Thermal noise & sensitivity

Receiver noise figure NF is in dB. To combine receiver noise with source noise, first convert it to linear noise factor $F$:

Using the standard 290 K noise-figure reference, the equivalent receiver-added noise temperature is $T_e=290(F-1)$ kelvin. This describes the receiver’s added noise referred to its input; it is not the hardware temperature.

Source noise temperature $T_s$ and receiver-added noise temperature combine to give total input-referred noise power $N$:

Temperatures are in kelvin and $N$ is in watts. Boltzmann’s constant $k$ is exactly $1.380649\times10^{-23}$ J/K. $B$ is equivalent noise bandwidth in hertz: the rectangular-filter width passing the same white-noise power at the same reference gain. It need not equal the −3 dB bandwidth. Source noise alone is $kT_sB$; the model assumes independent, flat source and receiver noise with matched available power at one input plane.

Sensitivity is the signal power needed for the required signal-to-noise ratio. Convert total noise to dBm, then add required signal-to-noise ratio and implementation margin $M$ in dB:

Required SNR may be negative; implementation margin is non-negative. These inputs raise or lower the sensitivity threshold without changing the calculated noise power.

The compact noise table compares source-only, receiver-added and total noise at the same input reference plane. Its Kelvin readings are equivalent noise temperatures, not the receiver’s physical temperature. Generated bandwidth and NF/temperature conversions display at most two decimals while retaining full precision in calculations and copied inputs until edited.