[hpsdr] mercury adc

jeff millar jeff at wa1hco.net
Sun Oct 19 11:57:15 PDT 2008


alex brown wrote:
> ***** High Performance Software Defined Radio Discussion List *****
>
> how do you get 119 db dynamic range of a 16 bit adc surely it should be 96 db
>
> alex
> m0gjr
> ps this is the 5th time ive sent this message with no response so if you receve it could you reply anyway even if you dont know the answer
>   
Let me add another way to think of it. 16 bits of of ADC corresponds
to 96 dB SNR in the _Nyquist band_, from 0 Hz to 1/2 the sampling 
frequency.
This random quantization noise, which comes from not knowing more than
16 bits of the true RF voltage gets spread evenly across the Nyquist band.

For a 100 MHz sampling frequency, quantization noise gets spread across
50 MHz of bandwidth. An SSB signal of 2.5 kHz bandwidth has only 1/20,000
of that noise, fully 43 dB less noise...corresponding to a 43 dB 
increase in
dynamic range (if that was the only factor).

So the dynamic range depends on _two_ things, the sampling rate and
the intended receive bandwidth.

A side effect of this relationship mean that one can increase the 
dynamic range
of a system by using a higher sampling clock which spreads the 
quantization noise
across a larger bandwidth.

jeff, wa1hco




More information about the Hpsdr mailing list