Electro-acustic influence of the measuring system on the photoacoustic signal amplitude and phase in frequency domain

Abstract

The paper discusses the most common impacts of the measuring system on the amplitude and phase of the photoacoustic signals in the frequency domain using the opencell experimental setup. The highest signal distortions are detected at the ends of the observed modulation frequency range from 20 Hz to 20 kHz. The attenuation of the signal is observed at lower frequencies, caused by the electronic filtering of the microphone and sound card, with characteristic frequencies of 15 Hz and 25 Hz. At higher frequencies, the dominant signal distortions are caused by the microphone acoustic filtering, having characteristic frequencies around 9 kHz and 15 kHz. It has been found that the microphone incoherent noise, the so called flicker noise, is negligibly small in comparison to the signal and does not affect the signal shape. However, a coherent noise originating from the power modulation system of the light source significantly affects the shape of the signal in the range greater than 10 kHz. The effects of the coherent noise and measuring system influence are eliminated completely using the relevant signal correction procedure targeting the photoacoustic signal generated by the sample.

Experimental photoacoustic observation of the photogenerated excess carrier influence on the thermoelastic response of n-type silicon

“…The transmission configuration was adopted to assure 1D heat conduction along the uniformly irradiated plate, which allows using simple theoretical models to interpret the obtained experimental data. Taking into account that frequency response of the microphone and accompanying electronics has bandpass (electronic filter) characteristics that depend on the design, manufacturing, and diaphragm properties, we applied a standard signal correction procedure 27,28 to obtain undistorted signals accurately representing the physical characteristics of the considered samples.…”

“…Unfortunately, noise and signal distortion often mask δp total (f ), so their analysis is an unavoidable task in accurate and precise PA measurements. 27,28 Main types of noise in PA measurements are flicker noise N FN (f ) and coherent signal deviations N CSD (f ). Flicker noise is typically generated by the random and nonreducible microphone noise.…”

“…Taking into account the main types of noise and signal distortions, it is clear that the correction of S exp (f ) is needed to obtain δp total (f ). If we measure and subtract N CSD (f ), assuming that N FN (f ) is negligible, the values of S exp (f ) can be fitted with S fitt (f ) given by 27,28…”

“…To find the sample properties, one has to fit δP(f ) to the δp total (f ) function determined by the well-known composite piston theoretical model, which, for semiconductors as the n-type Si considered in this work, establishes that [19][20][21][27][28][29]…”

Developing the Techniques for Solving the Inverse Problem in Photoacoustics

“…Resonant peaks observed in this part of frequency domain are, throughout literature, attributed to the influence of measurement chain, although, in measurements, they occur at frequencies lower than expected (frequency characteristic of the microphone, the amplifier, and other electronics) [27][28][29]. Minimum volume cell has already been observed as an electro-acoustic resonator and it has been modeled with cascade filter array, with transfer function represented as the combination of two Helmholtz resonators:…”