Why match LNA with capacitive input?

[analog_chip]

Hi friends,
Why is it necessary to match LNA at the input if input is a gate i.e. purely capacitive. I understand max power transfer and min noise figure concepts. My question is since we are concerned about voltage swing at the LNA input, why not drive the LNA directly with source ( Rs=50 ohm).

 

[johnjoe]

You have to match the complex impedance not only a real value. And conjugate Match and noise match can have different values.

 

[analog_chip]

Let us assume Zin=-j100 (purely cap). We can have voltage swing even without matching right? since cap does not consume power.

 

[johnjoe]

You will have some voltage swing, but it's not matched!

 

[analog_chip]

Once I have the voltage swing the LNA can give me amplified output, right? Why do we need to match since its not a T line and we dont need to worry about reflections etc.

 

[FvM]

Your initial assumption is wrong. LNA respectively FET input impedance is never purely capacitive. You possibly want to review s-parameters of a real LNA.

LNAs are typically used for applications related to their name Low Noise Amplifier. If low noise is an objective, impedance matching matters.

 

[johnjoe]

Your LNA will work SOMEHOW without matching. But I think, you'll have some specs which you have to fulfill.

 

[analog_chip]

Hi Fvm & jonjoe,

Thanks for the reply! I just wanted to confirm that in ideal condition (purely capacitive) you dont need matching at all.