[time-nuts] Why a 10MHz sinewave output?

John Miles jmiles at pop.net
Mon Feb 6 21:43:02 UTC 2012



> -----Original Message-----
> From: time-nuts-bounces at febo.com [mailto:time-nuts-
> bounces at febo.com] On Behalf Of bob grant
> Sent: Monday, February 06, 2012 12:41 PM
> To: Discussion of precise time and frequency measurement
> Subject: Re: [time-nuts] Why a 10MHz sinewave output?
> 
> 
> Some sine-wave outputs are not very symmetrical, in that the rising
> edges
> are much more sinusoidal in shape than the falling edges.
> 
> I guess my question is really about what type of input circuity and
> drive
> level are most common and which signal shape would provide the lowest
> jitter.
> 

Most line receiver chips are fine for squaring up sine-wave signals that
come in on a 50-ohm connection at +5 to +15 dBm.  If you need fast 10 MHz
edges, they are pretty easy to generate on demand without adding significant
jitter.  A two-transistor differential amp works well enough in almost every
case.

I think the answer to the "why sines?" question is a mixture of several
points others have raised.  1, 5, 10, and now 100 MHz sine waves have all
been used for reference-frequency distribution.   Equipment that needs to be
locked to a common reference does one of two things with that reference
signal as soon as it enters the enclosure: they either fan it out for use
where it's needed internally, or they phase lock their own internal
oscillator to it.  Before digital hardware became so prevalent, these
applications -- either miscellaneous RF circuits or analog PLLs -- both
tended to expect sine wave signals.  

Also, the signal would have originated in a crystal oscillator or similar
source, where it would naturally be available as a sine.  So it would have
taken extra circuitry to square it up, potentially extra circuitry to filter
it at the destination, and better cabling to transmit the harmonic-rich
signal.  Differential signaling was not yet in common use, and not entirely
free to implement... and EMI is always a concern even with 10 MHz sines.  If
you don't use good double-shielded cabling for 10 MHz distribution, your lab
environment will be full of 10 MHz energy from your house clock that is
difficult to keep out of sensitive circuits.  Things would be even worse if
we were at risk of 'broadcasting' harmonics from 10 MHz to daylight.

These days you might do things differently if starting from scratch, but
there is so much infrastructure designed to generate, carry, and use  5/10
MHz fundamentals that these have become an entrenched standard.  If there's
a trend away from 5/10 MHz, it seems to be towards 100 MHz fundamental
distribution.

-- john





More information about the time-nuts mailing list