Main Page | Report this Page
 
Loading..
   
Science Forum Index  »  Electronics Forum  »  Looping Binary Counter, Stable 32768 oscillator
Page 1 of 1    
Author Message
GMV
Posted: Fri Jul 29, 2005 10:57 am
Guest
I am in need of a circuit that will place minute mark
,second 59 or second zero of UTC, into my homebuilt circuit
for calibration purposes. I have thought of possibly building
a 32768Hz oscillator and using a looping binary counter that can be
calibrated using WWV or WWVB but this all sounds silly to me
since there must be calibration clocks already out there.
I need either one with a TTL output or even better an FET switch or something
like that with a mark I can feed directly into the back end of my
circuit just before the signal hits the A/D converter.
The stability of the oscillator would be most important
not to deviate more than +/-0.03 Hz.
My laptop has a RTC that looses one second of time every four point five hours
so it is completely unsuitable for anything other than getting the time
close enough to understand what the calibration mark means.

Could someone who is an engineer provide me with
a schematic to do this ?
The basic idea seems simple for those educated in the EE field.
It is the stable oscillator that eludes me.

Any help here is appreciated.
Sincerely;
gmvoeth@yahoo.com
Back to top
GMV
Posted: Fri Jul 29, 2005 3:09 pm
Guest
I think you might be from another country so i just
want to say here in the usa we have a time signal at
60KHz that keeps cheap maybe $20.00 clocks in sync
with the atomic one in colorado and if only some
genius would include an output on the back of one of those
clocks for second zero from the beginning to end of
second zero on the minute mark UTC that is all any amateur should need.
The GPS alternative is maybe 10 times more expensive then
necessary if you live in the USA.
But thanks for the input i will download and read it even
though I am pretty sure they are not producing the device I seek.
The accuracy of the time loop I am talking about will be set manually
or electrically by listening to WWV or WWVB broadcasts.
We are not talking millisends here but only +/- 0.2 seconds or
the human reaction time.

GMV


"Dan Hollands" <dhollan3@rochester.rr.com> wrote in message news:qEtGe.6828$8y3.1472@twister.nyroc.rr.com...
Quote:
There is a difference between stability and accuracy

The best way to get what you want is to use a GPS receiver with a 1PPS output

See http://www.gmat.unsw.edu.au/snap/publications/mumford_2003a.pdf for a review of several GPS receivers each less than $200

Dan

--

Dan Hollands
1120 S Creek Dr
Webster NY 14580
585-872-2606
QuickScore@USSailing.net
www.QuickScoreRace.com


"GMV" <gmvoeth@yahoo.com> wrote in message news:e664$42ea52e8$4fea0a0$5983@DIALUPUSA.NET...
I am in need of a circuit that will place minute mark
,second 59 or second zero of UTC, into my homebuilt circuit
for calibration purposes. I have thought of possibly building
a 32768Hz oscillator and using a looping binary counter that can be
calibrated using WWV or WWVB but this all sounds silly to me
since there must be calibration clocks already out there.
I need either one with a TTL output or even better an FET switch or something
like that with a mark I can feed directly into the back end of my
circuit just before the signal hits the A/D converter.
The stability of the oscillator would be most important
not to deviate more than +/-0.03 Hz.
My laptop has a RTC that looses one second of time every four point five hours
so it is completely unsuitable for anything other than getting the time
close enough to understand what the calibration mark means.

Could someone who is an engineer provide me with
a schematic to do this ?
The basic idea seems simple for those educated in the EE field.
It is the stable oscillator that eludes me.

Any help here is appreciated.
Sincerely;
gmvoeth@yahoo.com



Back to top
Charles Jean
Posted: Fri Jul 29, 2005 3:44 pm
Guest
On Fri, 29 Jul 2005 08:57:02 -0700, "GMV" <gmvoeth@yahoo.com> wrote:

Quote:
I am in need of a circuit that will place minute mark
,second 59 or second zero of UTC, into my homebuilt circuit
for calibration purposes. I have thought of possibly building
a 32768Hz oscillator and using a looping binary counter that can be
calibrated using WWV or WWVB but this all sounds silly to me
since there must be calibration clocks already out there.
I need either one with a TTL output or even better an FET switch or something
like that with a mark I can feed directly into the back end of my
circuit just before the signal hits the A/D converter.
The stability of the oscillator would be most important
not to deviate more than +/-0.03 Hz.
My laptop has a RTC that looses one second of time every four point five hours
so it is completely unsuitable for anything other than getting the time
close enough to understand what the calibration mark means.

Could someone who is an engineer provide me with
a schematic to do this ?
The basic idea seems simple for those educated in the EE field.
It is the stable oscillator that eludes me.

Any help here is appreciated.
Sincerely;
gmvoeth@yahoo.com



Looks like you are after about 1 ppm stability. Dalis/Maxim makes the
DS32kHz chip which is a temperature compensated crystal oscillater.
Don't know if this will work for you, but you might want to take a
look at the data sheet. With a couple of counter chips and this chip
you could get a 1 pulse/min output. The specs say frequency
stablility from 0-40 deg C is +/- 2ppm. Several years ago I built a
clock circuit from this chip, set it to WWV time, and measured its
time vs WWV over a period of 300 days. It gradually and linearly
gained time over WWV, and the total time gained over WWV was 13.7
seconds, or about 0.7 ppm. This was done inside the house, so the
ambient temp over this period probably varied +/- 4-5 deg F. Don't
know if this meets your requirements, since you didn't mention ambient
temp, but thought I'd post this in case you could use it. If you like,
I can send you the circuit and data.

Regards,
Charlie
"Sic hoc legere scis nimium eruditionis habes."
(If you can read this, you're overeducated.)
Back to top
James T. White
Posted: Sat Jul 30, 2005 2:15 am
Guest
"GMV" <gmvoeth@yahoo.com> wrote in message
news:4563$42eae7e7$4f0a87d$10773@DIALUPUSA.NET...
Quote:
Sounds like a good idea but at 4 bits per counter thats a lot of counters to
use.
It would be so nice to find a single chip that would be a 20 bit counter but I
know of none.
I am wary of sending 3.6MHz any distance because
of possible RFI generation.


RFI shouldn't be a problem since the GC-1000 feeds the 3.6 MHz out via a BNC
connector so you would be using coaxial cable to connect it to your counter
module.

You are correct that you would need several stages of dividers if you use
standard TTL or CMOS logic parts. To be more precise, you would need two
divide-by-2 stages, two divide-by-3 stages and five divide-by-ten stages. While
you could do this with a bunch of 4-Bit binary counter stages, I would probably
use some of the devices that provide two decade counters in one package. IIRC,
there are also counters that provide a divide-by-12 counter in a single package.

To me, the better way of doing this would be to use a CPLD programmed to do all
the dividing. I would probably use something like the Xilinx XC9536-15PC44C
($3.30 from DigiKey) which has more than enough macrocells and I/O pins for
something this simple. The Xilinx ISE WebPACK is free for download and can be
used to develop the logic and program the device.

--
James T. White
Back to top
 
Page 1 of 1       All times are GMT - 5 Hours
The time now is Fri Aug 29, 2008 3:26 pm
Science Forum Index

Powered by phpBB & ML | Disclaimer



Contact Us