User talk:JohnM7190/John's Noise Figure Page
Friis being mis-stated?
[edit]I read everywhere, including in multiple places of this article, that the Friis definition of noise factor F is SNR_in/SNR_out. However, looking at the 1944 paper by Friis "Noise Figures of Radio Receivers" I can't find that definition. What Friis is defining in equation (4) is F=(Sg/KTB)/(S/N). But I think what the author wanted to write is F=Sg/(kTB)/(S/N). Anyhow, this is not equivalent to the above definition quoted in many places in literature. The difference is that Friis states that the noise of the input signal must be kTB, whereas the misquoted statement only asks for a SNR_in, which completely leaves open what the input noise is, it just has to be in a ratio with the signal (SNR). Many authors will accompany this with an example of a signal picked up by an antenna, in case the input noise is indeed kTB (defined by the thermal noise floor), and therefore they do not realize the error. But in the general case any particular signal with a given SNR can have an arbitrary noise power. Mayrayday (talk) 23:32, 16 February 2024 (UTC)