Talk:Noise figure

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Physics (Rated Start-class, Low-importance)
WikiProject icon This article is within the scope of WikiProject Physics, a collaborative effort to improve the coverage of Physics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 Low  This article has been rated as Low-importance on the project's importance scale.

Noise figure in attenuators[edit]

"Attenuators have a noise factor F equal to their attenuation ratio L" I just can't see how this can be correct, for example an attenuator made of capacitors (voltage divider) will not introduce any new noise so the SNRin and SNRout will be the same, therfore the noise figure of such an attenuator will be zero, I am sure that there are many other examples where the NF is zero or irrelevant. A resistive attenuator's NF would be a function of the resistors used not the level of attenuation. — Preceding unsigned comment added by (talk) 04:31, 27 March 2014 (UTC)

Satellite communication systems[edit]

The article currently says: the case of satellite communications systems, where the antenna is pointed out into cold space, the antenna effective temperature is often colder than 290 K. In these cases a 2 dB improvement in receiver noise figure will result in more than a 2 dB improvement in the output signal to noise ratio.

From the definition of NF, I don't see how this can be true:

\mathrm{SNR_{out,dB}} = \mathrm{SNR_{in,dB}} - NF

The receiver doesn't know anything about the signal source. What am I missing? GyroMagician (talk) 10:29, 19 March 2010 (UTC)

External Links[edit]

Hello MrOllie. I'm curious - why did you remove the external links to Analog? I'm not saying you were right or wrong, just curious to know your reasoning. GyroMagician (talk) 17:30, 11 August 2011 (UTC)

I was thinning links to out (especially the ones that include fluff pieces on Analog employees and/or products) because I recently noticed that they have been systematically added by a single purpose editor for a fairly long period of time. - MrOllie (talk) 17:47, 11 August 2011 (UTC)
I agree with the removal of the ADC Noise Ref; it doesn't extend the article. I don't know about the log amp one; I can see reasons to keep or delete. I'm not sure a discussion of log amps belongs in this article. Glrx (talk) 03:44, 12 August 2011 (UTC)
Yes, that seems fair. I asked because I met an editor a while ago who was convinced that all company design guides, etc, were advertising (I think it related to an Agilent document describing S-parameters). Anyway, I'm happy to find you're more sensible MrO ;-) The log amp article seems useful, but I'd agree this is the wrong place for it. GyroMagician (talk) 18:35, 15 August 2011 (UTC)
For reference, the log amp link: Glrx (talk) 15:00, 16 August 2011 (UTC)


The equation and description here seem to refer to receivers and electronics. Can anyone add similar information for optical amplifiers? Madgenberyl (talk) 16:15, 11 July 2012 (UTC)

I'm not certain, but as the calculation is entirely in terms of the amplifier input and output SNR (with no attempt to model the noise source), I would imagine it would also apply to optical amplifiers (or mechanical, or any other mode, for that matter). GyroMagician (talk) 17:38, 11 July 2012 (UTC)
Many of the WP noise articles are RF-centric with the side-effect that impedance issues are ignored (many RF systems use a standard impedance such as 50 ohms). More sophisticated models include noise currents and noise voltages and illustrate tradeoffs between noise matching and gain matching. Glrx (talk) 18:31, 11 July 2012 (UTC)
In general I'd agree (and not just on WP) - but as NF is defined in terms of power, Z doesn't make any difference here, does it? GyroMagician (talk) 22:43, 11 July 2012 (UTC)
This article says what NF is, suggests how it might be measured, and how to calculate cascades if I know the stage NF. It doesn't describe how to compute a stage NF given a source impedance and the amplifier noise parameters. It does not suggest design tradeoffs.
Impedance matters. If a stage has a high noise current and a low noise voltage, then a lower impedance source may be attractive. If the noise current contribution inRs >> noise voltage en, then reducing the source impedance by a factor of 4 reduces the in contribution by a factor of 4 while the source's thermal noise voltage declines by factor of 2 (ideal transformer with 2:1 turns ratio gives the 4:1 Z ratio); SNR improves by 6 dB.
But there's another issue. Decreasing the source impedance may increase the stage mismatch loss and wreck the SNR another way.
Setting different bias conditions also impacts noise, gain, and noise figure.
Glrx (talk) 23:57, 11 July 2012 (UTC)
It's true that Z matters for amplifier design. Even in the RF world, low noise amplifiers are typically noise-matched rather than reflection matched (i.e. the input match stage is set to minimize the NF rather than reflection). But this is all about amplifier design, not NF itself. Maybe we do need a section describing how to calculate NF under different Z, talking about in and en, but I can't help feeling this would belong more on the low noise amplifier page than here (and that page could really do with some work).
I would have pointed Madgenberyl to the LNA article if it were helpful; that's why I made the many of the WP noise articles comment. The topic should be covered somewhere; it can have a short summary and link in other articles. I'm in basic agreement, but NF is both an amplifier and a systems issue. Glrx (talk) 14:06, 12 July 2012 (UTC)