[Moved]Definition of noise figure

[Gzmeuh]

Hello.

Let's imagine that you have a signal with it's associated noise of at least kTB is amplified (by 8 for example). The signal to noise ratio is still the same (OK, I don't take into account the noise of the amplifier) but your amount of noise is now higher than kTB. If you go through the attenuator after that (by 2 for example), the noise of the attenuator itself will be smaller the the attenuated noise from the input and NF should be close to 0.
Where am I wrong ?
I'm sure that the NF of a passive is equal to its attenuation but I agree with dd2001 that it appears most of the time like an axiom.

Thank you,

Gzmeuh.

 

[jiripolivka]

Hello.

Let's imagine that you have a signal with it's associated noise of at least kTB is amplified (by 8 for example). The signal to noise ratio is still the same (OK, I don't take into account the noise of the amplifier) but your amount of noise is now higher than kTB. If you go through the attenuator after that (by 2 for example), the noise of the attenuator itself will be smaller the the attenuated noise from the input and NF should be close to 0.
Where am I wrong ?
I'm sure that the NF of a passive is equal to its attenuation but I agree with dd2001 that it appears most of the time like an axiom.

Thank you,

Gzmeuh.

Please refer to any good textbook.Noise figure has nothing to do with signals. It is an expression referring device noise contribution to a matched resistor noise contribution.
Attenuators generate noise like resistors, their contribution depends on their loss and physical temperature. Their maximum loss generates the same noise as the matched resistor on the line.

 

[Gzmeuh]

Hi, thank you for your reply.

As you say, noise figue for an attanuator will depends on temperature and matching. So the relation between attenuation and NF should include some 'T' and some 'S22' terms.
Maybe an advice for a good textbook or a nice link ?

Thank you,

Gzmeuh.

 

[BigBoss]

Hi, thank you for your reply.

As you say, noise figue for an attanuator will depends on temperature and matching. So the relation between attenuation and NF should include some 'T' and some 'S22' terms.
Maybe an advice for a good textbook or a nice link ?

Thank you,

Gzmeuh.

The Noise Figure of an Attenuator is exactly Attenuation Value of the Attenuator.

So, ATT=10dB,NF=10dB..

 

[jiripolivka]

A good textbook is Van der Ziel: Noise, from ~1980. Certainly there is a lot of newer books, too.
Attenuator noise power is usually defined as noise temperature, taken from the basic formula Pn = kTB. So, T =Pn/(kB)
S22 is a measure of impedance matching to a transmission line.

A noise temperature from a matched resistor is equal to its physical temperature. For an attenuator having a loss L (times, not dB),we have

T= To (1-1/L), where To is the ambient temperature (Kelvins).

For a 3-dB attenuator, L=2, so the output temperature T = 1/2 To.

- - - Updated - - -

And for L>20, T ~ To