[hpsdr] About the Dynamic Range of Receivers

[Robert McGwier]

Here is an intuitive guide to help your thinking.  No math.

You cannot increase the dynamic range at the top by decimation.  Nothing 
helps you if you clip the A/D.  What you are hoping for is to increase 
the dynamic range DOWN.

Assume you have a large waveform in the stop band of the decimation 
filter  and a weak signal in the passband of the decimation filter but 
the strong one is not clipping the A/D.  Furthermore for simplicity, 
assume the strong one in the stop band is a tone.   In both the real and 
imaginary channels of this strong tone there are places where it crosses 
through 0 volts.  You are sampling larger than the "Nyquist rate" for 
the weak signal you are after,  so when the large signal is out of the 
way (down near zero in either channel),  the weaker signal, is more 
likely to wiggle a bit or two in the digital representation.  Now you 
filter off the strong signal and the bit or two of wiggling is 
interpolated by the filtering process.   All of that noise that was 
outside of the decimation filter passband that was helping to mask the 
impact of the wiggling is filtered off by the decimation filtering so 
the wiggling becomes more visible because so much of the noise power is 
now gone.

There is quite a bit of knoodling to get the dB enhancement (and 
assumptions you hope are true) to get the actual processing gain..  You 
can "see" this happening instantly with the SDR-1000 and the console 
software (for example) with a weak signal you can see.    As you 
decrease the filter bandwidth the noise power inside the filter (and 
therefore the minimum discernible signal) drops in proportion to the 
width of the filter.

Does this help at all? 

Bob
N4HY



Greg Overkamp wrote:
> ***** High Performance Software Defined Radio Discussion List *****
>
> I have a question regarding this explanation of
> instantaneous dynamic range. After decimation, the
> instantaneous dynamic range is measured only over the
> bandwidth of the output of the decimation filter. What
> if the weak signal is within the passband of the
> decimation filter, while the strong interfering signal
> is in the stopband of the decimation filter, but below
> fs/2 of the A/D converter. Can the dynamic range still
> be considered to be improved by decimation? In this
> example, can the dynamic range still only be measured
> over the bandwidth of the output of the decimation
> filter? If this same strong signal necessitates a
> reduction in gain ahead of the A/D in order to keep
> the A/D from saturating, can the instantaneous dynamic
> range still be considered to be increased by the
> decimation process and only measured over the BW of
> the decimation filter passband?
>
> This has been bugging me for awhile. Any help on
> clearing this up is appreciated.
>
> Greg Overkamp
> WD9DEX
>
>
>
>  
>   
>> About the Dynamic Range of  Receivers 
>>
>> It has nothing to do with the instantaneous dynamic 
>> range, which is defining 
>> the radio's capabilty to discern a weak signal (near
>>  the noise floor) in the 
>> presence of a much larger signal. The maximum
>> possible  ratio of the two 
>> signal's levels is called instantaneous dynamic 
>> range). 
>> The overall instantaneous dynamic range of the
>> receiver  is not limited by 
>> the nyquist dynamic range of the 16-Bit ADC (about
>> 95 dB). DSP  is decimating 
>> the ADC sampling rate from 100 kS/sec down to rates
>> appropriate to  the actual 
>> CW / SSB signal bandwidths of 300 / 3000 Hz (meaning
>>  Fs= 781 / 12.5 kS/sec). 
>> Noise coming from ADC (including the  quantization
>> noise) is reduced by just 
>> the ratio of ADC's nyquist bandwidth (=50  kHz) to
>> the signals bandwidth, thus 
>> giving a corresponding gain in dynamic range  of  
>>     
>
> _______________________________________________
> HPSDR Discussion List
> To post msg: hpsdr at hpsdr.org
> Subscription help: http://lists.hpsdr.org/listinfo.cgi/hpsdr-hpsdr.org
> HPSDR web page: http://hpsdr.org
> Archives: http://lists.hpsdr.org/pipermail/hpsdr-hpsdr.org/
>
>   


-- 
AMSAT VP Engineering. Member: ARRL, AMSAT-DL, TAPR, Packrats,
NJQRP/AMQRP, QRP ARCI, QCWA, FRC. ARRL SDR Wrk Grp Chairman
"An expert is a man who has made all the mistakes which can be
made in a very narrow field."  Niels Bohr


 1154692590.0