Hi Jorcy. Temperature is the Kelvin temperature of the medium carrying the electrical power. Any medium carrying electrical power, including circuits or cables. It's essentially electrons (or holes) banging around in the medium. Thermal noise is a function of temperature (the lower the temperature, the less random motion of electrons or holes) and the bandwidth of the circuit or communications channel (the wider the bandwidth, the more noise the circuit is subjected to).

In resistors, the total noise figure is also a function of resistance.

Vnoise = sqr(4KT * resistance * bandwidth)

where that resistance can refer to the equivalent resistance of a series and parallel resistor network.

This is a good explanation:

https://en.wikipedia.org/wiki/Johnson%E2%80%93Nyquist_noise

Hi Jorcy. Temperature is the Kelvin temperature of the medium carrying the electrical power. Any medium carrying electrical power, including circuits or cables. It's essentially electrons (or holes) banging around in the medium. Thermal noise is a function of temperature (the lower the temperature, the less random motion of electrons or holes) and the bandwidth of the circuit or communications channel (the wider the bandwidth, the more noise the circuit is subjected to).

In resistors, the total noise figure is also a function of resistance.

Vnoise = sqr(4KT * resistance * bandwidth)

where that resistance can refer to the equivalent resistance of a series and parallel resistor network.

This is a good explanation:

https://en.wikipedia.org/wiki/Johnson%E2%80%93Nyquist_noise

I should add that thermal noise is usually normalized, and listed as nV/sqr(Hz).

One example I ran into, where this became important, was when I initially couldn't figure out what resistance values to use in an opamp's input and feedback circuits. It's easy enough to pick values for the appropriate gain, input loading, and so on, but what order of magnitude to use? Kohms? 100Kohms?

Well, thermal noise ends up being the other constraint. You want to choose resistor network values that result in (hopefully) lower thermal noise values than the noise figure listed in the opamp spec sheet. So I tried this in real life. Choose higher resistance, you hear quite a bit of white noise. Choose resistor values that result in approximately the same noise as the opamp, and the result was much quieter operation. Then choose lower resistance values, so the resistors created less noise than the opamp, and the improvement was only marginal at best. Because the opamp was creating most of the noise now, and reducing the resistance noise didn't really help. (I suppose it would have helped if I had changed to a quieter opamp!)

Just as the text book says.

it is the medium under which the signal is being examined or measured, generally this gives us the present of noise power in a given bandwidth of observation/calculation/measurement - it is the noise floor

The equation refers to the noise generated in the signal due to temperature at which it is being operated. Here temperature refers to KELVIN.

The quantity KTB is the available noise power from a lossy element, such as a resistor, in bandwidth B Hz.  The temperature T is the absolute physical temperature in degrees Kelvin of the lossy element. It is the temperature as read by a thermometer in contact with, and in thermal equilibrium with, the lossy element.  K is Boltzmann's constant. By available is meant the power that could  be delivered to a matched load.  This quantity is quite independent of any "signal" that may be of interest to you. Any lossless transmission medium contributes no noise, whatever its physical temperature. It should be understood that the formula is a low frequency approximation. It is not valid in the millimetric wavelength range at low temperatures. Active devices, transistors etc, generate noise which is not representative of their physical temperature. Often the concept of equivalent noise temperature is used in such cases.

Mr. John

In this sense, what would be the Equivalent Noise in such active elements (e.g. AmpOps/Transistors) when considering its network Gain Characteristics ?

Best Regards

Equivalent noise depends on the electronics, as your question points out. Signal to noise ratio is degraded in the Low Noise Amplifier (which amplifies the received signal from the antenna) by the Noise Factor F, which commonly ranges from 0.1dB

Mr. Nettleton

Thank you very much for these slides and also for the very straightforward explanation about Noise Figure. 

Also, I hadn't actually observed which would be the equivalent Bandwidth, since the -3dB criteria is the usual "thumb rule" for most of the active filtering applications (or the Null-to-Null BW typically on wireless devices)

Best Regards