selfnoise

Self-noise, or intrinsic noise, refers to random fluctuations generated by a system in the absence of an external signal. It represents the fundamental limit on the smallest signal a device can reliably detect or reproduce and is a key consideration in acoustics, electronics, and precision instrumentation.

In electronic and audio contexts, self-noise is often described by its spectral density or as equivalent input noise. For a sensor or amplifier, the device’s output noise can be expressed as if it were produced by a hypothetical input signal, a quantity known as the equivalent input noise (EIN). In microphones, self-noise is commonly reported as a noise figure or equivalent input level and may be described relative to at-sensor sound pressure level. In voltage-sensitive devices, noise density is typically given in units such as nanovolts per root hertz. Thermal (Johnson–Nyquist) noise is a fundamental limit arising from the random motion of charge carriers and is proportional to the square root of temperature, resistance, and bandwidth: sqrt(4kTRB).

In seismology and instrumentation, a device’s self-noise defines the minimum ground motion it can faithfully record. This is usually presented as a power spectral density versus frequency and is compared to the ambient environmental or Earth background noise to assess sensitivity.

Significance and mitigation: Self-noise constrains the signal-to-noise ratio and the achievable fidelity of measurements. Reducing self-noise involves selecting low-noise components, proper impedance matching, shielding and grounding, cooling where feasible, and limiting system bandwidth. Digital techniques such as oversampling or careful clocking can also help minimize perceived self-noise.

The term is used across fields—from audio equipment to scientific instruments—where the internal noise floor determines performance independent of external conditions.