An Atmospheric Plasma Jet Generator Driven by a Current Source

Abstract

A novel system aiming to electrically supply various cold plasma generators is proposed. It operates as a programmable linear current source which is able to maintain a dc constant discharge current at various discharge voltages required to sustain the plasma jet. Its design is based on a specific electronic device called a switchable current regulator, which considerably simplifies the circuit topology. Experimental results carried out in real operating conditions confirm the practical purpose of the proposed solution.

1. Introduction

Cold plasma, also known as non-thermal plasma, is generated at atmospheric pressure and consists of a partially ionized gas. The bulk gas remains at low temperature (<100 °C), in association with high kinetic energy electrons, reactive species such as free radicals, heavy ions, and ultraviolet radiation. As a result, cold plasma produces various phenomena which can transform the surface state that it interacts with without the presence of certain thermal effects. This specific characteristic has led to the development of numerous applications related to biology [1,2], agriculture and food industry [3,4], medicine [5,6,7,8,9,10,11,12], and materials science [13,14]. There are many methods of generating plasma at atmospheric pressure; extensive reviews can be found in [15,16,17]. Plasma generating devices based on the dc electrical discharge can have various designs, since they do not include tuned or matching circuits as in RF or microwave discharges. Discharge voltage and current can have important variations depending on the surface state and geometry of the electrodes, the nature and flow rate of the working gas, the nature, geometry, and surface state of the object being in contact with plasma, and other random external factors. In certain applications, such as medical fields or plasma treatment of materials, it is necessary to control the discharge current in order to keep the plasma jet temperature low enough to avoid destruction of biological tissues, occurrence of the pain sensations in the case of living organisms, or surface damage in the case of materials. In addition, maintaining the discharge current at a constant and well-known value is necessary in order to achieve a high degree of repeatability on the part of the plasma treatment results. The aim of the current work is to devise a simple and robust electrical supply system for cold plasma generators that is able to keep the discharge current at a programmed value. Although this does not fully solve the problem of the plasma properties’ stability under various operating conditions, it is the most convenient approach because the electrical circuits can be designed to have stable characteristics over time. Current technological progress allows for the development of solutions based on non-conventional components and devices that simplify or improve the design of the electrical circuits.

2. Experimental Setup

A schematic diagram of the experimental setup is shown in Figure 1. Typically, in the case of a dc generated plasma, the voltage required to be applied to the electrodes for its ignition is much higher than the voltage necessary to sustain it [18,19]. To form the plasma jet, two dc negative voltage sources, respectively denoted EH and EM, were used. One of them, EH, is intended to ignite the electrical discharge, whereas the second, EM, is intended to sustain the electrical discharge. The EH voltage source only has the role of igniting the electrical discharge. A conventional dc voltage multiplier [20,21] was used for this purpose. After the plasma jet is formed and becomes stable, the EH voltage source operation can be switched off. However, to secure the re-ignition of the electrical discharge in the event of its accidental interruption, the EH voltage source was permanently switched on, giving a fixed output dc voltage $U_H\cong$−3.2 kV that was not changed during the experiments. Its output is connected to the cathode of the plasma generating device through a resistor $R_H$ = 10 MΩ. Because the current due to the EH voltage source (on the order of $I_H<U_H/R_H\cong$ 0.32 mA) can be neglected, its contribution to the discharge current $I_d$ (which is on the order of 5–50 mA) is considered as having no significance for the experiment outcomes.

The EM voltage source, giving an adjustable output dc voltage $U_M$ in the range from 0 to −650 V, is used to sustain the electrical discharge.

An IXYS IXCP 10M45S switchable current regulator is used as a current source device (CSD). This device has three terminals: the anode (A), cathode (K), and gate (G), and is intended to keep constant the anode (A) to cathode (K) current in the range 2–100 mA as a function of the voltage applied between the gate (G) and cathode (K) terminals [22]. The anode–cathode current can be programmed by means of an external negative dc voltage $U_{GK}$ applied between the cathode and the gate terminals.

To afford the control voltage $U_{GK}$, a single section of the integrated circuit (Avago Technologies ASSR-V621 type [23]) is used, which is a dual channel photovoltaic MOSFET driver able to generate 7 V at the output terminals (open circuit state) for input current $I_F$ = 10 mA. Its input and output circuits are galvanically isolated and the output voltage is determined by the input current $I_F$.

The EM voltage source output is connected to the cathode of the plasma-generating device through a circuit consisting of resistor R_S, CSD, resistor R_B, and diode D_H, all of which are series-connected. This voltage source provides the most important contribution to the discharge current $I_d$ and represents the main supply voltage.

To monitor the operating parameters of the electrical circuit and implicitly of the plasma source, four voltages $U_1,U_2,U_3$ and $U_4$ were measured at different points of the main supply circuit, as shown in Figure 1.

The resistor $R_S$ = 1 kΩ acts as a current sensor. The discharge current is calculated as $I_d=|U_2-U_1|/R_S$. To immediately observe the discharge current $I_d$, a battery powered digital voltmeter (not shown in Figure 1 and Figure 8) was connected across the resistor R_S, its indication expressed in volts being numerically equal to the discharge current $I_d$ expressed in milliamps. Next, $R_B$ = 10 kΩ acts as a ballast resistor; its presence is necessary in order to obtain a stable discharge for the range of variation of the discharge current taken into consideration. According to the Kaufmann criterion [24,25], in order to obtain a stable discharge, the following condition must be met: $R+d{U}_d/d{I}_d>0$. In the actual case considered here, $R=R_B+R_S$. This resistor also protects the CSD so that the voltage drop across the cathode–anode terminals does not exceed the rated value under the given experimental conditions. A higher value of ballast resistance improves discharge stability and the CSD safety, but reduces the discharge current and the range in which it can vary. The total ballast resistance $R_B+R_S$ was set at the design stage of the experiment to meet both constraints. Subsequent measurements certified that this was an appropriate value. The diode D_H (BY6–type) is inserted to block the current supplied by the voltage source EH from flowing through electrical circuit of the voltage source EM. Its forward voltage (<6 V/1 A [26]) is negligible compared to the discharge voltage $U_d$ (>100 V) as a consequence $U_d\cong \left|U_4\right|$.

Note that the CSD is not included in any additional feedback loop, acting only as a basic current regulator in its simplest form.

The four voltages $U_1,U_2,U_3$ and $U_4$ were measured using voltage dividers, each of them as shown in Figure 2. Due to spread of the resistor values around the nominal values indicated by the manufacturer, they were calibrated previously by applying a voltage of −500 V to their inputs and measuring the resulting voltage at the output. Outputs of the voltage dividers are connected to the inputs of a data acquisition board, which performs A/D conversion and sends the results to a personal computer for recording.

Figure 2. Schematic diagram of the voltage dividers.

Figure 2. Schematic diagram of the voltage dividers.

Due to the specific nature of the electrical discharges, the measured voltages show fluctuation. For this reason, the plotted or tabulated data represent time-averaged values over 100 samples, with the time step between two successive samples being ≈600 ms.

The axial section of the plasma-generating device is shown in Figure 3. The geometry is similar to that described in [27], consisting of an aluminum cylinder with a hollow cylindrical cavity 8 mm in diameter terminated at one end, with a conical ejector having a circular hole output 3 mm in diameter that serves to form and expel the plasma jet. This acts as an anode connected to ground. The cathode is formed by a thin metal bar 3 mm in diameter which slides along the longitudinal axis of the cylindrical cavity, its position being adjusted manually so as to obtain a specific optimal operating point that forms a stable plasma jet of maximum length. The working gas consists of Ar supplied through a lateral hole oriented perpendicular to the cylindrical cavity longitudinal axis. The gas is maintained at a constant flow rate by means of a gas flow controller (Alicat Scientific MC-20SLPM).

Figure 3. Axial section of the plasma generating device cavity (not to scale): 1—cathode; 2—anode; 3—high voltage insulator; 4—output hole; 5—working gas input.

Figure 3. Axial section of the plasma generating device cavity (not to scale): 1—cathode; 2—anode; 3—high voltage insulator; 4—output hole; 5—working gas input.

3. Results

To assess the effect of the CSD, a series of measurements without the CSD was performed. For this purpose, the CSD was removed and replaced by a conductive wire connecting points x and $x^{\prime }$ of the circuit. The current–voltage characteristic of the discharge for a series of experimental points within the discharge current range of interest is shown in Figure 4.

Figure 4. Current–voltage characteristic of the discharge. Experimental points are represented by black squares. The thin red line joining the first three experimental points was used to estimate the maximum differential resistance $d{U}_d/d{I}_d$. The working gas consisted of Ar at a flow rate of 4 LPM.

Figure 4. Current–voltage characteristic of the discharge. Experimental points are represented by black squares. The thin red line joining the first three experimental points was used to estimate the maximum differential resistance $d{U}_d/d{I}_d$. The working gas consisted of Ar at a flow rate of 4 LPM.

As can be seen in Figure 4, under the considered experimental conditions, the current–voltage characteristic exhibits negative resistance which varies as a function of the discharge current $I_d$. The maximum slope of the current–voltage characteristic appears to be in the region of the first three experimental points (joined by a thin red line). Based on this observation and taking into account the experimental values used to plot the current–voltage characteristic, it can be estimated as $|d{U}_d/d{I}_d{|}_{max}\cong$ 3.13 kΩ. When compared with total ballast resistance $R_B+R_S$ = 11 kΩ, this results satisfies the Kaufmann criterion.

Next, using the experimental setup shown in Figure 1, we measured the discharge current $I_d$ as a function of the main dc supply voltage $U_M$ under the specific experimental conditions listed in

Table 1. Experimental conditions related to the measurements.

No.	${\mathit{U}}_{\mathit{GK}}$ [V]	${\mathit{I}}_{\mathit{F}}$ [mA]
1	−2.9	5.2
2	−2.7	4.9
3	−2.5	4.6
4	N/A	No CSD

.

Figure 5 presents the variation of the discharge current $I_d$ as a function of the main dc supply voltage $U_M=U_1$, corresponding to the four cases listed in Table 1. The working gas consisted of Ar at a flow rate of 4 LPM.

Figure 5. Variation of the discharge current $I_d$ as a function of the main dc supply voltage $U_M=U_1$ for: (1—blue) $U_{GK}$ = −2.9 V; (2—green) $U_{GK}$ = −2.7 V; (3—red ) $U_{GK}$ = −2.5 V; and (4—black) without CSD. The $U_M$ voltage step ≅ 30 V. The experimental setup shown in Figure 1 was used. The working gas consisted of Ar at a flow rate of 4 LPM.

Figure 5. Variation of the discharge current $I_d$ as a function of the main dc supply voltage $U_M=U_1$ for: (1—blue) $U_{GK}$ = −2.9 V; (2—green) $U_{GK}$ = −2.7 V; (3—red ) $U_{GK}$ = −2.5 V; and (4—black) without CSD. The $U_M$ voltage step ≅ 30 V. The experimental setup shown in Figure 1 was used. The working gas consisted of Ar at a flow rate of 4 LPM.

During the experiments, a stable cold (non-thermal) plasma jet with a length of about 4 mm was obtained, as shown in Figure 6. To observe the variation of the gas temperature T as a function of the discharge current $I_d$, a K-type thermocouple was placed about 3 mm from the plasma jet output hole. The measured values are shown in

Table 2. Average discharge current $I_d$, discharge voltage |$U_d$|, threshold voltage |$U_{th}$|, and gas temperature T with the electrical supply system operating in the current source regime.

${\mathit{U}}_{\mathit{GK}}$ [V]	${\mathit{I}}_{\mathit{d}}$ [mA]	|${\mathit{U}}_{\mathit{d}}$| [V]	|${\mathit{U}}_{\mathit{th}}$| [V]	T [°C]
−2.9	10	252	362	54
−2.7	13	248	391	60
−2.5	16	245	421	72

.

Figure 6. Image of the plasma jet. The working gas consisted of Ar at flow rate of 4 LPM.

Figure 6. Image of the plasma jet. The working gas consisted of Ar at flow rate of 4 LPM.

To assay the long-term course of the experimental circuit, the variation of the discharge voltage and current was monitored for 1 h. The discharge was previously ignited and maintained for 10 min in order for the CSD to reach thermal equilibrium with the ambient environment. The measurement results are shown in Figure 7.

Figure 7. Time variation of the discharge voltage $U_d$ (top) and discharge current $I_d$ (bottom). Time step ≅ 5 min, $U_M$ = −450 V, and $U_{GK}$ = −2.7 V. The working gas consisted of Ar at a flow rate of 3 LPM.

Figure 7. Time variation of the discharge voltage $U_d$ (top) and discharge current $I_d$ (bottom). Time step ≅ 5 min, $U_M$ = −450 V, and $U_{GK}$ = −2.7 V. The working gas consisted of Ar at a flow rate of 3 LPM.

4. Discussion

The experimental results demonstrate that the topology taken into account for the electrical supply system works as expected. It is able to drive a cold plasma jet generating device while keeping the discharge current approximately constant at a programmed value. As can be seen in Figure 5, if the main dc supply voltage magnitude |$U_M$| exceeds a threshold value |$U_{th}$|, then the discharge current $I_d$ is roughly constant. As a result, beyond this threshold voltage, the electrical supply system operates in a current source regime. A minimum value of the main dc supply voltage magnitude is necessary to compensate for the voltage drop across ballast resistor (actually $R_B+R_S$) and discharge voltage |$U_d$|, which remains approximately constant for a given discharge current $I_d$. Therefore, this threshold voltage can be estimated as |$U_{th}$| = $|U_d|+(R_B+R_S)I_d$. Table 2 shows the average values for the discharge current $I_d$, discharge voltage $U_d$, threshold voltage |$U_{th}$|, and gas temperature T when the electrical supply system operates in the current source regime. It should be taken into account that the temperature value depends on the thermocouple position relative to the plasma jet, which is tabulated for the purpose of comparing the three cases.

Table 3. Maximum voltage drop across anode–cathode terminals $U_{AKmax}$ and dissipated power $P_{AKmax}$ for the CSD, recorded during measurements, corresponding to the considered experimental conditions.

${\mathit{U}}_{\mathit{GK}}$ [V]	${\mathit{U}}_{\mathit{AKmax}}$ [V]	${\mathit{P}}_{\mathit{AKmax}}$ [W]
−2.9	234	2.37
−2.7	212	2.98
−2.5	168	2.70

lists the maximum anode–cathode $U_{AK}$ voltage, calculated as $U_{AK}=|U_2-U_3|$, and dissipated power $P_{AK}$, calculated as $P_{AKmax}=U_{AKmax}\times I_{dmax}$, for the CSD, corresponding to the considered experimental conditions.

In the case of an IXYS IXCP 10M45S device, the maximum ratings values for the voltage between the anode (A) and cathode (K) is 450 V, whereas the maximum allowed dissipated power is 40 W [22]. To avoid overheating, the CSD was mounted on an aluminium heatsink. Therefore, by means of proper design of the electric circuit, both the discharge stability and running of the semiconductor component within the safe operational area can be achieved. This type of CSD may also operate in a current source regime using another configuration presented in Figure 8. The anode–cathode current is programmed by means of an external resistor $R_K$ connected between the cathode and gate terminals of the CSD. The datasheet indicates 10 mA as a typical value for $R_K$ = 300 Ω. In fact, the same principle is used as in the case of the circuit presented in Figure 1. The negative voltage applied between the cathode and gate terminals is due to the voltage drop across resistor $R_K$ as a result of the discharge current $I_d$.

Figure 8. Another possible configuration in which the electrical circuit operates as a current source. The anode–cathode current is controlled by the external resistor $R_K$.

Figure 8. Another possible configuration in which the electrical circuit operates as a current source. The anode–cathode current is controlled by the external resistor $R_K$.

The described solution based on inserting the CSD in circuit, as shown in Figure 1, allows for the creation of a system that generates plasma operating at a constant discharge current while being adjustable within a certain range. Although the CSD is inserted on the high-voltage side of the electrical supply circuit, the current discharge can be adjusted at any time by the user in a simple and safe manner, since the external voltage control can be applied through a circuit that is galvanically isolated. The variation of the discharge current $I_d$ can be achieved in this way without changing the ballast resistance, thereby preserving the electrical discharge stability. The described system does not contain processors or other digital signal processing elements. This affords it a high degree of immunity against sources that generate EMI and better reliability. As seen in Figure 7, under conditions where the discharge voltage $U_d$ gradually decreases, the discharge current $I_d$ exhibits variations with an amplitude of approximately ±0.5 mA in the long term. This variation can be considered acceptable in many applications. As mentioned above, the CSD acts as a basic current regulator without any feedback. Any variation of the chip temperature, for example, leads to variation of the anode–cathode current. Based on the fact that the discharge current $I_d$ can be adjusted via the voltage $U_{GK}$, better control of the discharge current can be obtained by including the CSD in a negative feedback loop by comparing the actual current value with a reference value. If the goal is to create a system that generates plasma while operating at a constant and predetermined discharge current that never changes, then the solution presented in Figure 8 is more practical because of its simplicity. An issue that should be noted is that the discharge current is extremely sensitive to variation of the external voltage. Small variations of $U_{GK}$ can lead to significant variations of the discharge current $I_d$.

In conclusion, the described technical solution meets the work aim and the results represent an experimental proof-of-concept (TRL3 level).

In Table 2, it can be seen that the bulk gas temperature T can be controlled by means of the discharge current $I_d$ and implicitly by the voltage $U_{GK}$. The next stage of development will consider including a paired CSD/temperature sensor in a negative feedback control loop, with the aim of maintaining the gas temperature at a constant and preset value.

5. Patents

An electrical supply circuit including a CSD is described in the patent application Compact dc power supply module for cold plasma generation, filed with State Office for Inventions and Trademarks Romania (OSIM) and published under number RO138932 A2/2025.