You are on page 1of 12
17 Harmonic Composition of the Current The method of analyzing a periodic function by means of Fourier's integral will be explained and applied to the alternating line currents of a rectifier connection. This leads to a general law that compre- hensively describes the harmonic composition of the various current It was explained in Chapter 16 that the alternating currents of a rectifier connection are not sinusoidal. Therefore we interpret them as combinations of a fundamental and harmonics, and we shall now learn how to determine the amplitudes of these components, In order to keep the investigation as simple as possible, the exciting current of the transformer will be disregarded throughout this chapter, and regular operation will be assumed (infinite inductance in the load circuit). Furthermore, the harmonics will be expressed in terms of the fundamental. Thus the problem consists of two parts: 1, Compute the fundamental, 2. Determine the harmo: relation to the fundamental, ‘The Fundamental For determining, the fundamental of the current let us recognize the following facts. (@ With zero phase control and instantaneous commutation the effective power delivered to the d-c circuit including the power dissipated in the resistancé ofthe rectifier connection) is equal to P,, since under these conditions By = Ey, (® Under the same conditions the effective input power also equals Pay ‘because it must be in balance with the oufput power Gncluding the losses). (© With the additional assumption that the alternating voltages are undistorted, the apparent power of the fundamental must also be 2 Harmonic Composition of the Current 293 equal to P,,, because there is no reactive power and the harmonics of the current do not contribute to the effective power as long as there are no corresponding harmonies in the voltage. @ Phase control does not change the apparent power because it does not change the waveshape or the magnitude of the alternating ccurrents, as long as the operation remains regular (no free-wheeling operation), and the magnitude of the load current remains un- changed. Hence the following relation must be true over the entire range of phase control, for 3-phase as well as single-phase connections haPy any where P, = apparent power of the fundamental, disregarding the trans- former exciting current For a 3-phase supply system there is Py BE where E, = rms value of the line-to-neutral voltage 1, = rms value of the fundamental of the line current Hence the rms value of the fundamental is defined by ‘This formula is the key for many computations.since it establishes a relation between the a-cand the d:c power. Let is again'tite the conditions under which itis correct: 1. Regular operation. Free-wheeling operation will be investigated in Chapter 20. 2. Fistantaneous commutation. A correction factor that takes the com- mutating angle into account will be derived in Chapter 21. It changes (17.2) into Pas Py C___correction factor Lefnition 3. Undistorted supply voltages (included in the general assumptions stated at the beginning of this book). 294 ALC Characteristics Example (Computing the fundamental) A rectifier system with £,, = 300 V and a load current of Jy = 4000 A is connected to a 3-phase supply line with a line-to-line voltage of 13.8 kV, Determine the rms value of the fundamental ofthe line current, disregarding the transformer exciting current. Solution ae _Colaion 172) = S0LV]_4000 [4 _ 59 fa) Vix Bemevy SS) ‘The Method of Determi jing the Harmonics Although it would be sufficient for the following investigation to present the final formulas for analyzing a nonsinusoidal current, a brief intro- duotion into the background of the method will be presented in order to deepen the understanding of the results and of the conclusions. ‘The manner in which a sinusoidal quantity varies as a function of ot is usually described by a mathematical expression of the form Slot) = esin(ot + 9) where ¢ = amplitude, @ = phase angle 7 Flot) = sin ot + b cos ot where the amplitude, is defined by Va? + 6%, and the phase angle by Yan g = dja. Ina similar form, quantities can be presented that vary with 2,3,+++, n times the angular frequency «and the general expression Slot) = ay + ay sinsot + ay sin Zot ++ ++-F a, sin neot +B, 008 wit 4B, 058 wt °F b, cos not (173) describes any combination of such quantities, including a constant portion 4,88, for example, a d-c component that is superimposed on an alternating current. Such a cbmbination is not a sinusoidal function any more, but still a periodic one, and may sometimes be presented in a completely different form as, for example, (rh for —2nf3 Sot 50 Slot)\= +l, for 0S ot S +2n/3 74) = 0 for +2n/3 Sat < +4n/3 Harmonie Composition of the Current 298 cor graphically, as shown in fig. 17.1. Both expressions (17.3) and (17.4) are just different forms for the same function, and are equal if the coeffi- cients a,,°°+, a3 b,,°**5bq3 and ay are given the right values. The only problem is to convert a presentation similar to (17.4) into the form (17.3) which is desirable for certain investigations since this form displays the harmonic composition. With a few tricks, theconversion is surprisingly a 7 LT ycte Fig. 171. Current curve analyzed in sumnple calculation. simple, and the method will now be presented. Itis based on the following statement: ‘The prodisct of two sine or cosine functions or of a sine and a cosine function with frequencies that are 1, 2, 3, , times a common basic frequency, integrated over one full eycle of this basic frequency, is zero, unless two sine or two cosine functions of the same frequency have been multiplied. This statement is illustrated for the two functions /(at) = sin wt and Flot) = sin ot and their product by fg. 17.2, and it can be seen that the ‘positive and the negative areag under the product curve are in balance. Hence the integral value over ohe full cycle is'Zro, Wliereas in fig. 17.3 a sine function has been multiplied by itself, and the integral value of the product is—in accordance with the statement just given—different from ‘zero. From this statement, which can be exactly proved for the general _¢a8e, we derive the following fundamental conclusion. If a periodic function contains a sine (or a cosine) component of @ certain order, this component discloses its magnitude if the function is. ‘multiplied by a sine (or a cosine) function of the same order and is then. integrated over one full cycle, FOR SAID COMPONENT 18 THE ONLY COM- PONENT OF ‘THE FUNCTION THAT CONTRIBUTES TO THE VALUE OF THE INTEGRAL, Thus it can be detected. F In order to develop this conclusion into a general formula, let us assume that a function f(«1) presented in any form contains among many others the component Aa 296 A-C Characteristies Ht) = sin ot ig flat) = sn 20 \ Hat) = nt sn 28 Fig. 172. ‘Two diferent sine functions and their product. and let us multiply this function by sin nox, and then compute the value of the integral over one full eycle of the basic angular frequency «, in order to determine a "Hox are nat doot =| “ay sin nest sin noot deot analog only component of) that contributes to value of integral at) = in ot J Hat) = int en wo Fig. 173, ‘Two identical sine funcijons and their product, Harmonie Composition of the Current With afew conversions, this results in auf — 4 08201] don = ay Hence the result ofthe integration is just w times a, or, seen from the other side i a= 2 [7a sin not dot (7s) To _analyzing funetion ‘A similar approach leads to the amplitudes ofthe cosine terms: 4 "Hot cos nox dor (176) ™ (analyzing function ‘Whereas ays the average value of the analyzed function 1pm =2 [render 77) With these formulas, called “Fourier’s integrals,” any periodic function can be analyzed, and with the following example such a calculation willbe practiced Example (Determining the harmonics) ‘Analyze the current shown in ig. 17.1. Solution. Since f(ot) is « different function in all three sections, and sine it is zero in one of them, we compose the value ofthe integral from two parts and compute according to relation (17.5) ° ss ayn 2P Ctosintintdors ! fg sia mat don Sth con? Si 1 ede mp te dae In order to express the harmonics in terms of the findamentl, we insert this result in elation (178): 12 a n= 022] — cosn2@ a: n= 221 — coon] as) =H for'n =1,24,5,7.8, = 0 forn=3,6.9, "+ 298 A-C Characteristics ‘There are two factors inthis result. ‘The term 1/n describes the amplitudes, while the second factor specifies the existing harmonies, No calculation is necessary for cosine components. Since the curve is symmetrical with respect to the point wr = 0, but reverses-the polarity, all harmonics must also be symmetrical with reversing polarity with respect to this point, and no cosine component would satisfy this condition, Actually, this point has intentionally been selected for wf = in order to eliminate the cosine components. Otherwise, it would have been necessary to add geometrically the amplitudes of the sine and cosine components of the same order. ‘The Harmonic Composition, General Law For the curve just analyzed, relation (17.9) states that the amplitudes of the harmonics decrease reciprocally to their order (factor 1/n), and that certain harmonics do not exist at all. In a similar manner other current curves ean be analyzed, and it will be found that for all curves with a “regular shape” (curves of regular connections with inductive load and {instantaneous commutation) the following rule applies: (@ The amplitudes of the harmonics decrease reciprocally to their order. (®) There are only harmonics of the order ng + 1 This is really a simple result, and it does not even exclude phase control since phase control does not change the waveshape of the alternating ‘currents as long as there is no free-wheeling operation. This rule will now ‘be demonstrated with an example. Xzxample (Harmonici‘of a 6-pulse current) Determine the harmonic composition of the alternating current of rectifier connection with 6-pulse performance. Solution. According to the above statement we find: fundamental ='100 % Sth harmonic = 100/5 = 20 7th harmonie-= 100/7 Ath harmonic = 100/1 13th harmonic = 17th harmonic 59% 19th harmonic = 100/19 = 5.3% 23th harmonic = 100/2: 25th harmonic = 100/25 = and so on 43% 40% iiss Harmonic Composition of the Current Complementary Currents In Chapter 16 it was explained that under certain conditions two currents are “complementary” and can be combined to form a current of twice the pulse number, From this we conclude that complementary currents have certain harmonics that cancel out if the currents are combined, because the general law states that a current with @ higher pulse number has a smaller assortment of harmonics. As an example, two complementary 3-pulsé currents must have the harmonics of the order 2, 4,8, 10, ++ with ‘equal iimplitudes but opposite polarity, because these harmonies are not included in a 6-pulse current, whereas harmonics of the order 5, 7, 11, 13, ++ must be equal in amplitude and must be “in phase,” because these components are present in a 6:pulse current and have there the same amplitudes in relation to the fundamental as they have in the individual currents. Concluding Example (Complementary 3-pulse currents and their total) ‘Compute the harmonic composition of a 6-pulse current with & = 30° in two ways: (@) Analyze this current by means of Fourier’s integral (®) Analyze the complementary 3-pulse components and superpose their harmonics Solution (jitustrated by figs. 17.4 and 17.5). The current to be analyzed is represented by the solid curve of fig. 17-4c. With the selected location of wv = 0 shown in this figure, ve curve is symifetrical with respect to ‘of = 0, Hence there are only cosine components, and the integration can bbe reduced to 4 cycle if the result is multiplied by 2. With equation (17.6) this leads to 2 [ig cos not dor +2 f° (1,05 not dat ah een = aint + sa] any nl 3 ‘With n = 1 we determine the amplitude of the fundamental a3), any sais asnd-g Kremaudines out Jo UoReMEMOD “LE BE mean 6 eos ye|z tau BPP Peel | A-C Characteristics Harmonic Composition of the Current: current of i. 17.40 autet of ‘ie 1746 a: he Fig. 17.5, Combination of two complimentary -pulse connections ‘This expression inserted in (17.10) gives the final result: pF (a7.12) factor specifying existing harmonics; $1 form = 1,7,-° O for n = 2,3, 4,6, 8,9,--- 1 for n= 5, 11,++* ‘general law defining amplitudes of harmonics ‘The fundamental and the fifth harmonic as well as their total are shown in fig. 17.4c, and it can be seen that the total, represented by the heavy broken ‘curve, approaches the real current curve, and the approximation will ‘become closer the more harmonics are included. ‘The complementary components are shown in figs. 17.4a and b, ‘These currents may be the primary line currents of two 3-pulse midpoint connec- tions operating with opposite polarity (ig. 17.5). Again, the curves are symmetrical with respect {0 cof = 0 so that it is sufficient to analyze just 302 AC Characteristics one half-cycle, For the curve of fig. 17.4a we calculate: pea : bye 2 [aj cosner dor +2 [ al 4 hein H,)oos nat dot with beS sy, (a7.13) this results in (a7.14) In a similar manner we find for the curve shown in fig. 17.46: eit = bt sin 17.15) bam ba a arts) +1 for n Oforn =I for n Where by is defined by (17.13). Comparing (17.14) and (17.15) we notice that the fiindamental and the ‘harmonics of the,order 5,7, 11, 13,-++ are equal in value and polarity ahd total up, whereas ‘the other harmonics are opposite and disappear, which accords precisely with the result computed for the total current. This i graphically presentedin fig. 17.4at the right, showing the “spectrum” of the currents. Furthermore, the total of the: fundamental and the harmonics up to the fifth order is presented by the heavy broken curve in fig. 17.4a, and it can be seen that this curve approaches very well the real current curve. rms Value of the Current We shall compute the rms values of nonsinusoidal currents and derive a simple relation between the rms value of the alternating line current of a rectifier connection and the rms value of the fundamental of this current. Computation of the rms Value The rms value of a current of any waveshape is defined as the value of the direct current that is equivalent with respect to the generation of heat, if carried by a linear resistor. Hence for computing the tms value of a ‘current we assume that this current passes through a resistor with the value of resistance R, and determine the heat that is generated during the duration of one eycle (1/f): “ teat = [Ri ae ais. * Theat generated ducing de Usually,"the current is defined as a function of (wt). ‘Therefore we sub- stitute et/o for 1, which changes relation (18.1) into 1 [pa tech "ara os ‘Then weexpress the same amount of heat according tothe definition ofthe rms vale By heat = Rife toe = [2 ["F dai]*

You might also like