Input Noise Matching on LNA

[Faizan shafi]

Hi
I am designing a LNA and for that I am simulating few topologies and seeing the nf , gain of my circuits. as you can see the schematic and graphs, you can tell my noise figure would be large because of no input output matching.

My supervisor told me to play with the values of Zs so that you get less noise figure. He told me specifically not to build a matching network but to just change or sweep the values of Zs and get a noise figure which is smaller than what it is now.
I don't want to sweep the values of Zs. Instead I just want to clear few concepts.

I have plotted Sopt and Zopt on the graph.

At 61Ghz If i look at the values of Zopt, they are 55+41j .
So now If i just change my Zs to 55+41j would it be considered as input matching for minimum noise figure?


Similarly if i look at Zin and lets say it is 12+14j.
And now i change my Zs to 12-14j . Would it be considered as input matching for maximum noise..


AM i correct?

Also can someone explain to me how can I use S11, Sopt for matching my circuits? I am little confused about it. Thank you

 

[BigBoss]

IF Z_opt is known well, Z_source has to be equal to this impedance for Lowest Noise. (Noise Matching)
IF Z_in is known well AND IF you make Z_source Complex Conjugate of this Impedance you obtain Maximum Power Transfer for Input Side. But this does not give you max. Noise. There is no such definition Max. Noise.
If you do Simultaneously Conjugately Matching ( Condition: IF k>1 ) you obtain Max. Available Gain.

Read Guillerme Gonzalez's Textbook for detailed explanations.

 

[Faizan shafi]

Thank you fro replying
"
IF Zopt is known well, Zsource has to be equal to this impedance for Lowest Noise. (Noise Matching)
IF Zin is known well AND IF you make Zsource Complex Conjugate of this Impedance you obtain Maximum Power Transfer for Input Side"

So my concepts were correct regarding these two points..

But you still havent answered my original question.


I have plotted Sopt and Zopt on the graph by using the Zopt block from S paramter-simulation palette.
At 61Ghz If i look at the values of Zopt, they are 55+41j .
So if i change my Zs tp 55+41j- It would be good for impedence matching at the input for low noise figure. I should get a smaller value of noise figure on the input side . right?

But thats not happening at the moment. When i change the value of Zs to match the value of Zopt . and simulated my circuit . I dont get the results what i wanted.
MY zopt also changes by changing Zs.
My noise figure at the input also doent get lower.

Mind you, I dont wanna build the matching network at the moment. I just wanna play with Zs values to get a small noise figure.
Does my question makes sense?

 

[BigBoss]

I have plotted Sopt and Zopt on the graph by using the Zopt block from S paramter-simulation palette.
At 61Ghz If i look at the values of Zopt, they are 55+41j .
So if i change my Zs tp 55+41j- It would be good for impedence matching at the input for low noise figure. I should get a smaller value of noise figure on the input side . right?

But thats not happening at the moment. When i change the value of Zs to match the value of Zopt . and simulated my circuit . I dont get the results what i wanted.
MY zopt also changes by changing Zs.

Don't change the Characteristic Impedance (50 Ohm). The Impedance which you've been searching is the Optimum Noise Impedance.
Do you matching and check the Noise Curve.