Rick Westerman wrote:
> Hey Sinisa, what in the world are you trying to prove now?
> "Can I recommend a more repeatable noise figure test,
> which even doesn't require any signal on the band?"
> What is your point? Who among us is interested in tests which do
> not require a signal on the band. I assume we are all in this
> game to hear signals on the band. What you are suggesting is
> nothing more than mental masturbation. Sorry for the French but
> that's the only words I can find for this situation!
> Sorry Sinisa, this time I just can't see your point!
Dear Rick, there is a very good technical reason for the test.
Far from being a [French word], the test checks a vital condition
which influences receiver's ability to receive weak signals,
when the main obstacle is noise, and not interference from other signals.
The total noise power audible from the speaker is a sum of
* receiver's internal noise, plus
* external noise (man made, atmospheric, galactic...).
In a given circumstances (location, antennas, propagation conditions)
we have no control of the external noise. If there was no
receiver noise, then the only limiting factor in weak signal
reception would be the external noise, i.e. the ultimate signal
to noise ratio is determined by the ratio of the external noise
to the signal. No receiver can improve on that.
Real receivers add noise of their own, which must be low enough
in order not to degrade the signal to noise ratio available
from the antenna.
For simplicity, let's walk through an example.
external noise power = 1 unit
signal power = 10 units
--------------------------------------
signal to noise ratio = 10 : 1 (10 dB) as available from antenna
receiver noise power = 7 units (8.5 dB above external noise)
total noise power = 8 units (1 + 7)
----------------------------------------
signal to noise ratio = 10 : 8 (1 dB) as audible from speaker
In this example signal to noise ratio was degraded by 9 dB
due to receiver's noise far in excess of external noise.
Please note that the amount of degradation depends ONLY
on the ratio of receiver's noise to external noise,
and NOT on the signal level.
This explains both why is this test important
and how it can work without a signal.
A good receiver for the application would have a noise which
is ~12 dB (2 S units) below external noise:
receiver noise power = 0.063 units (12 dB below external noise)
total noise power = 1.063 units (1 + 0.063)
----------------------------------------
signal to noise ratio = 10 : 1.063 (9.7 dB) as audible from speaker
This receiver degrades signal to noise ratio by only 0.3 dB.
If a perfect receiver was available, it could not improve on it
by more than 0.3 dB, which is insignificant for most purposes.
To summarize:
a) receiver's noise must be weaker than external noise
if the receiver is not to degrade the signal to noise ratio;
b) there is no point in making receiver's noise weaker
than external noise by more than 10-12 dB; nothing
can be gained, and receiver may be overloaded if
too much gain is used (low receiver noise is usually
accompanied with more gain);
c) if the reception is limited by receiver's noise
(and not by interference, intermodulation, etc, etc),
then there is a limit on how much improvement is possible,
as described above. Merely having more gain helps
absolutely nothing. Having a less noisy receiver
helps only to a point when receiver's noise
becomes ~12 dB below external noise. Further lowering
of receiver noise is useless.
d) optimum receiver settings obviously depend on circumstances
such as frequency, location, antennas and propagation conditions.
I hope this explanation helped.
73,
Sinisa YT1NT, VA3TTN
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|