Enhanced statistical detection of random telegraph noise in frequency and time domain

Owen Gauthier (1, 2) , Sébastien Haendler (1) , Patrick Scheer (1) , Alexandre Vernhet (1) , Quentin Rafhay (2) , Christoforos Theodorou (2)

1 ST-CROLLES - STMicroelectronics [Crolles]
2 IMEP-LAHC - Institut de Microélectronique, Electromagnétisme et Photonique - Laboratoire d'Hyperfréquences et Caractérisation

In this work, a new statistical detection method of Random Telegraph Noise (RTN) in the frequency domain is presented. An algorithm for the automated detection of Lorentzian spectra in the noise power spectral density (PSD) of a device is proposed, which enables the processing of a large amount of experimental data. Using 40 nm Bulk CMOS technology as a test vehicle, we demonstrate that the detection of Lorentzian spectra in the noise PSD allows an easier, faster, and often more precise detection of RTN presence compared to the time domain detection.

Mots clés: Random Telegraph Noise; Lorentzian noise; detection algorithm; statistical analysis; CMOS transistor

Domaines: Micro et nanotechnologies/Microélectronique; Electronique; Analyse de données, Statistiques et Probabilités [physics.data-an]

Owen Gauthier, Sébastien Haendler, Patrick Scheer, Alexandre Vernhet, Quentin Rafhay, et al.. Enhanced statistical detection of random telegraph noise in frequency and time domain. Solid-State Electronics, 2022, 194, pp.108320. ⟨10.1016/j.sse.2022.108320⟩. ⟨hal-03762370⟩