does noise attenuate like signal through a transmission line

[mona123]

Hi, If I connect a long lossy transmission line to the output of an amplifier and measure the noise output power, will my measured noise power be shaped by frequency response of transmission line? Thanks.

 

[johnjoe]

Sure but it has a low pass filter characteristic. But if you have a high gain amplifier the influence is quite small.

Forget the last sentence, we're talking about noise power and not noise figure, my fault!

 

[mona123]

Sure but it has a low pass filter characteristic. But if you have a high gain amplifier the influence is quite small.

Thanks John. So if I measure lets say -100 dBm output noise power @1GHz at the output of the die and I have cable loss connected at the output of die of 3dB @1GHZ. Then the output noise power measured using cable would be -103 dBm?

One more confusion related to this is that why do we have noise figure of a passive device equal to loss of passive device, if the signal and noise both will be attenuated by same amount? won't the signal to noise ration at the input will be equal to signal to noise ratio at the output, hence noise figure unchanged. I am sure I am missing something big here?

Thanks

 

[johnjoe]

I uploaded some pdf for you, it should help to understand the relation between noise figure and attenuation. It is the same for a cable (passive device), input noise is attenuated but "new" noise is "generated" inside the cable.

 

[mona123]

Thanks. So does it mean its opposite. I will have 3dB more noise power that means I should measure -97dBm in the earlier example I gave? I am confused...I want to understand only in terms of noise power not noise figure..

 

[RealAEL]

In simple terms if Noise is calculated using:

Noise Power Density = kTB + NF + Gain (dBm/Hz)

Then for your case with the amplifier and transmission line:

Noise Power Density = kTB + NF(amp) + Gain(amp) + NF(line) + Gain(line)

For any passive device, like the transmission line, the Noise Figure is equal to the loss in the line so if we take the example of a 3dB loss then

Noise Power Density = kTB + NF(amp) + Gain(amp) + 3dB - 3dB

So the two terms for the transmission line 'cancel out' and the noise power density is the same at the output of the transmission line as it is from the amplifier. Given that any frequency response of the passive line is simply a change in attenuation at different frequencies then the noise figure would surely change in exactly the same manner and the two will always cancel out at all frequencies. So any 'frequency response' for the noise power density at the output of the transmission line would simply be a repeat of that at the output of the amplifier, if any, and would not be altered by the transmission line.

 

[kspalla]

Thanks John. So if I measure lets say -100 dBm output noise power @1GHz at the output of the die and I have cable loss connected at the output of die of 3dB @1GHZ. Then the output noise power measured using cable would be -103 dBm?

One more confusion related to this is that why do we have noise figure of a passive device equal to loss of passive device, if the signal and noise both will be attenuated by same amount? won't the signal to noise ration at the input will be equal to signal to noise ratio at the output, hence noise figure unchanged. I am sure I am missing something big here?

Thanks

You are right. no confusion. Remember F1+F2-1/G1 formula.

- - - Updated - - -

Thanks. So does it mean its opposite. I will have 3dB more noise power that means I should measure -97dBm in the earlier example I gave? I am confused...I want to understand only in terms of noise power not noise figure..

In my view, if you work in terms of noise power than NF , you learn a lot.
in your example -100dBm is the noise power at amplifier output, loss is 3dB then result of noise power is -103m as noise also gets attenuated by same loss.
NF calculation on higher noise powers is easy by just substantiating( remember [F-1)/G] due to high G,
Lets take example where G is small. let say Noise power is -164dBm and is passing through same 3dB transmission loss. what is the effective noise power at this stage? it is not -167dB am sure.
if you find the answer means you understood noise power and hence noise figure calculations.
Hope this trigger your thoughts.

Regards.

 

[RCinFLA]

"does noise attenuate like signal through a transmission line "

Depends on signal and noise power level but general answer is no. At low input noise levels the physical temperature of the losses (cable in your question) will add more noise then signal and input noise attenuation. This is most important on space communications systems.

 

[kspalla]

"does noise attenuate like signal through a transmission line "

Why not? we are talking about large noise powers generated by previous stages. Treat noise as the signal to the input for next stage that amplifies or attenuated by the next stage.

 

[johnjoe]

Lets take example where G is small. let say Noise power is -164dBm and is passing through same 3dB transmission loss. what is the effective noise power at this stage? it is not -167dB am sure.
if you find the answer means you understood noise power and hence noise figure calculations.
Hope this trigger your thoughts.

It is more obvious, if you take even smaller values for input noise power ;-).