BROADBAND LEVEL INDICATOR

M. S. Roitman, V. N. Studenikin, V. M. Sergeev

(Presented by the scientific and technical seminar of the Department of Radio Engineering)

To evaluate the frequency error of balanced voltage dividers (attenuation boxes), it is necessary to have a highly sensitive broadband level indicator.

Usually such instruments, used in measurement practice, are devices similar to voltmeters and have the indicator scale graduated in transmission units (nepers or decibels). As the measuring stage, a rectifier circuit on diodes is used, or a bridge circuit on thermistors.

A drawback of level indicators built with detectors is their comparatively low sensitivity (not more than 2.5–4 nepers), the influence of ambient temperature, and a number of other additional factors [1].

Level indicators in which thermistors are used, although capable of providing higher measurement accuracy, impose stringent requirements on the design of these instruments with respect to thermostating of the controlled element.

Known level indicators have a gain factor \(K \leq 30000\) units in a frequency band up to 0.6 MHz and make it possible to measure an input signal attenuated by not more than 7–8 nepers, with a high noise level. The measurement ranges are changed by means of stepped dividers using active resistances. In this connection, because of reactive conductances, a noticeable additional frequency error is observed.

In developing the level indicator described below, additional requirements were taken into account, especially the increase of the gain factor and reduction of the noise level at large attenuations (over 7 nepers), an increase in the sensitivity of the circuit, and the elimination of parasitic couplings through the power-supply circuits of the amplifier units.

The functional diagram of the level indicator is shown in Fig. 1. The input signal to be amplified passes through a balancing transformer to a smoothly adjustable low-resistance divider and then to the amplifier units, the gain factor of each of which is equal to 30. By means of the double switch \(P\), stepwise regulation of the overall gain factor of the entire amplifying path is carried out, and a constant signal level is maintained at the photoelectric converter.

The broadband level indicator has three identical amplifier stages with an overall gain

\[ K_{\text{total}} = 80000\ \text{units}. \]

Fig. 1

The schematic diagram of one decade of amplification is shown in Fig. 2. As can be seen from the circuit, the decade is built on two transistors of types P 415-B and GT 301-E with different conductivity and is a two-stage amplifier encompassed by deep negative current feedback.

Fig. 2. Decade amplifier

Deep negative feedback ensures high stability of the circuit operation without additional correcting elements, a uniform passband over a wide frequency range, and also little influence from changes in the supply-source voltage. The frequency characteristics for one and three decades of amplification are presented in Fig. 3. The gain of an individual decade, without taking into account the base divider, at medium frequencies has the following form [2]:

\[ K_0 = \frac{r_{kэ} + r_0}{r_0}, \]

where \(r_{kэ}\) is the resistance in the collector circuit of the second transistor,
\(r_0\) is the resistance in the feedback circuit.

The depth of negative feedback in the operating frequency range is approximately 4.3 nepers.

The sensitivity threshold of the readout device is to a considerable degree determined by the level of noise and interference of the first amplifier stage. To reduce their influence, the balancing transformer is placed in a double shield and the nominal operating mode for the circuit has been selected.

of the first amplifier decade. In addition, decoupling is provided in the collector power-supply circuits. These precautions ensured a noise level of the first decade not exceeding 4 μV.

Fig. 3. Frequency characteristics of the amplifiers:
a—one decade, b—three decades

The output stage (Fig. 4) of the level indicator is assembled on three transistors with an overall gain factor \(K \leqslant 3\). As the load

Fig. 4

a photoelectric transducer PT is used, comprising an incandescent lamp and a photoresistor. The magnitude of the output signal, taken from the load of the emitter follower built on transistor P605, is converted into a change in the conductance of the photoresistor. Owing to certain difficulties in constructing theoretical characteristics of photoelectric transducers, its analysis is carried out-

based on experimentally obtained curves [3]. As can be seen from Fig. 5, the ohm-ampere characteristic of the photoconverter has considerable nonlinearity and, consequently, a high transmission coefficient. Its value for the specified operating mode of the circuit is approximately equal to 3 units.

As the indicating device, a high-sensitivity microammeter with a linear scale is used, connected into one of the diagonals of a single bridge of direct current.

The equilibrium condition of the bridge measuring circuit, corresponding to the value of the photoresistance \(R_{\phi}=15\ \text{kohm}\), is maintained by changing the overall gain by switching amplifier stages for specified attenuation values of the four-terminal networks.

Fig. 5

One of the principal characteristics of the level indicator is its sensitivity, which ensures measurement of the partial error of the four-terminal networks being compared with a specified accuracy. It is necessary that a sufficiently small change in the measured quantity should cause a noticeable deviation of the null indicator from the equilibrium position.

By the relative sensitivity of the level indicator \(S_{уy}\) we shall understand the ratio of the change in the indicator reading \(d\alpha\) to the relative change in the measured quantity causing it.

Since the measured quantity is voltage, the relative sensitivity of the level indicator has the following form:

\[ S_{уy}= \frac{d\alpha}{\dfrac{dU}{U}} = \frac{d\alpha}{dU_{\text{m}}}\cdot \frac{\partial U_{\text{m}}}{\partial R_{\phi}}\cdot \frac{\partial R_{\phi}}{\partial U_{\text{out}}}\cdot U_{\text{out}} = S_{1}\cdot S_{2}\cdot S_{3}\cdot U_{\text{out}}, \]

where \(S_{1}\), \(S_{2}\), \(S_{3}\) are, respectively, the sensitivities of the galvanometer, the bridge circuit, and the photoelectric converter.

Provided that the resistance of the bridge power-supply source is very small, the sensitivity of the bridge circuit is determined by the following expression:

\[ S_{2}=E\,\frac{R_{2}}{(R_{\phi}+R_{2})^{2}}, \]

where \(R_{\phi}\) and \(R_{2}\) are the resistances of the photoelectric converter and the upper arm of the bridge.

The sensitivity of the photoelectric converter is

\[ S_{3}=\frac{dR_{\phi}}{dU_{\text{out}}} = -\frac{R_{\phi}\lvert K_{i}\rvert}{I_{\ell}}\cdot \frac{dI_{\ell}}{dU_{\text{out}}}, \]

where \(I_{\ell}=\dfrac{U_{\text{out}}}{R_{\ell}}\) is the current flowing through the measuring converter.

Transforming the expression for the current \(i_{\ell}\) and introducing the coefficient

\[ m=\frac{dU_{\ell}\cdot i_{\ell}}{di_{\ell}\cdot U_{\ell}}, \]

which characterizes the degree of nonlinearity of the filament lamp,

…of amplification, we obtain the final expression for the sensitivity of the FP:

\[ S_3=-\frac{1}{m}\left|K_i\right|\frac{R_{\phi}}{U_{\text{out}}}. \]

The relative sensitivity of the level indicator is one division per a 0.02% change in the current from the nominal value (an M-26 microammeter with a \(+50—0—50\ \mu\text{A}\) scale, \(R_{\text{pr}}=2.5\ \text{k}\Omega\), was used as the balance indicator).

REFERENCES

1. N. N. Solov’ev. Fundamentals of Measurement Technology in Wire Communications. Svyazizdat, 1957.

2. M. S. Roitman, V. M. Sergeev. Decade measuring amplifier. Collection of Papers of TIRiET (in press).

3. M. S. Roitman, E. I. Tsimbalist, B. A. Perminov, N. P. Fefelov. An incandescent lamp and a photoresistor as a controllable element of an electrical circuit. Izvestiya TPI, vol. 141, Tomsk, 1966.

ERRATA NOTED