SCHEMES OF VALVE-MECHANICAL COMMUTATION OF DIRECT-CURRENT MACHINES

O. A. Bratkovskii, V. V. Ivashin, I. A. Miloradov

(Presented by the scientific seminar of the Departments of Electrical Machines and General Electrical Engineering)

Among the diversity of electrical machines, direct-current machines constitute a significant group. High regulating qualities, a large starting torque, and other positive properties place them among the best for adjustable electric drives.

However, practice shows that in a number of cases, especially for machines operating under severe conditions (a wide variation in rotational speed, considerable overloads, a pulsed character of the load, etc.), insufficient commutational stability prevents the full use of these advantages (by commutational stability is meant the ability of machines to operate without sparking in various modes). This explains the heightened interest of individual researchers and large teams in the problems of commutation.

Work is being carried out not only on the study and improvement of ordinary commutation, effected by the mechanical commutator–brush-collector unit, but also on the development and study of new principles of machine commutation using semiconductor devices.

At the present time, three types of commutation of direct-current machines can be distinguished:

1) ordinary, or mechanical,
2) valve, or electronic,
3) valve-mechanical.

The fields of application of these types of commutation are determined by their advantages and disadvantages and depend on the specific requirements imposed on the machine.

With ordinary commutation, sparkless operation of the machine is ensured by the brush contact, which is capable of withstanding certain deviations of the commutation process from the optimum; these deviations are inevitably caused by nonidentities of commutation cycles of mechanical and electromagnetic nature [1]. The general disadvantage of this type of commutation is the limited commutating properties of the brush contact. It is quite difficult, and sometimes impossible, to ensure sparkless operation solely at the expense of the commutating properties of the brushes.

Electronic commutation [2] eliminates the main drawback of ordinary commutation—insufficient commutational stability. In those cases where special characteristics, a long service life, or the inadmissibility of a sliding contact for any reason are required of the machine, electronic commutation is preferable.

At the same time, electronic commutation has a number of disadvantages: complexity in controlling the semiconductor devices, a large number of elements in the electronic commutator, its large weight and dimensions in comparison with a mechanical commutator, increased cost, insufficient overload capacity, etc.

Attempts to eliminate these shortcomings have led to the use of mixed, valve-mechanical commutation (VMC), which successfully combines some advantages of the mechanical commutator (simplicity, small dimensions and weight, increased overload capacity, etc.) with the merits of semiconductor devices (SD), which are ideal switches. In VMC, in most cases, the collector and brushes retain only the function of conducting current, while the function of sparkless switching-off (or switching-on) of the sections being commutated is performed by the SD.

The first ideas for applying VMC in direct-current machines date to 1930–1940; however, because of the low reliability of the mercury valves available at that time, VMC did not become widespread. Only the development of semiconductor technology laid a real foundation for its application and made the study of VMC relevant.

At present a fairly large number of VMC circuits have been proposed. Experimental verification of some of them has shown that VMC substantially increases commutational stability, makes it possible to increase the commutating emf and thereby the utilization of direct-current machines.

The present article gives an analysis of known VMC circuits, considers their advantages and disadvantages, and selects optimal circuits for direct-current machines with additional poles.

According to the design of the brush-collector unit and the number of SDs, VMC circuits of direct-current machines can be divided into two groups:

the first group—with the usual design of the collector and brushes and with a number of SDs proportional to the number of collector plates [3, 4];

the second group—with subdivided brushes and collector plates and with a number of SDs proportional to the number of poles of the machine [5–13].

In circuits of the first group, connection of the SD is possible either in parallel with the section being commutated (between collector plates) [3, 4], or in series (in the “cut” of the loop). In the first case the SDs shunt the section at the moment when the residual current is interrupted by the brush. This current is intercepted in the SD, and thus sparking is eliminated.

In the second case the SDs switch off the current in the collector plate before it is interrupted by the brush, which likewise eliminates sparking. Obviously, with such methods of connection the number of SDs placed on the armature is proportional to the number of collector plates.

The advantages of VMC circuits of the first group include loading the SDs with a part of the brush rocker current and the possibility of “extinguishing” residual currents of different polarity.

An obvious and substantial disadvantage of these circuits is the large number of SDs and the fact that they must be installed on the rotating part of the machine. It should also be noted that the utilization of the SDs is low, since they operate only when the sections pass through the commutation zone. The rest of the time (when the sections move under the main poles) their presence is undesirable.

The most probable field of application of VMC circuits of the first group may be machines with a small number of collector plates [3].

In VMC circuits of the second group (in Fig. 1 they are shown schematically), the brushes and collector plates are subdivided in such a way that the SDs installed on the stationary part of the machine can alternately

be connected into the circuits of the sections being commutated in order to switch off the current in the commutator plates before the brushes leave them. The possibility of subdivision is achieved by a certain reduction in the number of armature sections and an increase in the number of turns in them. However, this is permissible, since with VMC an increase in the number of turns in a section does not cause deterioration of commutation, as it does in conventional machines [8–13].

The design features of the circuits make it possible to connect the CV into the circuits of the commutated sections for the entire period during which they are short-circuited by the subdivided brush (Fig. 1, a, b, c) and for part of it (Fig. 1, d, e). In the first case the CV will be loaded by the full current of the brush bracket, and in the second by part of it.

VMC circuits of the second group (Fig. 1) are free of the shortcomings inherent in the circuits of the first group. Owing to the subdivision of the brushes and commutator plates, the number of CV is substantially reduced, and their utilization is improved. Such VMC circuits are sufficiently simple and are the most acceptable for practical application. Let us consider them in more detail.

Fig. 1. VMC circuits with subdivided brushes and commutator plates

Let us compare the VMC circuits of the second group by the width of the sparkless zones (according to V. T. Kas’yanov) for the trailing edge of the brushes. We shall assume that reverse current of the valves is absent and that an infinitesimally small residual current causes sparking of the brushes.

In this case VMC circuits (Fig. 1, d, e) with the CV connected for part of the period during which the section is short-circuited by the brush have two boundaries of the zone [8], as shown in Fig. 2, a. At the upper boundary of the zone, the trailing edge sparks.

Fig. 2. Sparkless zone for the running edge of the brushes: a) for the circuits in Fig. 1, g, d; b) for the circuits in Fig. 1, a; c) for the circuits in Fig. 1, b, c; \(I_{\mathrm{p}}\) — field current; \(I_{\mathrm{a}}\) — armature current.

brushes \(A\), and on the lower one, brushes \(B\). As the duration of switching-on of the auxiliary poles into the section circuit increases, the width of the zone will increase owing to an increase in the slope angle of its upper boundary. In the limiting case, when the auxiliary poles are switched on for the entire period during which the section is short-circuited (the circuits in Fig. 1, a, b, c), the upper boundary of the zone “disappears,” and the zone assumes the form shown in Fig. 2, b (for the circuit of Fig. 1, a) and in Fig. 2, c (for the circuits of Fig. 1, b, c). The sharp increase in the width of the zone in Fig. 2, b under negative excitation and zero armature current (shown conventionally in the drawing) is explained by the presence of rectifiers in the circuit of the commutated section. VMK circuits with the auxiliary poles switched on for the entire period during which the section is short-circuited have wider sparkless zones, which is their advantage over circuits with the auxiliary poles switched on for part of the period.

From the standpoint of the influence of the commutating properties of the brushes on the change in current in the parts of the subdivided brush \(A\) and \(B\) (Fig. 1), all circuits

are divided into two subgroups. In one of them (Fig. 1, a) the transition of current from part \(A\) to part \(B\) of the brush and back occurs mainly under the action of the commutating emf from the auxiliary poles. The commutating properties of the brushes are in fact not needed for operation in the nominal mode. Their presence in the brushes only increases the width of the zone.

In the other subgroup (Fig. 1, b, c, d, e) the circuits are made with a smaller number of AP and, in individual cases, collectors; however, the transition of current from part \(A\) to part \(B\) of the brush can be effected only owing to the commutating properties of brush \(A\). For this, the voltage drop under brush \(A\) must be greater than the sum of the voltage drops under brush \(B\) and across the valve, which is achieved by selecting the characteristics of the brushes and valves. This is the disadvantage of the VMC circuits of Fig. 1, b, c, d, e in comparison with the circuits of Fig. 1, a.

Another disadvantage of them is that the difference between the voltage drops under brush \(A\) and brush \(B\) with the valve, which is necessary for normal commutation of current from brush \(A\) to brush \(B\), has a retarding effect on the commutation of current in the sections. This influence is especially strong at low machine speed, which leads to sparking during starting. To improve commutation in these circuits during starting, special means are necessary.

The VMC circuit of Fig. 1, a is free of the indicated disadvantage, since the voltage produced by the brushes during starting accelerates commutation and the starting process is facilitated.

It is known that sparking of the leading edge of the brushes may also be harmful to the sliding contact and limits the sparkless operating zone of the machine. The structural features of the VMC circuits open up sufficiently simple possibilities for eliminating it. Let us consider, for example, the method set forth in [10, 11], which consists in the use of controlled valves (thyristors), which open only after reliable contact has been established between the brush connected with them and the leading collector plate.

To implement it in the circuit of Fig. 1, a, it is necessary to replace the uncontrolled valves by thyristors; in the circuits of Fig. 1, b, c, to include them in the circuit of brushes \(A\); and in the circuits of Fig. 1, d, e, a third “leading” brush with a thyristor is necessary. The number of AP in the circuit of Fig. 1, a remains the same, while in the others it doubles, which eliminates their advantage of having a smaller number of AP. For normal operation of the circuits in this case, special sensors are necessary to control the operation of the thyristors.

A structural feature of all VMC circuits of the second group is the presence of idle and working collector plates made of one material (copper). Analysis of such VMC circuits (Fig. 1) makes it possible to formulate the following rules for their construction:

1. Each of the brushes \(A\) and \(B\) (Fig. 1, a, b, c) or brush \(B\) (Fig. 1, d, e) must not short-circuit the sections being commutated.

2. An idle collector plate must not short-circuit the divided brushes \(A\) and \(B\) with each other.

Fulfillment of the latter requirement in the circuits of Fig. 1, a, b, d is ensured by the presence of a doubled collector, while in circuits with one collector it is necessary to divide the idle plate (Fig. 1, c) or to choose the corresponding distance between the divided parts of the brush (Fig. 1, d).

Conclusions

1. The VMC circuits of the second group with subdivided brushes and commutator segments, and with a number of SCs proportional to the number of poles, have the smallest number of SCs, make better use of them, are sufficiently simple, and are the most acceptable for application in direct-current machines.

2. The VMC circuits of the second group are divided into circuits with the SCs connected into the sections for the entire period of their short-circuiting by the subdivided brushes (with the SCs loaded by the full current of the brush set), and circuits with the SCs connected into the sections for part of the period of their short-circuiting (with the SCs loaded by part of the current of the brush set).

3. Among the VMC circuits with the SCs connected into the sections for the entire period of their short-circuiting (with the SCs loaded by the full current of the brush set), the most reliable in operation is the VMC circuit of Fig. 1, a. In it, commutation in the nominal mode is practically independent of the properties of the brush contact, and in the starting mode the conditions of its occurrence are considerably facilitated in comparison with other circuits.

References

1. M. F. Karasev. Commutation of commutator direct-current machines, Gosenergoizdat, 1961.

2. I. E. Ovchinnikov, N. I. Lebedev. Contactless direct-current motors for automatic devices, “Nauka,” 1966.

3. Wisotzky W. Funkentlöschung mit Zenerdioden, 1960, Bd. 9, H. 11, Elektronik.

4. Priess A. Schaltung zur Vermeidung von Funkenbildung bei elektrischen Kommutatormaschinen. FRG patent, No. 1216418, cl. 21 d¹, 1965.

5. Dalen S. C. Commutating device. German patent No. 655112, cl. 21 d¹, 1933.

6. N. S. Klimov. Device for improving commutation in electrical machines. USSR author’s certificate No. 58081, cl. 21 d¹, 1938.

7. G. T. Vasil’ev. Electrical machine. USSR author’s certificate No. 79604, cl. 21 d¹, 1939.

8. V. V. Ivashin, I. A. Miloradov. Investigation of sparkless-operation zones of a direct-current motor with brush-valve commutation. Izvestiya TPI, No. 172, 1967.

9. N. S. Klimov. On improving commutation conditions in direct-current machines. “Elektrichestvo,” 1959, No. 8.

10. Bates I., Sridhar T. Dynamo electric machine. English patent No. 1079792, cl. H2A, 1966.

11. Bates I., Sridhar T. Thyristor-assisted sliding contact commutation, Proc. Inst. El. Engrs., 113, No. 2, 1966.

12. V. V. Ivashin, I. A. Miloradov. On valve-mechanical commutation of direct-current machines. Izvestiya TPI, vol. 160, 1966.

13. O. A. Fratkovskii, V. V. Ivashin, V. V. Miloradov. Experimental verification of a direct-current machine with valve-mechanical commutation (see the present collection).