I think the rectifier transformer connections can be related to harmonic cancellation concepts.
In general, semiconductor power converters for both ac and dc conversions have long been known to produce harmonic currents that affect other electrical equipment.
Harmonic currents can create harmonic voltage distortions and somewhat increase the core loss, the exciting current, and sound levels while leading to potential core-saturation problems. Other stray losses are generally proportional to the current squared times the harmonic frequency.
Eddy-current losses in the windings are affected mostly by harmonic currents. The eddy-current loss is proportional to the square of the load current and the square of the harmonic frequency. The transformer designer must make efforts to reduce the winding eddy current losses due to the harmonic amplification of these losses.
Also I2R losses increase as the rms current of the transformer increases. A transformer with a higher harmonic spectrum will draw more current from the system.
Some harmonics can be cancelled, depending on how the windings are constructed or on the transformer circuit selected. Cancellation considerations are vital to the proper design and cooling of the transformer windings.
All of the harmonic cancellations discussed so far have used simple delta and wye windings in order to achieve them. Additional phase shifts are used to achieve higher pulse orders. This is accomplished by using extended delta windings, zigzag wye windings, and polygon windings. The windings are extended the required amount to produce the degree of shift desired. These are usually done at 7 1/2
, 10, 15, and 20 degree, although more increments are possible in order to achieve the desired converter pulse.
The type of winding combinations required for the converter pulse desired can be quite comprehensive.
A table of typical winding shifts is shown in below table, but other combinations are acceptable. The table shows the required number of secondary windings needed with a single-primary-winding transformer.
This is usually used with higher-voltage, low-current secondary windings. It can also be used as the required number of primary-winding shifts and transformers, when the transformers have the same single-secondary-winding phase relationship. This is usually used with high-current secondary windings.
Using different phase shifts on the single winding of the transformer, whether the primary or secondary winding, can increase the number of phase shifts. For instance, two 12-pulse transformers can make a 24-pulse system by using a delta primary on one transformer and a wye primary on the other. In cases where it may be desirable to have an interchangeable spare, it is sometimes beneficial to use two 15° phase-shifted primary windings. The spare transformer can then be made with a reconnectable winding for 15°shift.
It is important to note that harmonic cancellation is generally not perfect. This is due to several factors, such as unbalanced loading, inaccurate phase shifts, differences in commutating impedances, and tap changes. That may be acceptable at some times but not at others. It is common to assume a 5% residual of lower harmonics to accommodate these realities.
When the phase shift is incorporated in the primary winding, the degree of shift will vary somewhat as taps are changed on the transformer unless a tap changer is used in the main part of the shifted winding and the extended part of the shifted winding. Even then, there may be a slight shift.