What Is 1f Noise?

1/f noise, also called flicker noise or pink noise, is a type of electrical noise whose power spectral density is inversely proportional to frequency, meaning that noise power increases as frequency decreases. The relationship follows a 1/f^α law, where α is typically close to 1. This distinguishes 1/f noise from white noise, which has a flat spectrum, and from other colored noise types. 1/f noise appears in a wide range of physical systems, from electronic components to biological signals, but its most consequential engineering manifestation is in semiconductor devices, where it sets a fundamental limit on low-frequency circuit performance. The effect was first systematically characterized in vacuum tubes by J.B. Johnson in the 1920s and later studied extensively in transistors and resistors.

The noise is characterized as excess noise, meaning it adds to the irreducible thermal noise floor rather than replacing it. At frequencies above a threshold known as the corner frequency, thermal noise (white noise) dominates; below the corner frequency, 1/f noise becomes the dominant contributor.

Physical Origins

Despite decades of study, the microscopic origin of 1/f noise in semiconductors is still an active area of research, with two competing models the most widely cited. The carrier number fluctuation model, developed by Andrew McWhorter in 1957, attributes the noise to random trapping and detrapping of charge carriers at oxide-interface trap sites. In MOSFETs, these traps are located at or near the silicon-silicon dioxide interface, and their random occupancy modulates the channel carrier density. The mobility fluctuation model, proposed by F.N. Hooge in 1969, instead attributes the noise to fluctuations in carrier mobility throughout the bulk of the semiconductor. In practice, both mechanisms contribute to different degrees depending on device type and bias conditions. Research on noise in semiconductor devices and thin-film materials documents both mechanisms and their dependence on fabrication process parameters.

Frequency Characteristics and Corner Frequency

The corner frequency is the frequency at which the 1/f noise power equals the white noise power, and it varies widely across device types and operating conditions. In bipolar junction transistors (BJTs), the corner frequency is typically in the range of hundreds of hertz to a few kilohertz. In CMOS transistors, it can range from tens of hertz to hundreds of kilohertz, with PMOS devices generally exhibiting lower 1/f noise than NMOS devices in standard bulk CMOS processes due to differences in trap density and carrier type. High-electron-mobility transistors (HEMTs) used in low-noise RF amplifiers can exhibit corner frequencies below 1 MHz. The Analog Devices technical article on understanding and eliminating 1/f noise provides quantitative guidance on how to estimate corner frequency from datasheet noise figures and how it affects system-level noise performance in precision analog and sensor signal chains.

Mitigation Techniques

Engineers manage 1/f noise through circuit topology choices, device selection, and signal processing. Chopper stabilization shifts the signal band to a frequency above the corner frequency, effectively moving the signal out of the 1/f region before amplification and then demodulating it back. Auto-zero amplifier architectures periodically sample and subtract their own offset and low-frequency noise. At the device level, larger gate-area transistors have lower input-referred 1/f noise because the larger volume averages over more traps. EDN's analysis of 1/f noise behavior across circuit applications explains why buried-channel devices and careful layout practices also reduce its impact in precision circuit design.

Applications

1/f noise analysis and mitigation is relevant across a range of fields, including:

• Low-noise amplifier design for radio astronomy and wireless receivers
• Precision analog circuit design in instrumentation and data acquisition
• Phase noise in oscillators and frequency synthesizers for communications
• MEMS sensor design for inertial measurement and chemical sensing
• Biomedical signal acquisition where low-frequency physiological signals are measured