[time-nuts] Rb Oscillator - rather fundamental question
dforbes at dakotacom.net
Tue Feb 23 22:18:29 UTC 2010
Matthew Smith wrote:
> Quoth Poul-Henning Kamp at 2010-02-23 20:22...
>> In message <4B83A33C.1010903 at smiffytech.com>, Matthew Smith writes:
>>> Simple and rather fundamental question: does the common or garden
>>> rubidium oscillator constitute an atomic clock?
> Many thanks for the responses and ensuing discussion that has
> considerably value-added to the yes/no nature of my original question ;-)
> Now I know a lot more about primary/secondary standards than I did a
> (9,192,631,770 Cs wobbles * 86400) ago.
> I can now proceed with my unconventional calendar design (a cascade of
> dekatrons) knowing that it will be driven by an atomic ticker.
> BTW: does anyone know if a 0.55V p-t-p sine wave from an Rb source would
> be enough to clock an Atmel AVR microcontroller? The crystal/clock
> input *is* an amplifier, but didn't know if I'd need to do anything to
> the signal first, to get it closer to the 5V logic level.
The signal you describe could be fed through a small-value capacitor (perhaps
100 pF) right into the microcontroller. The amplifier must be internally biased,
otherwise it wouldn't work with a crystal.
You will want to put a protection circuit on the input if you have an exposed
coax connector leading to the outside world, in case of a static discharge to
the center pin of the cable. A series resistor of several Kohm followed by the
capacitor in series with the signal, plus a pair of 1N4148 diodes connected in
reverse-bias to Gnd and Vcc from the input pin of the MCU, are usually sufficient.
More information about the time-nuts