Re: [Discuss-gnuradio] PNG's of a neutral hydrogen observation at PARI with a USRP.

[Marcus Leech]

Lamar Owen wrote:

> The one on 26 West is a RAS LNA and feed; NF has been measured at 0.38dB.  Our 
> standard of comparison for 70MHz spectroscopy is a RAS SpectraCyber 
> custom-built for us by RAS.  We also have a rack mounted 1420-70MHz 
> downconverter in the control room, but find it better to downconvert in the 
> feedbox to 70MHz; there's 2,000 feet of 7/8's heliax in the line to contend 
> with, and even 7/8's loss at 1420 isn't great.  At 70 it's negligible.
In all of those situations, the noise-figure of the back-end receiving 
system is largely
 irrelevant, since whatever noise-figure you have in the back end is 
overwhelmed by
 the gain of the front end.  Front-end noise figure is the absolute 
critical parameter
 in determining system noise temperature.

But other factors in the back-end, like short and long-term stability 
are important.
 Noise figure isn't really one of them. For example, a gain change in 
the back end,
 due to temperature drift, etc, of only 0.05dB can completely swamp 
your radiometer-mode
 observations, particularly if such changes are occuring at similar 
periods to the signals
 you're looking for.  Techniques like beamswitching, etc, are 
traditional in dealing with
 these issues.  But having a very stable, temperature-controlled, 
back-end helps
 tremendously :-)


--
Marcus Leech                            Mail:   Dept 1A12, M/S: 04352P16
Security Standards Advisor        Phone: (ESN) 393-9145  +1 613 763 9145
Advanced Technology Research
Nortel Networks                          address@hidden