The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> Thermal noise
https://designers-guide.org/forum/YaBB.pl?num=1419514653

Message started by raja.cedt on Dec 25^th, 2014, 5:37am

Title: Thermal noise
Post by raja.cedt on Dec 25^th, 2014, 5:37am

Dear all,
If input impedance of two port network is resistive (irrespective of the circuit) then we can conclude noise at that port is 4KT*Real(Zin). This was stated very clearly in Probability, Random Variables and Stochastic Processes.

Now the question is in case of inductively degenerated LNA input impedance is resistive however there is no resistive element here, so it should have similar noise performance compared to 50Ohm terminated LNA's, but generally people easy it is less noisy(for the time being assume transistor noiseless)

Thanks,
Raj.

Title: Re: Thermal noise
Post by aaron_do on Jan 6^th, 2015, 5:10am

Hi Raj.,

do you have a reference for this?

Quote:

If input impedance of two port network is resistive (irrespective of the circuit) then we can conclude noise at that port is 4KT*Real(Zin). This was stated very clearly in Probability, Random Variables and Stochastic Processes.

The way that you have stated it, it is clearly not true since I can always design a port which looks like 50-ohm but has much more noise (i.e. not equal) than 4kTR. In fact even a common-gate LNA can have less than 3 dB NF without the use of any kind of feedback (calling CG negative feedback is debatable) or filtering/transformation network.

Aaron

Title: Re: Thermal noise
Post by rfcooltools.com on Jan 9^th, 2015, 5:57pm

if source is a 50 ohm resistor and the input to your circuit is indeed a 50 ohm resistor then the noise is as you state. Intuitively it can be viewed as NF will be 3dB since 6dB of signal voltage loss and noise voltage drops by 3dB due to the parallel combination of the two resistors.

With the active circuit with no noise it the input the signal voltage is reduced by 6dB and the noise voltage by the same 6dB due to the parallel combination of the source resistance and noiseless input resistance. in this ideal case the noise figure will be 0.

http://rfcooltools.com