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

[Jim Lux]

On 2/6/12 6:47 AM, paul swed wrote:
> Well right you are thats why todays chips have equalizers and such.
> But then its all getting crazy complicated even though its in a itty bitty
> chip.
> My distribution is made of high quality television analog amps and I have
> in general made amplifiers and such with parts I can still easily pickup
> and solder to.
> But still I always do wonder about tinkering with a square wave dist
> system. Though I doubt I will ever actually do anything.
> KISS is the general principal.
> Regards
> Paul.
>


"adaptive equalizer" and "precision frequency/time distribution" are 
going to be very uneasy bedfellows..

Of course, if you're just looking for distribution of house black burst 
or analog video, that's probably ok.  You're looking for good waveform 
fidelity, rather than precise knowledge of time delays.

In most of the precision measurement systems I fool with we look for 
parts in 1E10 or better.  Say, 1 degree of phase at 32 GHz.. if you're 
multiplying up from a 10 MHz reference for that, the x3200 
multiplication means you need to be pretty savvy about how your 
references are distributed (and, as well, how a phase change in the 
harmonic content might screw up the zero crossings)