Noise figure calculation

[mw_rookie]

Given the values of insertion loss and return loss (X and Y dB resp.), how to obtain the value of the noise figure.

Thanks.

 

[DDavid]

If X is IL than NF will be NF=X plus what u have after that until will have active device

 

[mw_rookie]

Thanks for the reply. So, return loss has no impact on the noise figure. But when you say that NF = IL + whatever you have after it till an active device, I am not sure what it means because this is the only information I have to calculate the NF i.e. RL and IL.

 

[DDavid]

let take example:
if u have cable with IL of 5dB and after cable LNA with NF=2dB
than total NF will be around (at least) 7dB.

another
IL cable =5dB, IL BPF=3dB, NF LNA=2dB
NT total>=10dB

 

[mw_rookie]

Thanks for the reply and also, thank you very much for clearing this up.

 

[vfone]

There could be a situation when things are different.
Let's say you have a BPF with 3dB insertion loss, and an LNA with 2dB NF when its input is 50 ohms, pure resistive.
Theoretically total NF if you cascade them, should be 5dB, but when you connect together you find that the total NF is 4dB (1dB better).
What happened? The LNA NF was not optimized for gamma-optimum at the input and when attached the BPF, the LNA sees a complex impedance which happened to be exactly the gamma-optimum for lowest NF. So, the NF of the LNA gets better when attached the BPF at the input.

 

[mw_rookie]

Thanks for the insight.

 

[RCinFLA]

There could be a situation when things are different.
Let's say you have a BPF with 3dB insertion loss, and an LNA with 2dB NF when its input is 50 ohms, pure resistive.
Theoretically total NF if you cascade them, should be 5dB, but when you connect together you find that the total NF is 4dB (1dB better).
What happened? The LNA NF was not optimized for gamma-optimum at the input and when attached the BPF, the LNA sees a complex impedance which happened to be exactly the gamma-optimum for lowest NF. So, the NF of the LNA gets better when attached the BPF at the input.

Usually things get worse because a NF optimized match may not terminate the filter properly resulting in higher ripple in filter bandpass. It is important to strive for good S11 at optimum noise figure if LNA is to terminate a filter.