Induction motor control using static power frequency changer

Abstract

When I originally decided to use a cycloconverter for induction-motor speed control, I was unaware of the fact that the more properly named ‘naturally commutated cycloconverter’ was one of a wider field of ac-to-ac converters. Books which treat power electronics in general, usually, name only the cycloconverter with the phase shifting control method, and ignore the rest. As was seen though the cycloconverter is but one of the established ac-to-ac converters which are now becoming more popular and useful. The reason that this class of converters are still laboratory specimen, is that the operating principles, performance characteristics and application advantages of such systems have been established only fairly recently, but once this is overcome, frequency changers have may potential applications which are only limited by the designers imagination. At present, the most intersting applications of frequency changers are variable speed control, variable-speed-constant-frequency supplies for aircraft applications, controllable VAR generator for power factor correction, and ac system interties where two independent power systems are linked together, and power flow in between them is controllable (notice that the frequencies of the two systems can either be equal or different). Seeing that this was a vast interesting field I thought it worthwile to, instead ofanalysing the NCC alone, analyse the basics common to all systems then study a typical application. The cycloconverter applied to variable speed systems is nowadays an established application in traction purposes, especially on heavy vehicles like trians and mineral ore transporters, as well as in mill and heavy drilling applications. This system has one major disadvange, taht of offering only a limited range of control, usally a maximum of 75% the supply frequency, but this range is generally acceptable for most applications. The unrestricted frequency changers is the only system where the speed range is not limited. The cyclocnverter system is though very efficient indeed, extracts fro mthe IEEE transactions on ‘Power Semconductor Applications’ quote some large systems having an efficinecy of over 99%. Naturally, when large systems are used filers are an additional precaution to avoid harmonics imprinting on the supply lines. Harmonics on the supply are the major (though very much reduced) drawback of all power swtiching systems, because they might affect the shape factor of the sinusoidal waveform and change it say from 1.11 to 1.08 which is of great detriment to other systems connected onto the same lines and have transfomers on their inputs. Most cycloconverter systems presented to the IEEE are controlled by the phase-shifting method described in chapter 5. The system using frequency modulation gives better results than that using phase-shifting but the latter method is more widely used, and I can understand this perfectly well, for the modulation is the crucial part of the system and is very difficult to design. I was maybe a bit ambitious and tried the second unpopular alternative but failed to design the required modulator and the resulting switching instants were unsatisfactory. The Armstrong modulator was an inadequate method to obtain the firing instants for the power switches. This is because the modulator used was primarily a modulator used for amplitude modulation, and with the surrounding alterations, the resulting system was only applicable to relatively low signal frequency modulation. The same system could be used for wider applications if both carrier and signal are multiplied by some convenient factor. Otherwise, a varactor diode modulator must be used (not available). Unfortunately, the writing was already done when the following, much more simpler methd was conceived. The same effect of amplitude dependent frequency modulation required to extract the timing instants, is obtained from the outputs of a multiphase transformer at whose star point the reference required output voltage waveform was applied. The following drawings give a visual account of the principles involved. The first drawing shows the three-phase supply waveforms and the required sequence of timing pulses for a zero d.c. mean output. The second diagram shows the required timing instants for a sinusoidal mean output, and the deviations with respect to the quiescent, of the timing pulses. Fig. 2 also shows the actual output waveform for the theoretical timing instants obtained from the already met cosine crossing scheme (i.e. Timing for phase A is done with phase B). The last drawing shows graphically how by using the transformer method, the required exact firing instants are derived, for a single existence function, by detecting the zero crossings appropriate to the specific case. It shoud be noted that the case show is specific for a three-pulse scheme producing a positive output voltage waveform. For a negative output waveform, it is necessary to invert the action of the zero crossing detector from one of negative going switching, for a positive type out put to a positive-going switching for a negative type output. The circuit built is for the production of a positive type output with the zero-crossing detector transfer characteristic inverted because there is a 180° shift later on. For a six-pulse converter, the system will consist of a six phase mains synchronised star connected transformer instead of the three-phase one, but the same principles for cosine timing apply (i.e. phase A to time B, B to time C, C for D etc.). The other difference is in the sampling procedure, where in an attempt to economise on the circuit elements, small sampling times are introduced by inserting electronic switches on the timing waves which are on or off according to a pre-established manner, as shown previously in the last chapter, from the shift-register’s output. The same distribution of pulses is then followed as seen before. A three-pulse system was built for demonstrating the principles, but the output power waveform was not used as it is highly unrecommendable to use a three-pulse output on a motor.