[TowerTalk] Coax cable economics

[Jim Lux]

On 2/19/15 9:09 AM, James Wolf wrote:
> Patrick,
>
> The problem with your setup is that when you short across the input to the
> receiver you can easily change the characteristics of the input circuitry
> and the first RF amplifier.
> Since the receiver was designed for ~50 ohms input, and you want to measure
> the radio's output noise level under a no-signal condition, you will want to
> terminate the input with 50 ohms.
>

It's not entirely clear that a typical ham HF receiver has a 50 ohm 
impedance OR that this is where the noise is lowest.

Typically, HF receivers design isn't driven by noise figure (strong 
signal handling, and instantaneous dynamic range are probably more 
important for a lot of applications).

So if your receiver had a 100 ohm input impedance, and you reflect back 
some of the power from your antenna, you're also reflecting back the 
atmospheric noise, so the "received SNR" stays the same.

And most amplifiers have input noise that varies with the impedance it 
sees on the input.  Microwave LNA designers obsess about this kind of 
thing, trading off loss in the matching network against improved NF of 
the amplifier.   There's even amplifiers with negative noise figure 
(narrow band, not necessarily stable, with carefully designed feedback 
that cancels the input noise).  And, at microwave frequencies, external 
noise sources (particularly looking up at the sky) are quite small, 
compared to thermal noise from the environment.


take home message is that expecting a HF receiver to look like a perfect 
load might be unrealistic, since it doesn't affect the "as used" 
performance very much.