MHz to 20 MHz.
 
P  With
another circuit of 74HC390’s divide 10 MHz to 200 KHz.  Then double it twice to 800 KHz with LM 1496
DBM’s.  Apply the two frequencies to a LM
1496 DBM and use a LPF to get the 19.2 MHz.
 
P  Hardware
complicated?  A bit.
 
J  >Only a bit?  Only the filter to reject the products that you will have spread in all
places, spaced 200 kHz,
 
P  I’m not so sure of that at all. The DBM doubler
has a difference frequency of zero. And I don’t understand where did you get 200
KHz spaces?
 
J>  and
mainly to remove the 20.8 MHz spurious that you will have as a result of the
last mixing, makes this approach difficult.
 
P  You are
correct that the sum frequency is 20.8 MHz.  However the difference frequency being 19.2 MHz makes a 1.6 MHz
difference which will allow one to use a simple parallel L-C circuit to get the
frequency of choice rather simply. A parallel resonance filter (ignoring stray
capacitance for the moment) using a .1uH inductor and a 687pF capacitor
resonates at 19.20175 MHz.
 
 J  I would favour a PLL, and since for the
application, short-term stability seems irrelevant, even using a conventional
VCO and not a crystal would be enough. A 74HC4046 can reach 19.2 MHz, and you
only need a couple of dividers to get a 200 kHz reference to feed it.
 
P  That may
be another way to do it.  I didn’t say my
way was thee only way.  I said it might
be a way.  And until one built my circuit
idea criticizing that it won’t work well is hypothetical and somewhat arrogant.  Building it is the real proof.  Never forget about the duck-billed platypus,
the animal that couldn’t be, according to *scientific laws*.  
 
P  According
to the data sheets I have for a 74HC4046 to use it at 19.2 MHz you have to be
brand specific.  Some are only rated to
14 MHz
 
P  I thought
that the 10 MHz source was a rubidium.  That may have been incorrect.  If
the short term stability is irrelevant as you have stated then other
possibilities are also feasible.  
 
P  But if
one uses a 74HC4046 with a VCO (which would be fine with me) one still has to
get two matching frequencies. 
 
J  I figured
out the 200 KHz division.  
 
P  How would
you get 200 KHz from the 19.2 MHz for the PLL and where does this come from and
how does it fit into your circuit.
 
P  Can you
use matching harmonics in a PLL? I don’t know.  I’m a hobbyist and have to go with what I know or can get from others.
 
J > However one doesn’t have to search for a
microprocessor that you program and may not be available in a couple of
years.  The IC’s are cheap and have been
and will be around forever.
 
J  The
LM1496 was discontinued long ago.... it was a second source of the MC1496 (that
is in production). 
 
P  OK.  But a rose is still a rose. Mouser currently
has stock of over 4,000 MC1496 for immediate delivery. They are
interchangeable.  So what is your
point?  The specified type of DBM is
currently available.  Sir, methinks,
respectfully, that thy are quibbling over nothing, 
 
J  But never
think it will be around forever (yes, as a hobbyist, surely you can find a
single piece forever,
more if price does not matter too much). 
 
P  Well some
parts whose use is so pervasive, for example, the 2N3904 and 2N3906 transistor,
they will be around *forever* and reasonably priced.  I believe the MC or LM 1496 falls in this
category.  Everyone had and still uses
them.  Although there are a few other IC
DBM’s they really are insignificant and very hard to find.  
 
J  >Also,
I'm not a bit fan of PICs, quite the contrary, but for example the PIC16F84 has
been available from more that 16yr and it is in production... so following your
LM1496 criteria, will be available forever :)
 
P  OK, but
you still have to be able to program any PIC one chooses.  Most prevalent posters to the list are
programmers or have been trained in programming or like it enough to learn it
and do it.  Fine. Go for it.  Their skill is indispensible  in this world.
 
Now comes the BUT part.  IMNSHO, I believe that their (programmers)
judgment is quite slanted to using microprocessors as a solution to all
electrical problems.  This leads to
over-complicated solutions to many problems.  This is actually poor engineering practice.  Good engineering is finding the least
expensive, reliable, simplest and reproducible method to solve a given problem. 
 
Many of us have no skill to use microprocessors.
So we find a work-a-round. Microprocessors are fine.  Microprocessor use gives us a great quality
of life.  Modern life as developed
nations have would not exist without them.  We couldn’t be spied upon without them!  
 
P  > IMHO
sometimes an older *brute force* circuit proves that more can be less in
implementing what you desire to accomplish.
 
J  Brute
force is usually brute :)
 
P  Perhaps.  But if a circuit works reliably, fulfills its
objective, and is easily reproduced by others that can easily outweigh
sophistication.  An example: the Brooks
Shera GPS system.  It is/was wonderful
but only as long as the custom programmed chips were available and Brooks was
alive to do the source code.  But the
project is useless W/O the programmed chip and the source code.  Now people are scrambling to save it and
program chips.  Still it is a project
that is useless to most of the world as the chip is for them is unobtainium.
 
P  Now
perhaps calling my circuit idea *brute force* was a poor choice of
wording.  Perhaps if I had used words
like *basic or *simple* technology that would have been better.
 
J  So
instead people obtain atomic synced clocks at Wal-Mart.
 
P  How does
this fit into the discussion?  And as an
aside, many they sold do not work with changed WWVB format.
 
Regards,
 
Perrier  
 


________________________________


More information about the Time-nuts_lists.febo.com mailing list