[time-nuts] Fast frequency counting question
bruce.griffiths at xtra.co.nz
Mon May 5 18:13:35 EDT 2008
Peter Vince wrote:
> Forgive me if this is nonsense, but what about mixing with a pair of
> quadrature signals from another continuously running 40 MHz signal,
> and using the low-pass filtered outputs to directly show the relative
> phase, and then grab that on a sampling 'scope?
> Peter Vince (G8ZZR, London, England)
In principle it would appear to work if the mixer output low pass filter
had adequate bandwidth.
However in practice, to achieve picosecond phase resolution the ADC
would need to have an effective resolution of at least 15 bits.
Few digital sampling scopes achieve this resolution.
Since the transfer functions of the mixers wont be known to 15 bit
precision they would need to be calibrated.
An equivalent performance is achieved by sampling the 40MHz signal
directly using an ADC that has an ENOB =15 with a 40MHz input signal.
More information about the time-nuts