Noise in materials and electronic devices, noise as a diagnostic tool

Prof. L. K. J. Vandamme
Department of Electrical Engineering, Eindhoven University of Technology, Eindhoven, The Netherlands

20 hours, 5 credits (final test)

October 2008

Dipartimento di Ingegneria dell'Informazione: Elettronica, Informatica, Telecomunicazioni, via G. Caruso, meeting room, ground floor

Contacts: Prof. Massimo Macucci

Outline

I. Introduction

• Short history of noise research, some objectives and definitions, measurement set ups.

II. Noise sources, origins, and remedies

• Noise sources: thermal, shot, generation-recombination, RTS (random telegraph signal noise) and 1/f noise.
• Noise equivalent circuits for passive components.

III. Resistance fluctuations in networks and continuous media

• Sensitivity coefficient for two ports.
• Sensitivity coefficient for four ports.
• Current crowding increases the effect of conductance fluctuations.

IV. Experimental facts on 1/f noise in metals and semiconductors

• Hooge’s empirical relation for the 1/f noise.

V. Noise in electronic devices: resistors, MOSFET, diodes, BJT

• Resistors (Au-layers; poly Si; poly SiGe; thick film resistors).
• Diodes (shot noise; 1/f and RTS noise; perimeter or bulk contributions; faster diodes are noisier).
• MOSFETs:
• Origin of 1/f noise in MOSFETs, as a surface or bulk effect, related to carrier trapping or mobility fluctuations.
• Circuit oriented noise equations in terms of the Hooge parameter for biasing conditions in the Ohmic/Saturation or Sub-threshold region.
• Relationship between device speed and noise.
• Review of experimental results.
• Misunderstandings about 1/f noise in MOSFETs: RTS noise is an additional number fluctuation noise on top of 1/f noise.
• BJTs and HBTs.
• Flicker noise in bipolar transistors: is it a perimeter effect?
• Corner frequency f_c, is it a figure of merit?
• Other possible figures of merit.
• An equivalent circuit: problems and formulas.
• How to measure the dominant noise source by means of coherence techniques and changing the termination resistance.
• Experimental results on poly-emitter BJTs, HBTs.
• Noise figure at medium frequencies.

VI. 1/f noise as a diagnostic tool for quality evaluation of materials and devices

• Different noise types such as: shot noise, burst noise or RTS, generation recombination noise and the omnipresent 1/f noise have different meanings in view of reliability diagnostics.
• Current crowding makes 1/f noise a sensitive quality indicator of external and internal contacts in materials and devices (e.g. submicron-MODFETs, conductive adhesives, thin film transistors, poly silicon layers).
• 1/f noise in contacts.
• Low frequency noise as an indicator of reliability problems in diode type devices (e.g. p-n junctions, Schottky-barriers and bipolars) and in resistor type devices.
• Some case studies: polymer transistor, nano-particle dielectrics, MESFETs, MODFETs, solar cells and LEDs.
• Correlation between noise in devices and reliability.
• Noise measurements are often more sensitive and less destructive than classical tests after accelerated life tests.

VII. Low frequency noise measuring set ups

• Time domain analysis for non-Gaussian noise mixed up with Gaussian noise (RTS with e.g. 1/f noise) and analysis in the frequency domain.
• Derivation of the resistance fluctuation spectrum from that of voltage and/or current fluctuations.
• Noise characterization of amplifiers.
• Voltage amplifiers in series.
• Amplifier choice, low noise current or voltage amplifier, transformer amplifier, or correlation measurements.
• Sample design based on a high corner frequency value: four contacts on a sample are not always a guarantee of contact noise suppression.
• Measuring setup (band pass filtering or line spectra, FFT).
• Shielding.