[TowerTalk] Re; Need Help Remembering Something I Read

[jimlux]

On 7/7/18 12:41 PM, Richard (Rick) Karlquist wrote:
> Interesting paper, lots of stuff to absorb.
> The discussion of receiver noise figure brings
> up a topic that has bothered me for a long time:
> the use of excessively high gain external antenna preamps
> (as high as 40 dB).  Somehow, this is supposed to
> make up for a really low gain receiving antenna,
> say -40 dBi.  Most practical external preamps have
> an NF of around 3 dB.  With a gain of 10 dB, they
> should be able to overcome receiver noise in most
> cases, especially of the receiver's internal preamp
> is selected.  If high gain is used, you are simply
> listening to amplifier noise, instead of band
> noise.
> 

with beaucoup gain at the antenna, you can use really lossy cable back 
to the receiving point. Most "active GPS" antennas (which is virtually 
all of them) have about 30-40dB gain, so there's little point in 
running, say, LMR-400 (other than for shielding effectiveness).

The issue with excessive gain is on the top end - is environmental noise 
going to saturate the final stage.  40dB seems high, unless you've got 
some selectivity in the front end.

I've got a HF receiver in LEO feeding a 1V ADC and it has about 30 dB 
gain from a 6m long dipole. We went back and forth on the gain, but 
settled on "enough gain to get the LNA noise just above the ADC noise", 
and have a switchable 20dB pad in the front end.

The noise figure of the amplifier should be low enough that 
environmental noise is above the amplifier noise.  For HF, if the 
ionosphere is transparent, your dominant noise is galactic background 
which Hilary Cane measured quite accurately Plus local thunderstorm 
noise. If the ionosphere is reflective it's manmade + thunderstorm noise.

The ITU-R P.372-8 report (which you can download from itu.org)  gives 
the curves, but for general purposes, the noise figure at 10 MHz is 
about 30dB. That is, an isotropic antenna will have a noise that is 30dB 
above kTB (-174 dBm/Hz).  Atmospheric noise will raise that to 40dB, and 
"median business area man made" might be around 50dB.

At this kind of level, NF of the amplifer is probably irrelevant.

With a "short antenna", the situation changes a bit.. at 10MHz a full 
size dipole is 5m long, so with a 1m whip, you're down about 14 dB, and 
it gets worse with lower frequencies..  At 1 MHz, even though the 
environment might be 50-100 dB above the thermal noise floor (into an 
isotrope), with a short antenna, you might have 30-40 dB of "mismatch loss"