Comparative study of noise in low-current Townsend discharge in nitrogen and argon

Noise in a low-current Townsend discharge in argon and nitrogen is studied. Experiments were performed in a “semiconductor–gas-discharge” structure constituted by a short plane-parallel discharge gap and a high-resistivity photosensitive semiconductor electrode. The noise of optical emission from gas excited by the discharge was investigated with a photomultiplier. Experiments were done at the room temperature of the discharge structure. According to the obtained data, the noise of discharge glow in Ar is astonishingly low—similar to that earlier found for the cryogenic discharge in this gas. At equal time-average photon fluxes emitted by discharges in both gases, the noise of gas glow in nitrogen substantially exceeds the corresponding value for argon. A statistical processing of signals of glow brightness for the two gases revealed the origin of this difference: While the power spectral density of noise in nitrogen is specified by a band of frequencies with the increased density, in argon it is nearly constant in an extended frequency range. At the same time, intensity of noise is practically the same for both gases in the ranges of low and high frequencies. The relationship of the increased noise in nitrogen with a low critical current density for oscillatory instability of the Townsend discharge in this gas is considered.

Figure 1

Schematic presentation of the experimental setup. For the details, see text.

Figure 2

Examples of dependencies of the time-average PMT signal on current in the SDG structure. (1) $d=0.1$ mm, $p=120$ hPa; (2) $d=0.26$ mm, $p=46$ hPa; (3) $d=0.18$ mm, $p=50$ hPa. The straight lines correspond to the linear rise in the gas brightness with increasing current in the structure.

Figure 3

Dependencies of the standard deviation of the PMT signal on its time-average value. Full symbols represent the data for the glow of a discharge gap filled with argon: (1) $d=0.18$ mm, $p=50$ hPa, and (2) $d=0.84$ mm, $p$ = 100 hPa. Open circles (3) correspond to recording light from an incandescent lamp. The solid curve shows dependence (1) that approximates sets of experimental points (see text).

Figure 4

Dependencies of the standard deviation of the gas glow signal on its time-average value obtained in two conditions of filling the discharge gap by nitrogen at $pd\approx \text{const}$ (full symbols): (1) $d=0.1$ mm, $p=120$ hPa, and (2) $d=0.26$ mm, $p=46$ hPa. The open circles (3) represent the data for light from the incandescent lamp. The solid curves correspond to the results of fitting of the approximating dependence (1) to experimental data.

Figure 5

Spectral distributions of the noise power density for discharge glow in argon and nitrogen at the same time-average PMT signal $m=0.058$ and 0.0081 V.

Figure 6

Probability distribution functions for fluctuations in the discharge glow in an SDG structure. (1) discharge in Ar, $d=0.18$ mm, $p=50$ hPa, $I=248 \mu \mathrm{A}$; (2),(3) discharge in ${\mathrm{N}}_2, d=0.26$ mm, $p=46$ hPa, (2) $I=234 \mu \mathrm{A}$, (3) $I=307 \mu \mathrm{A}$. The solid lines correspond to the Gaussian distribution of the probability density. The distributions are normalized to the maximum probability density.

Figure 7

The ratio of the standard deviation of the PMT signal to its time-average value as dependent on the time-average value. (1) discharge in ${\mathrm{N}}_2, d=0.26$ mm, $p=46$ hPa; (2) discharge in Ar, $d=0.18$ mm, $p=50$ hPa.