Random fluctuations in a system caused by the thermal motion of its constituent particles.Thermal noise, also known as Johnson–Nyquist noise, is the random electrical voltage or current that arises in any conductor or electronic component due to the thermal agitation of its charge carriers (usually electrons). At any temperature above absolute zero, particles are in constant, jittery motion; in a conductor, this motion translates into tiny, fluctuating electric fields that appear as a background hiss in any measurement. It is a fundamental, unavoidable limit: you cannot reduce it by improving the quality of a device, only by cooling it or narrowing the measurement bandwidth.
The effect was first measured by John B. Johnson at Bell Labs in 1926, and explained theoretically the same year by Harry Nyquist, who derived the formula for the noise power using thermodynamics and statistical mechanics.
In a resistor, electrons are constantly colliding with the lattice of atoms, and their random thermal velocities create a fluctuating voltage across the terminals. The average power of this noise is proportional to the absolute temperature and the frequency bandwidth being observed, independent of the material or resistance value. This is why sensitive detectors, from radio telescopes to neural recording rigs, must be cooled to cryogenic temperatures — to lower the thermal energy and thus the noise floor that masks the tiny signals of interest.
Thermal noise sets the fundamental sensitivity limit in every electronic measurement, from gravitational-wave detectors to the amplifiers in your smartphone's receiver; it is also exploited as a natural source of random numbers in cryptographic hardware.
The same thermal noise that limits the sensitivity of neural recordings is also what allows certain biological neurons to fire spontaneously even in the absence of a stimulus, a phenomenon that helps keep the brain's circuits primed.




