[time-nuts] tcxo

Jim Harman j99harman at gmail.com
Thu Dec 11 11:57:56 EST 2014


More Arduino clock and timekeeping notes:

If the part connected to the processor's XTAL1 and XTAL2 pins has 2 pins
and external capacitors it is a crystal. If it has 3 pins and no caps it is
a ceramic resonator. If you have a crystal, you can fine-tune its frequency
by replacing one of the capacitors with a variable cap.

You may have trouble using an external oscillator instead of whatever is
there because the chip is supposed to be set up differently when you do
that. See Sec. 9 of the ATMega datasheet for further info.

Another and potentially more flexible approach (again depending on your
requirements) would be to leave the processor clock alone and connect an
accurate external oscillator to a timer input. Then you can use timer
interrupts and/or the timer capture function for precise timekeeping.

On Thu, Dec 11, 2014 at 9:31 AM, folkert <folkert at vanheusden.com> wrote:

> > Some of the Arduinos (not sure about Mini 04 but I am suspicious) use
> > ceramic resonators rather than real crystals and thus may have extremely
> > poor frequency stability. See here
> > http://jorisvr.nl/arduino_frequency.html
> > for an example.
>
> ah!
> Very intriguing material, those crystals. I wonder what happens in them
> that they start to tick(?) at precisely 16MHz for example.
>
> > Is there some reason you are using a 16.9344 MHz oscillator rather than
> > 16.0? The processor will probably work, but timing of Arduino functions
> > like millis() and software serial baud rates will be affected.
>
> Misread I guess :-|
> Like that time that I bought 20 max690 instead of max680 ics. sigh.
>
> > On Wed, Dec 10, 2014 at 2:19 PM, folkert <folkert at vanheusden.com> wrote:
> >
> > > > I'm experimenting a bit with time keeping.
> > > > For that I use cheap low power hardware like raspberry pies and
> > > > arduinos.
> > >
> > > Hi people,
> > >
> > > Thanks for all the replies! Took a bit to respond but I had the flu.
> > >
> > > My objective is, to get the best precision/accuracy possible with said
> > > hardware.
> > >
> > > The first step is preparing an arduino. I found one that has an easy to
> > > desolder crystal:
> > >
> > >
> http://upload.wikimedia.org/wikipedia/commons/thumb/0/04/Arduino_Mini.jpg/170px-Arduino_Mini.jpg
> > > Bought a couple of those and some of these:
> > >
> > >
> http://www.aliexpress.com/item/Precision-TCXO-square-wave-oscillator-TCXO-16-9344MHz-line-rectangle/1104200137.html
> > >
> > > While waiting for these and the parts for the schematics, I'm toying
> > > around with implementing ntp servers (well, sntp) for arduino. With
> > > lots of blinkenlights of course and with ethernet. Not trivial I can
> > > say.
> > > This is the jitter plot of a test where time is fed from an RTC module.
> > > I did not plot the drift but it is +/- 120 seconds in the 3.8 days I
> > > measured it.
> > > http://vps001.vanheusden.com/~folkert/aRTC.png
> > >
> > >
> > > Folkert van Heusden
> > >
> > > --
> > > ----------------------------------------------------------------------
> > > Phone: +31-6-41278122, PGP-key: 1F28D8AE, www.vanheusden.com
> > > _______________________________________________
> > > time-nuts mailing list -- time-nuts at febo.com
> > > To unsubscribe, go to
> > > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> > > and follow the instructions there.
> > >
> >
> >
> >
> > --
> >
> > --Jim Harman
> > _______________________________________________
> > time-nuts mailing list -- time-nuts at febo.com
> > To unsubscribe, go to
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> > and follow the instructions there.
>
>
> Folkert van Heusden
>
> --
> Multitail est un outil permettant la visualisation de fichiers de
> journalisation et/ou le suivi de l'exécution de commandes. Filtrage,
> mise en couleur de mot-clé, fusions, visualisation de différences
> (diff-view), etc.  http://www.vanheusden.com/multitail/
> ----------------------------------------------------------------------
> Phone: +31-6-41278122, PGP-key: 1F28D8AE, www.vanheusden.com
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>



-- 

--Jim Harman


More information about the time-nuts mailing list