Telephone Line Transpositions

The circuit of Fig. 1 has been redrawn in Fig. 9 with a transposition at the center of the telephone line; this balances the telephone line with respect to the power wire. Voltages will still be induced in each telephone wire by the varying magnetic field; these are indicated by e₁, e₁', and e₂ and e₂' of Fig. 9. Since the two telephone wires now^r occupy the same relative positions, e₁ + e₁' = e₂ + e₂', and, since these sums are substantially equal and the voltages are opposite, they will almost neutralize each other. Telephone line transpositions therefore tend to equalize the voltages induced in series in each line wire by magnetic coupling. This often greatly reduces the noise. The effect of transpositions on induction from power-line voltages through the electric field will now be considered. As shown in Figs. 4, 5, and the accompanying discussions, this field tends to raise the telephone wires to unequal voltages above ground.

Figure 9. A transposition in the telephone line balances it with respect to the power line. With this arrangement, e1 + e1' = e2 + e2', and negligible resultant voltage will exist to force a noise-producing current through the connected telephone sets.

With the circuit transposed, conditions change from those of Fig. 5 to those of Fig. 10. In this figure, voltage E_ffi = E_ff->' and E_ff2 = E_ffi, but the two former exceed the two latter in magnitude. These voltages at a given instant are assumed to act in the direction shown. The equivalent circuit is also shown in Fig. 10, and from this the effect of transpositions can be explained.

Analyzing the equivalent circuit, it will be found that the voltage E_g2'-E_g1' will tend to send a noise current I' up through the telephone at the right, and that the voltage E_g1-E_g2 tends to send a current I down through the same telephone set. This current I will be less than I' because of the series impedances in the line wires. Hence, a resultant noise current will flow up through the telephone set at the right. Thus, a difference of potential, determined by the magnitude of the line series impedances (other facts being comparable), exists between the ends of the two wires at the right (and at the left as well).

If the transposition sections are shortened, the series line impedance per section will be reduced, and the difference of potential across the telephone sets will be lessened. Although the transpositions reduce the noise-producing voltage between the wires as just explained, they do not appreciably affect the voltages to ground which may, therefore, act through unbalances to cause noise.

Figure 10. Actual and equivalent circuits showing how telephone line transpositions reduce noise caused by the telephone line wires being raised to unequal voltages above ground.

In both the explanations just given, it was pointed out that transpositions tend to reduce noise. It has sometimes been thought that one transposition located in the center of a disturbed circuit of any length will reduce the noise to a very low level (In fact, such an arrangement was tried without success on the early New York-Philadelphia circuits.)

One transposition is not effective for several other reasons. The first is that because of line attenuation the magnitudes of the current and voltage are not exactly the same at each point along the power wires. This will prevent complete equalization of the voltages. Although this effect may be negligible for many power systems, it exists nevertheless. It is especially important when considering crosstalk from telephone circuits which have comparatively high attenuation.

The second important reason that very long transposition sections are not effective is due to phase relations. As was shown in Chapter 6, a finite time is required for the propagation of current and voltage impulses, and, although these might have the same magnitude (neglecting attenuation) in different parts of a circuit, they would have different phase relations. It is apparent that, if the phase relations are greatly different in two sections of a line, it would be possible to have the induced voltages tending to add. Phase relations are particularly important for the higher frequencies (shorter wavelengths) encountered in communication circuits; it also becomes of increasing importance for the higher power-line harmonics which are in reality the components causing noise.

The practical solution is to have the telephone transpositions (and also the power transpositions as will be shown later) close together, so that the distance between transpositions is but a small fraction of a wavelength of the impulses in the disturbing circuit. Also, keeping the wires of the disturbing circuit close together will tend to reduce the stray fields and the influence factor, and, similarly, keeping the wires of the disturbed circuit close together will tend to reduce the susceptiveness factor.