# Transducer (electrical engineering)

A transducer or magnetic amplifier is an electromagnetic component for controlling alternating currents by direct currents by means of premagnetization of the magnetic core of a choke .

## Principle and structure

The mostly air gap-free magnetic core consists of a material with a strongly non-linear, low-hysteresis magnetization characteristic. For low frequencies this is an iron core, for high frequencies ferrites or nanocrystalline alloys are used. The magnetization is done by two coils, one of which carries the alternating current to be controlled. This coil is the controllable choke. The second winding carries the control direct current, but at least a current with a large direct component.

Fig. 1: Basic circuit diagram of a transducer: on the right is the direct current control circuit

The core volume in the AC circuit is premagnetized by the direct current, which varies its relative permeability. This can go as far as the saturation of part of the volume or the entire volume. In this way, the choke loses its (highly permeable) core partially or completely and thus a large part of its inductance.

DC and AC windings are often arranged so that no AC voltage is transformed into the control winding. This can be done using two similar, appropriately connected transducers or a three-legged core whose windings are connected accordingly (see section push-pull circuit ).

If AC voltage is induced in the control winding, this can be used directly for control purposes by rectifying it in such a way that a direct current is generated in the winding. Such transducer circuits can be controlled with just a variable resistor - a DC voltage source is not required.

## Operating modes

Two modes of operation are possible. In the area of ​​small alternating currents (small modulation), the magnetic characteristic curve at the operating point can be viewed as linear. In this case, the control direct current is used to shift the operating point. In this operating mode, the transducer is a coil with an electrically variable inductance on the AC side and is used, for. B. the amplitude control. The alternating current is sinusoidal.

The second operating mode (high modulation) drives the core far into saturation and is used for power control. In this case, the alternating current can deviate significantly from the sinusoidal shape.

### Low modulation (small signal operation)

The circuit shown on the right is a typical example. The inductance ( L = u 1 / d i 1 ) effective on the alternating current side of the choke is changed by changing the pre- magnetization of the associated non-linear magnetic core. For this purpose, a control direct current i 2 of variable magnitude flows through the second winding . In the area of ​​small control currents, the greatest inductance results, which decreases the more the control current shifts the operating point of the magnetic circuit in the direction of saturation. By changing the inductance, the impedance of the choke changes in the AC circuit . In this way, the power in the effective resistance R is also changed. Without control current, the power is small, for maximum control current, the performance sought in the resistance against u w / R . Due to the requirement of linearity, the power amplification of such a transducer P ( controlled ) / P ( control ) is usually below 1.

Fig. 2: Characteristic magnetization curve (greatly simplified)

### High modulation (large-signal operation)

Figure 3: Time curves of the voltages in the circuit Figure 1

Large-signal operation must be provided in order to be able to control higher voltages and powers. In order to explain the function of the circuit in this case, the simplest circuit should be used and the boundary conditions greatly simplified. The alternating current components of i 2 , which occur due to the transformed voltages and generate losses, are to be reduced to negligible values ​​by a very large inductance, so that i 2 is almost a direct current. For the sake of simplicity, the abstract magnetization characteristic shown in Figure 2 should also apply. Apart from the load resistance shown, there should also be no parasitic ohmic components.

In order to achieve periodic operation immediately, the observation begins after the highest saturation or the highest current. At point 1 of the characteristic curve, the load current i 1 is equal to zero. The magnetization takes place only through i 2 . At point 2 the core comes out of saturation and the current i 1 is negative and almost as large as - i 2 . Because of the beginning of desaturation, tension appears on the windings. It results to u 1 = u w - R · i 1 . Because of the applied voltage, the current falls slightly further into the negative. It reaches its most negative value at point 3, at which the direction of the flux change is reversed because of the zero crossing of the voltage u 1 . Due to the positive voltage, the core is magnetized again to saturation (point 2). Because of the saturation, the full generator voltage is switched to the load. This results in the flowing current i 1 = u w / R . Point 1 is exceeded and is reached again in the next current zero crossing. The process repeats itself.

## Push-pull circuit

Fig. 4: Push-pull circuit
Fig. 5: Characteristic magnetization curve

If two arrangements are connected in parallel on the AC side as shown in Figure 1 and the control current flows through the two secondary windings in different directions, the control current in the partial transducers acts out of phase. The voltages induced in the direct current circuit compensate each other. For small modulation, the currents and voltages remain almost sinusoidal. Because of the parallel connection, the possible load current doubles and because of the compensation of the induced AC voltage, the losses in the DC circuit are significantly reduced.

For large-signal operation, the processes must be considered again, since the superposition theorem does not apply due to the (magnetic) non-linearities that are used. This is based on the circuit shown in Figure 4 and the associated arrows. For the figures, point 4 of the magnetization curve in Figure 5 was selected as the effect of the control current i 2 = ⅔ I S and the number of turns ratio equal to 1.

Figure 6: Timing of currents and voltages in the circuit in Figure 4

At the origin of the magnetization characteristic, the load current is equal to zero and the currents i a and i b must be just as large as the direct control currents , but counteract them. With the same winding direction and the arrows according to Figure 4, i a = - i 2 and i b = i 2 . For the total current, i 1 = i a + i b = 0. In point 4, i a = 0 and i b = 2 · i 2 = i 1 . With the onset of saturation for winding a, the alternating voltage on both windings is zero and the current i 1 is limited by the load to which the mains voltage is applied. The winding currents are i a = i 1 + I S - i 2 and i b = - I S + i 2 . If the load current (purely ohmic) becomes zero at the zero crossing of the voltage, the relevant core comes out of saturation. The cores are magnetized back during the negative voltage half-oscillation. During this time the load current is zero and the currents i a and i b go back to the values ​​at the origin of the magnetization characteristic. Then its negative branch is run through in the same way. The essential time courses are shown in Figure 6.

In terms of losses, the push-pull circuit is significantly cheaper than the circuit according to Figure 1, because the AC voltages induced in the DC circuit compensate each other and because, as can be seen in Figures 2 and 5, a relatively smaller control current is required.

You can also use another advantage if you connect the power windings in anti-parallel and let the control current flow through the windings in the same way. In this case, the control windings can be combined into one. The structure of such an arrangement is shown schematically in Figure 7. In this case the iron body has three legs. The simplified theory does not result in any significant material savings, since in this case the control winding has to enclose twice the iron cross-section. If the iron and copper losses that occur in practice are taken into account, the material savings are greater.

Figure 7: Schematic representation of the winding arrangement of a transducer

The current and voltage curves and the control function of the push-pull circuit correspond to those of an AC power controller ( power electronics ) with phase control . Since this is cheaper, better in efficiency and also lighter and smaller, it has replaced the transducer as a control device for alternating currents.

## application

Transductors have been almost completely displaced by semiconductor technology, but there are still individual applications where a high level of robustness and reliability are important. Transductors have a long service life and, in contrast to semiconductor circuits, are hardly at risk from overvoltages and short circuits. In the past (1930s to 1960s) the line frequency transducer was a common solution for controlling, for example, cinema lighting. It does not require any moving parts like a variable transformer and is wear-free. Tube monitors and televisions partly had transducers for cushion equalization, which regulated the line amplitude. In the 1990s, in particular, switched-mode power supplies contained transductor controls on the secondary side. Also audio amplifier were performed with transducers, transducers for direct current can also be built with transducers.

Another application is in the area of power transformers for electric arc furnaces . Here transducers are used for fine control of the output current. A transducer for three-phase loads can be built from three push-pull circuits.

A special feature is the Pungs throttle (named after Leo Pungs ), which was used for amplitude modulation . It was set up as shown in Figure 7 and works at the transmission frequency. It worked directly at the transmitter output on the tuned antenna resonant circuit. The output voltage waveforms shown above cannot therefore be transferred to this application.

## literature

• Walter Schilling: transducer technology. Theory and application of controllable chokes . Publisher R. Oldenbourg, Munich 1960
• Fritz Kümmel: Control transducers: theory and applications in control engineering . Springer-Verlag; 456 pages; Berlin / Göttingen / Heidelberg 1961 - limited preview in the Google book search