[Elecraft] Update to KSYN3A FAQ (some measured dynamic range numbers)

[Ian White]

>> If both RX and TX are using the same synthesizer, TX phase noise
>> should be lowered.
>
>Not necessarily -- it depends on the design. Some rigs have low phase
>noise on RX but not on TX.
>

The noise that we hear on transmitted signals, or through reciprocal
mixing in our receivers, is always the COMPOSITE noise - the vector sum
of both the Phase Noise and the Amplitude-Modulated noise. 

Most modern test equipment measures exclusively phase noise, so that's
what the equipment reviews quote as well. Then the advertisers jumped on
the bandwagon so now it's "Phase noise, phase noise, everybody's talkin'
'bout phase noise!" We are carelessly sliding towards labeling every
kind of noise as "phase noise" when very often it isn't. 

Don't ever forget that AM noise component. At some frequencies AM noise
can be even more important than the phase noise.  

If the transmit and receive paths use the same frequency-determining
signals (oscillators) then the phase noise - true phase noise, that is -
in the TX and RX paths should be very similar. Generically they should
be the same, so you'd need to identify some quite specific reasons for
PN to be different between TX and RX in any particular transceiver.

But a transceiver also contains many, many sources of AM noise - not so
much in the oscillators but in the TX and RX signal pathways. Because
those pathways are different, the levels of AM noise generically *do*
differ between TX and RX. 

Therefore we should always expect the *composite* noise to be different
between TX and RX - and that will probably be mostly because of
different AM noise levels.

Problems with AM noise on transmit are often made worse by designers
forgetting that the problem even exists. In older analog rigs, for
example, there may be significant AM noise on transmit because the TX
chain starts out at too low a signal level, and the designer forgot to
use low-noise techniques in that weak spot. 

In modern digital rigs, the analog TX signal comes from a DAC (Digital
to Analog Converter) which is heavily laden with AM digital noise that
needs careful filtering. There is an awful example in SM5BSZ's 2013
article on 'Band Pollution by Amateur Transmitters'
<http://sm5bsz.com/dynrange/dubus313.pdf>.   Figure 8 shows a very high
noise level at the output of the DAC (enough to make the sine-wave look
like a fuzzy caterpillar). Then that noise is insufficiently filtered,
leaving serious TX noise sidebands of only -75dBc at +/- 300kHz. Because
this is almost entirely AM noise, any test that focuses exclusively on
phase noise will miss it... and that particular manufacturer did miss
it.

In contrast, the K3 came top of the table for suppression of wideband TX
noise (Table 1, which pre-dates the KX3).
 

73 from Ian GM3SEK