NF of LNA remains the same as load changes

[hhxuexia]

Hi,guys!

I am designing a two stage LNA, and confused with the following problem.

when the load impedence changes, the gain of second stage changes, but the NF remains the same. In my opinion, as the gain changes, NF should change, since noise contribution of the second stage changes.

So, why NF remains the same? Thanks!


hhxuexia

 

[johnjoe]

Hi,

your dominating noise source is the first stage and when it has sufficient gain the noise contribution of the second stage is less -> Friis-Formula (as approximation because this formula is just true under perfect matched conditions)

The impact on noise figure due the load impedance should be quite low.

Regards

 

[hhxuexia]

Thanks, johnjoe.

yes, you are right. The dominate noise source is the first stage. But, according to my simulation, noise contribution of the first stage changes due to the variation of load impedance, and the noise figure remains exactly the same. I can not understand it.

Hi,

your dominating noise source is the first stage and when it has sufficient gain the noise contribution of the second stage is less -> Friis-Formula (as approximation because this formula is just true under perfect matched conditions)

The impact on noise figure due the load impedance should be quite low.

Regards

 

[johnjoe]

Can you post maybe a schematic of your LNA for clarification?

What be helpful

 

[drkirkby]

Thanks, johnjoe.

yes, you are right. The dominate noise source is the first stage. But, according to my simulation, noise contribution of the first stage changes due to the variation of load impedance, and the noise figure remains exactly the same. I can not understand it.

The noise figure will be dominated by the first stage. Whilst the overall noise figure will depend on later stages, their contibution might be sufficiently small you can't actually measure it relieably. There's a fair degree of uncertainty when making noise measurements - it's not like frequency, which can be measured very accurately. Are you for example updating the noise figure meter if the room temperature changes, or do you have a modern instrument where the temperature sensor is in the noise souce? Or are you like 90% of the people I've seen using noise figure meters, unaware that this is a necessary step for accurate measurements?

 

[johnjoe]

The noise figure will be dominated by the first stage. Whilst the overall noise figure will depend on later stages, their contibution might be sufficiently small you can't actually measure it relieably. There's a fair degree of uncertainty when making noise measurements - it's not like frequency, which can be measured very accurately. Are you for example updating the noise figure meter if the room temperature changes, or do you have a modern instrument where the temperature sensor is in the noise souce? Or are you like 90% of the people I've seen using noise figure meters, unaware that this is a necessary step for accurate measurements?

You are right, but he is still simulating and not measuring his LNA. In simulation you can quite well estimate the noise contribution of each stage, but as i said just in simulation!

Best regards

 

[hhxuexia]

I do not have a schematic at hand, try to post it tomorrow.

Thanks!

hhxuexia

Can you post maybe a schematic of your LNA for clarification?

What be helpful

 

[vfone]

I think you set the simulator to check only the NF of the first stage, and not of the system (both stages).
If is looking only to the NF of the first stage, the gain variations of the first stage would not influence its own NF.

Based on Friis equation: "The NF of the second stage is reduced by the gain of the first stage"
F_total = F1 + (F2-1)/G1 + ...