> There are 3 key issues involved:
> 1. IMDDR3 (two-tone dynamic range)
> 2. BDR (single tone dynamic range)
> 3. Phase noise (which can override both)
All a very good explanation Bill, but there is a fourth,
fifth, and sixth issue. Also I think it is fair to mention
people in a position to design radios often do not listen to
external input and design for themselves and what they think
others want. They then get ruffled and want to force
end-users to agree that whatever they did is perfect and so
the radio must always work perfectly in any situation and no
one is allowed to point out flaws!
So adding to the good list you made.....
4.) AGC can affect readability of weak signals near noise.
Humans do NOT want a perfect AGC when copying weak signals
near noise. They do not want a perfect AGC in pileups
either. Designing a receiver with no volume change for
signal level change is a mistake for HF DXing or contesting
5.) Selectivity. There has to be a good compromise between
skirt rolloff and "ringing".
6.) User interface. If you can't easily use it for what you
are doing or understand it, then the best performance is
Then there is another issue on top of all that. A two tone
test is great in a radio that only passes a narrow slice of
signals, but when a radio passes a very wide swath the
testing must also include all the expected multiple signals.
While looking at close spaced and wide spaced testing with
two tones works with a radio that only passes a narrow
window to sensitive stages, it does not work when the radio
passes everything within 50kHz on a very crowded band. This
is because the accumulated power of all the signals in a 50
kHz window is significantly higher than the power of all the
signals within only a few hundred Hz area of the band.
Suppose we have two radios, one with 100 dB close spaced and
wide spaced spec, and another with 80 dB close spaced and
140dB wide spaced specs. When the band is nearly empty the
100dB DR spec radio will probably always be better. When the
band is congested, the 80 dB close spaced spec radio might
actually be better with close spaced signals than the 100dB
radio!!! This is because the 100dB radio has to handle ALL
the signals with that spec, both close and far in frequency.
The 80dB radio might only have to handle three or four
strong signals, and the rest fall into the -140dB wide
spaced area and can be ignored.
Engineers run into this problem designing broadband
amplifiers. It's a whole lot easier to design an amplifier
that only handles two carriers than twenty. Receivers are no
different, and we can't force it to be otherwise through
Topband mailing list