[BC] The first 3 dB...

[Lamar Owen]

On Saturday 18 February 2006 16:53, DANA PUOPOLO wrote:
> The 3 db would come from improving the noise level of the tuner.
> A GOOD UHF tuner might have a noise figure of 9-12 db. With a cheap GasFet,
> this could easily be improved to below 2 db. Remember, we're only talking
> 700 mHz here.

And here I'm used to seeing NF's in the 0.3dB range...and those are noisy by 
radio astronomy standards (as these are room temperature LNA's; cryo can get 
your NF down further; cool that LNA to 4K and NF's go through the floor; we 
have a lab coldfinger good to 10K).  Similar frequency; a good 1.4GHz LNA 
will need an NF well below 1dB to do much good.  A source for this kind of NF 
LNA's is Radio Astronomy Supplies (http://www.nitehawk.com/rasmit/jml0.html 
for the LNA's and other receivers).  

If they were to be mass-produced, these LNA's could easily be less than the 
price you see quoted on RAS's page (typical .5dB NF LNA around $150-$200).  
The 610MHz LNA would be closest to the UHF band; 18dB gain with a 0.5dB NF 
for $150; 406MHz with 18dB gain and 0.33~0.37 dB NF same price.  RAS 
hand-builds these at these prices.

Of course, all you really need is a good LNA in front of a so-so tuner to get 
Tsys down.

However, noise figure is not the most important spec to quote; third order 
intercept is also critical for low signal work.  We like to spec front ends 
with at least 20dBm at minimum, with 40dBm better (a 26 meter diameter 
parabolic has loads of gain at UHF and above) and our RFI survey LNA is 
running an NF of 1dB and intercept at 60dBm.  Requires a substantial 
heatsink.
-- 
Lamar Owen
Director of Information Technology
Pisgah Astronomical Research Institute
1 PARI Drive
Rosman, NC  28772
(828)862-5554
www.pari.edu