UDC 681.327.66

CHARACTEROGRAPH

FOR STUDYING THE DYNAMIC CHARACTERISTICS OF THERMOELECTRIC DETECTORS

E. M. Belov, A. I. Kuznetsov

(Presented by the scientific seminar of the Faculty of Physics and Engineering)

The dynamic characteristic of thermoelectric detectors, which determines their inertial properties or the dynamic measurement error, is the principal parameter in monitoring nonstationary fields of electromagnetic radiation. Highly sensitive low-inertia thermoelectric detectors, in dynamic terms, approximate a first-order inertial element; therefore their inertial properties are characterized by a time constant, defined as the time during which the thermo-emf generated by them changes by a factor of \(e\) after a simultaneous stepwise change in the intensity of the electromagnetic radiation.

If such thermoelectric detectors are used to monitor an electromagnetic-radiation field that induces high radioactivity, then under laboratory conditions their identification by the time constant is considerably complicated, and rejection according to dynamic characteristics proves practically impossible.

As shown in [1], the dynamic characteristics of thermoelectric detectors can be studied by an indirect method without irradiating them, by setting up a “clean” experiment (in the sense of radiation safety), based on the Peltier effect. The Peltier effect appears when a thermoelectric detector is connected to an external DC voltage source and reduces to the establishment of a certain temperature gradient between neighboring junctions of the thermopile (as a result of additional heating of some junctions and cooling of other junctions), which is equivalent to the action of electromagnetic radiation on the detector.

The detector time constant is calculated from the recorded curve of the change over time of the thermo-emf after switching off the heating current of the thermopile; alternatively, this operation is performed automatically by clipping the output signal after it has decreased by 63% from the initial amplitude and recording the time constant by means of an electronic stopwatch or by the time marks of an electronic oscilloscope.

By varying the amplitude and duration of the heating current in a certain way, one can associate this with the simulation of the action of electromagnetic radiation of different intensity and of different quantities of working medium deposited on the working junctions of the thermopile. Then such expe-

experiments provide an understanding of the static characteristics of the detectors under study.

The indicated functions, associated with the investigation of the characteristics of thermoelectric detectors under laboratory conditions by modeling heat sources at the thermopile junctions with the aid of the Peltier effect, are assigned to a specially developed device called a characterograph.

The characterograph (Fig. 1) consists of a trigger (1), a relaxation multivibrator (2), an emitter follower (3), a power amplifier (4), a transistor switch (5), an ammeter in the circuit of the thermoelectric detector (6), and a low-frequency electronic oscilloscope (7).

Fig. 1. Block diagram of the characterograph; 1 — trigger; 2 — multivibrator; 3 — follower; 4 — amplifier; 5 — switch; 6 — ammeter

The schematic diagram of the characterograph is shown in Fig. 2. The trigger (1) is assembled on transistors \(T_1\) and \(T_2\) and performs the role of a zero device, fixing the instant of comparison of the developing thermo-emf with the set-point voltage taken from potentiometer \(R_3\). The relaxation multivibrator is assembled on transistors of different conductivity, \(T_3\) and \(T_4\), and makes it possible to form thermopile heating-current pulses of different duration, determined by the time constant \(R_9 R_{10} C_i\) \((i = 1, 2, 3, \ldots)\). The emitter follower (transistor \(T_{10}\)) serves to match the electrical circuits, and with the aid of potentiometer \(R_{22}\), located in the emitter circuit of \(T_{10}\), the amplitude of the heating current is regulated.

The power amplifier (4), assembled on transistors \(T_5\)—\(T_8\), is made according to a cascade circuit with parallel connection of the output stages and makes it possible to regulate the magnitude of the thermopile heating current within the range from 0.1 A to 4.0 A. The transistor switch (\(T_9\)) ensures connection of the thermoelectric detector to the vertical-deflection amplifier of the electron oscilloscope beam only at the instant when no heating current is flowing through the thermopile. The heating current is recorded by means of an ammeter, the measurement ranges of which are set by means of a set of shunt resistors \(R_{17}\), \(R_{18}\).

Fig. 2. Schematic diagram of the characterograph

As the recording device of the characterograph, a low-frequency electronic oscilloscope of type S1-19A is used; directly from the plates for horizontal deflection of its beam, a feedback voltage is taken, serving as the cutoff signal when the initial amplitude generated by the thermocouple is reduced by a factor of $e$.

The characterograph is powered from an alternating-current mains supply (220 V, 50 Hz) through a matching transformer. The instrument’s power-supply circuit includes rectifiers assembled with D305 diodes, capacitive smoothing filters, and a parametric voltage stabilizer (D810).

In the initial state, transistor T$_1$ of the symmetrical trigger is open, and transistor T$_2$ is closed. This state of the trigger provides the self-oscillating mode of the relaxing multivibrator (2). Transistors T$_3$, T$_4$ are initially open, and the capacitance $C_i$ $(i = 1, 2, 3,\ldots)$ is charged through resistor $R_8$ to a voltage determined by the locking of transistors T$_3$, T$_4$. Then the capacitance $C_i$ is discharged through $R_7$, $R_9$, $R_{10}$ and the internal resistance of the circuit’s power source. A decrease in the voltage across capacitance $C_i$ causes transistors T$_3$, T$_4$ to open; capacitance $C_i$ is recharged, etc.

If trigger (1) is in the second stable state (T$_1$ is closed and T$_2$ is open), then the relaxing multivibrator (2) is in the inhibited state: by the open transistor T$_2$, transistor T$_3$ is shunted and locked.

When the multivibrator operates in the self-oscillating mode, then during formation of the charging current of capacitance $C_i$, the power amplifier (T$_5$—T$_8$) is locked by a positive voltage pulse at the base of the emitter follower (T$_{10}$). No heating current flows through the thermobattery. From the moment transistors T$_3$, T$_4$ are locked, the power amplifier opens and a heating current flows through the thermobattery. The duration of the heating current is determined by the discharge time of capacitance $C_i$ to the opening voltage of transistors T$_3$, T$_4$.

When the heating current flows, the transistor switch (T$_9$) is locked by a negative potential at the emitter. When no heating current flows through the thermobattery, the transistor switch is opened by the divider current, and the thermal emf generated by the thermobattery is fed to the vertical-deflection amplifier of the low-frequency oscilloscope.

The electronic oscilloscope operates in the triggered sweep mode of the electron beam. The absence of a signal at the vertical plates corresponds to a relatively small voltage on the plates for horizontal deflection of the beam, which does not ensure a change in the state of the trigger.

Fig. 3. Dependence of the cooling rate of the thermobattery on the mass of the “hot” junctions

Fig. 4. Dependence of the cooling rate of the thermobattery on the width of the thermoelectrode strip

When a signal from the detector is applied to the vertical plates of the oscilloscope, then by the voltage from the horizontal plates the trigger

is transferred from the initial state to another stable state, in which the multivibrator is in the inhibited state.

The starting circuit (trigger–multivibrator) remains in this state until the generated thermo-emf decreases by a factor of \(e\). The threshold for switching the trigger to its initial state is adjusted by means of potentiometer \(R_3\). Switching the trigger to its initial state causes the multivibrator to operate again, and the process of heating the thermobatteries is repeated.

Experimental studies carried out with the aid of the characterograph using strip chromel–alumel thermobatteries are presented in Figs. 3–6. Fig. 3 illustrates the rate of change of the generated thermo-emf as a function of different masses of the “hot” junctions of the thermobatteries. Fig. 4 illustrates the rate of change of the generated thermo-emf as a function of the width of the thermoelectrodes of a strip thermobattery, all other conditions being unchanged. Figs. 5 and 6 respectively show the dependences of the rate and amplitude of the generated thermo-emf on the duration of the pulse and on the amplitude of the heating current of the thermobatteries, all other conditions being equal. These results qualitatively confirm the expected character of the dependences under investigation.

Fig. 5. Dependence of the amplitude of the thermo-emf and the cooling rate of the thermobattery on the duration of the heating-current pulse

Fig. 6. Dependence of the amplitude of the thermo-emf and the cooling rate of the thermobattery on the amplitude of the heating current

The measured and calculated values of the time constant of the thermobatteries agree with one another to an accuracy of no worse than \(\pm 10\%\).

References

1. E. M. Belov, A. I. Kuznetsov. Study of the dynamic properties of thermoelectric detectors by modeling heat sources with the Peltier effect. Present collection.

2. E. M. Belov, A. I. Kuznetsov. Characterograph for studying the dynamic characteristics of thermoneutron sensors. Author’s certificate No. 412538. Bulletin No. 3, 1974.