STABILIZED VOLTAGE CONVERTER OF A STORAGE BATTERY INTO A SERIES OF DC VOLTAGES

O. S. VADUTOV, I. A. GONCHAR, Yu. S. MELNIKOV,
V. I. STEPANOV

(Presented by the scientific seminar of the Department of Automation and Telemechanics)

For supplying electronic equipment, the need often arises to obtain a series of stabilized direct voltages of different magnitudes. A storage battery is used as the primary power source. Thus, the problem arises of creating a stabilized multichannel voltage converter. In solving this problem, the following basic requirements must be taken into account: high efficiency, reliability, small dimensions, weight, and cost.

A number of works by Soviet and foreign authors [1, 2, 3, 4] consider particular questions of creating voltage converters, usually connected with their specific implementation; critical analyses of various principles for constructing converters are given. Questions of the analysis and development of multichannel stabilized voltage converters are practically not covered in the literature, although in some works, for example in [3, 5], circuits are presented for multichannel converters consisting of a combination of independently operating single-channel converters.

Let us briefly consider the main functional units and block diagrams of single-channel and multichannel converters and possible methods for stabilizing their output voltages.

Fig. 1. Block diagram of an unstabilized single-channel DC transformer: PS—power source; I—inverter; Tr—transformer; R—rectifier; F—filter; L—load

Figure 1 shows the block diagram of an unstabilized single-channel direct-voltage converter. A device that links two DC systems with different voltage magnitudes is commonly called a direct-current transformer (DCT). We shall use this term below.

The DCT circuit (Fig. 1) contains two basic converters connected in series: an inverter and a rectifier. An intermediate

the AC circuit of the TVC makes it possible to have an increased operating frequency, which leads to a reduction in the dimensions and weight of the converters.

As inverters in a TVC, self-oscillators without power amplification and with power amplification may be used. The latter have become most widespread.

In the present article a circuit with a two-stage inverter (a self-oscillator with power amplification) will be considered.

Stabilization of the output voltage of a TVC can be carried out in two ways: parametric (without feedback) and compensating (with feedback). Parametric stabilization, despite its simplicity of implementation, has a number of significant shortcomings (unregulated output voltage, its strong dependence on the parameters of the elements, etc.). In this connection, in a TVC—especially in a multichannel version—the use of the parametric method of stabilization is undesirable.

Depending on the point at which feedback is introduced in a TVC, the following ways of stabilizing the output voltage may be distinguished:

1 — acting on the inverter supply voltage;

2 — acting on the output rectified voltage of the TVC (stabilization of the output voltage);

3 — acting on the shape of the output pulses of the TVC inverter;

4 — acting on the unrectified output voltage of the inverter (control of the rectifier).

As is known, the amplitude of the inverter output AC voltage depends substantially on the supply voltage. Consequently, if the corresponding action on the supply voltage is applied, the output voltage of the TVC can be stabilized. Two cases are possible:

a) stabilization of the inverter supply voltage (Fig. 2,a);

b) variation of the inverter supply voltage as a function of the output voltage of the TVC (Fig. 2,b).

A stabilizer of the first type eliminates the effect of changes in the supply voltage on the output voltage of the TVC. In this case, however, the elements of the TVC are not covered by feedback, and the output voltage will obviously depend on changes in the load parameters and on changes in the circuit parameters. Consequently, the stabilization method under consideration can be used only when the load and the parameters of the TVC elements vary insignificantly.

In the TVC (Fig. 2,b), the inverter supply voltage is changed so that the output voltage remains unchanged. In this case the feedback encompasses all the elements of the TVC; consequently, changes in the parameters of the load and of the elements have less influence on the output voltage.

To stabilize the output voltage of the TVC, a stabilizer may be inserted after the rectifier or filter (Fig. 2,c). The instability of the output voltage will be determined only by the parameters of the stabilizer. This stabilization method is recommended for use when the TVC output voltage is less than 50–70 V.

With any method of voltage stabilization, power amounting to 10–30% of the useful power is dissipated in the stabilizer; i.e., when a stabilizer is present, the overall efficiency of the converter decreases noticeably, and this circumstance must be taken into account when incorporating additional stabilizers into the block diagram of the converter.

Therefore, in our opinion, the most promising is stabilization of the output voltage of the TVC by acting on the parameters of the output

Fig. 2. Block diagrams of stabilized single-channel TPTs; RE—regulating element; UE—control element; ME—measuring element; RVS—reference-voltage source; the remaining designations are the same as in Fig. 1.

... of the inverter pulses: duty factor or frequency (Fig. 2c). This method makes it possible to combine inversion and voltage stabilization, thereby reducing the number of elements in the circuit and increasing reliability and efficiency.

In principle, regulation of the output voltage of a TPT is possible at the rectification stage by means of controlled valves. The use of silicon controlled valves often proves inadvisable because of the low power of the converter. In addition, the need to include two controlled elements (in a full-wave rectification circuit) inevitably leads to an increase in the total number of circuit elements. This drawback is especially pronounced in multichannel TPTs.

Let us turn to questions of constructing multichannel TPTs. The simplest method of construction is, obviously, to implement each channel separately. The channels are connected with one another only through the common power source. The disadvantage of this method lies in the presence of a large number of identical elements that perform the same functions.

It is expedient to construct a multichannel TPT with a common inverter and transformer. The required conversion ratios in the individual channels can be obtained by means of the transformation ratios between the secondary and primary windings of the output transformer.

Fig. 3. Block diagram of a stabilized multichannel TPT

The most obvious solution to the problem of stabilizing multichannel TPTs is to include voltage regulators at the output of each channel. However, a large number of regulating elements leads to a considerable reduction in the overall efficiency. The need for additional filters leads to an increase in the dimensions of the converter, and the presence of a large number of reference-voltage sources makes temperature compensation difficult.

Therefore, for stabilizing the output voltages of multichannel TPTs it is proposed to use a single feedback circuit, whose signal is applied to the inverter, which simultaneously performs the functions of the regulating element. A block diagram of such a multichannel converter is shown in Fig. 3. The feedback signal is taken from an additional winding of the output transformer. The output circuits of each channel, including the secondary winding of the transformer, the rectifier, and the filter, are not included in the feedback loop, which leads to a certain increase in the instability of each channel.

With this method of constructing a multichannel TPT, the question arises of the influence of a change in the load in one of the channels on the output voltages in the other channels.

Suppose that the load in one of the channels increases, which leads to a decrease in the output voltage not only in this channel but also, because the power of the inverter is limited, to a decrease in the output voltage in the other channels, including in the feedback circuit. Consequently, even with a single feedback circuit, the influence of the load parameters in one of the channels on the other channels is reduced.

The advantage of this method of constructing the converter is that the number of elements required for stabilizing the TPT output voltages is minimal.

Fig. 4. Schematic diagram of a stabilized multichannel TPT

Figure 4 shows the schematic diagram of a multichannel stabilized TPT constructed according to the last block diagram (Fig. 3).

The pulse shaper [6] (triode \(T_3\), transformer \(\mathrm{Tr}_3\)) converts rectangular pulses arriving from the driving generator (triodes \(T_1, T_2\)) into bipolar rectangular pulses with a pause at zero. The pulse duration is inversely proportional to their amplitude, that is, to the supply voltage of the converter. To obtain the required TPT power, a power amplifier is used, made according to a bridge circuit on triodes \(T_4\)—\(T_7\), operating in switching mode.

The feedback signal acts on the driving generator, changing the pulse repetition frequency. When the output voltage of the TPT increases, the frequency decreases and the pause between pulses increases, since the duration of the pulses from the shaper remains unchanged. The output voltage of the converter thereby approaches the specified level.

Let us consider the operating principle of the pulse shaper. The voltage from the DG, through diodes \(D_5\) or \(D_6\) and transistor \(T_3\), is applied to the saturable transformer \(\mathrm{Tr}_3\). While the transformer is not saturated, a negative voltage is applied from the feedback windings, through diodes \(D_8\) or \(D_9\), to the base of transistor \(T_3\), keeping the transistor in the open state. When the core of transformer \(\mathrm{Tr}_3\) becomes saturated, transistor \(T_3\) is locked, since the negative voltage at its input disappears. Thus, the pulse at the output of transformer \(\mathrm{Tr}_3\) is shorter than the output pulse of the DG.

As the supply voltage increases, saturation of the core of \(\mathrm{Tr}_3\) occurs more rapidly, which leads to a shortening of the pulse duration and a corresponding increase in the pause at constant frequency. When the supply voltage decreases, the opposite picture is observed, but in any case the relation

\[ U_{\text{sh}} \tau_{\text{i}} = \mathrm{const}, \tag{1} \]

where

\(U_{\text{sh}}\) is the amplitude of the pulses at the shaper output;

\(\tau_i\) — pulse duration.

The mean value of the voltage at the amplifier output is

\[ U_{\text{out}} = 2 f U_{\text{PA}} \tau_i, \tag{2} \]

where

\(f\) — frequency of the master oscillator;
\(U_{\text{PA}}\) — voltage amplitude at the output of the power amplifier.

With constant load and temperature, it may be assumed that the voltage amplitude at the amplifier output and the voltage amplitude of the pulse shaper are directly proportional to the input voltage:

\[ \begin{aligned} U_{\text{sh}} &= a U_p;\\ U_{\text{PA}} &= b U_p, \end{aligned} \tag{3} \]

where

\(U_p\) — converter supply voltage;
\(a, b\) — constants.

The mean value of the voltage at the amplifier output, taking into account (2) and (3), is

\[ U_{\text{out}} = 2 f \frac{b}{a} U_{\text{sh}} \tau_i . \tag{4} \]

Thus, stabilization of the output voltages of the TPT when the supply voltage changes is carried out by means of the pulse shaper (parametric stabilization). Stabilization of \(U_{\text{out}}\) when the parameters of the converter output circuits change is carried out by varying the frequency of the MO through the action of the negative-feedback signal.

Fig. 5.

In the experimental study, the operating frequency of the converter was 970 Hz. Fig. 5 shows oscillograms of the voltages at the outputs of the MO, the pulse shaper, and the power amplifier. When the supply voltage was varied from 22 to 30 V and the load in all channels by 20%, the stability of the output voltage did not exceed 1.5%. The efficiency of the converter is 0.7.

The results of the experimental investigations make it possible to conclude that construction of a multichannel TPT according to the proposed block diagram is expedient. In the event that higher stability requirements are imposed on the output voltages of individual channels of the converter, the inclusion of additional stabilizers in these channels may be recommended.

REFERENCES

1. A. A. Zhuravlev, K. B. Mazel, Transistor DC Converters. “Energiya,” Moscow–Leningrad, 1964.

2. M. I. Kuzmenko, A. R. Sivakov, Semiconductor DC Converters. GEI, Moscow–Leningrad, 1961.

1. M. B. Leitman. Semiconductor voltage stabilizers for powering electronic circuits. Izv. vuzov, Elektromekhanika, No. 8, 1963.

2. V. F. Basovskii, V. N. Tubarevich. A magnetic-semiconductor converter with an adjustable output-voltage value. In: Electromechanical and Semiconductor Devices of Converter Technology. “Naukova dumka” Publishing House, Kiev, 1966.

3. Kleiner Charles T. Adjustable inverter. U.S. Patent. Class 321-18, No. 3260919.

4. V. I. Stepanov. Semiconductor converter. USSR Author’s Certificate No. 163262, class \(21d^2, 12_{03}\).