You are on page 1of 38
Room tenet Load Characteristic 13 ‘The Method For deriving the desired relation we utilize the following facts. @ Ina conducting secondary branch the difference between the line- to-neutral transformer-emf and the direct voltage, also referred to the (existing or assumed) neutral point of the secondary transformer windings, must be absorbed by the line-to-neutral commutating inductance L,, Hence this potential difference, divided by Ls represents at any instant the rate of change of the current in this line, difdt. (© The total of all individual currents is the d-c load, which is assumed to be without ripple. Hence the total of the rates of change of all contributing currents must be zero at any instant. ‘These two conditions will be satisfied only by the real currents and the real direct voltage and will therefore be used to determine these quantities. ‘Then, knowing the curves of the currents and of the direct voltage, two voltage-time integrals will be computed, one representing the reactive voltage drop and one representing the load current. ‘The relation between these two integrals is the desired relation between the reactive voltage drop and the current. Defi Of Tas In previous calculations all voltage quantities have been expressed in terms of the general reference E,,, and this proved to be very practical. For presenting the load characteristic in a scale independent of the constants of the circuit, a similar reference for the current is needed. Some publications employ a quantity that is based on rated current and the per-unit eommutating reactance. However, with this quantity the current seale is not always the same for the load characteristics of the various types of connections. Therefore a different reference quantity for the ‘current will be introduced, In accordance with E,, it will be denoted by Tuy and it will be defined as tam tala nn This definition is illustrated by fig. 10.1. We notice that /,, represents the load current at which the reactive voltage drop equals E,,, assuming. regular operation. In most cases, however, the operation is not regular Load Characteristic ‘This chapter investigates the operation of a rectifier connection under extreme load conditions and derives the relation between the direct voltage and the current for various types of connections ‘or the entire range from no load io short circuit The load characteristic of a voltage source represents the output voltage as a function of the current. The output voltage of a rectifier connection is By, reduced by the reactive and the resistive voltage drop (disregarding, phase control). The resistive voltage drop is, according to Chapter 6, conceived as part of the output voltage, Hence, only the reactive voltage drop determines the “slope” of the load characteristic, and this voltage drop will now be computed for the entire range from no load to short circuit. It is proportional to the current, as long as the commutation occurs in the manner described in Chapter 8, This is because the time integral of the voltage absorbed during commutation by the line-to- neutral commutating inductance of the succeeding phase represents both (@ The value of the commutated current, because the voltage-time integral equals Z, times the current rise from zero to this value, © The reactive voltage drop, because the voltage-time’ integral is withheld from the d-e circuit. Beyond a certain load, however, novel phenomena occur: 1, The individual commutations overlap, 2. The commutation between two phases delays the pickup of the third phase. ‘These phenomena affect the relation between the reactive voltage drop and the current. Hence we must investigate them in order to present the load characteristic for the entire range. ‘The investigations are rather complex and will therefore be restricted to operation with infinite in- ductance in the load circuit, with zero phase control and no losses in the commutating circuits im 148 D-C Characteristics and to exclude transient conditions. Thus the whole problem was only partially solved, In this chapter a completely different approach is used. The short cireuit path is conceived as a solid connection of the rectifier system. Certainly, this is a highly idealized condition, and the results may sometimes differ considerably from the real values, because they dis- regard any short circuit path impedance; but there are also aspects in favor of this method: @ The computed values represent the short circuit current under the worst condition and are therefore of special interest for the circuit design. (@ It is common practice to determine the impedance of a rectifier connection from measurements made with a bus bar across the de terminals. The results of the following investigations will serve as a guide for evaluating such measurements. ‘The assumption of a bolted short circuit permits us to simplify complex rectifier connections very generously, in most cases to single-phase circuits, and this is the important point. In these simple circuits the short circuit current can be computed by well- known methods including the transient component, and systems can be analyzed where different commutating loops are partially linked by common impedance. ‘Thus the treatment of this chapter, as compared to Chapter 10, provides supplementary information on the short circuit condition and is therefore useful, especially since the results satisfy closely the characteristics com- puted under the assumption of infinite inductance in the d-e circuit in Chapter 10. ‘Asalready announced, complex rectifier systems will be reduced towards single-phase connections. Therefore, the investigation starts with this category of circuits. Single-Phase Single-Way Connecti In a reactive circuit connected to a source of alternating voltage, the steady state current is displaced in relation to the voltage by the phase angle 4, defined by tan ¢ = wL/R, and the ratio of the erest values of voltage and current is determined by the circuit impedance, Z = (Fe + (oD). This is well known and is illustrated in fig. 11.1. At the instant of closing the circuit the current must start from zero, and the discrepancy between the steady state value and this starting value decays ‘Short Cireuit Currents 149 with the time constant of the circuit, 7'= Z/R, as explained in Chapter 1 ‘The “transient component” has been regarded in fig. 11.1 by displacing the zero line for the current curve in such a manner that all current steady state Fig. 14, Voltage and current in a reactive circuit values are presented with respect to the dashed-dotted curve, assuming ot, as the instant of closing the circuit, It can be seen that under the selected condition the current starts with negative polarity and turns to positive values at o,, remaining positive for more than one half cycle. primary secondary ipedance ipedance Wit shirt cut path t+----t4---4 { GQ Co) o Fig. 11.2, Single-phase single-way connection and equivalent crcl With an uncontrolled diode in the circuit, as shown in fig, 11.26, the situation is illustrated by fig. 11.3. The diode starts to conduct at ct since at this instant the voltage across the diode changes from the reverse to the forward direction and cannot be absorbed any more. Thus the arrent flow in the circuit is represented by the crosshatched areas, and "Connections with controled diodes are not investigated since the short circuit currents are usvally computed forthe worst condition, which sin the ease of controlled “lode, fring at the east posible instant, and which therefore doesnot difer fom the situation with regular diodes. Short Cireuit Currents 151 ‘we notice that the current is accompanied by a maximum transient component every cycle. The average value of the current is therefore considerably higher than a value computed from only the positive or the negative half-waves of the alternating current. The circuit of fig. 11.26, just referred to, is the equivalent for a singh phase single-way connection, shown in fig. 11.2a. This connection is [developed into the equivalent circuit by referring all values of impedance to the secondary side, as explained in Chapter 8, and representing the transformer by a voltage source that rigidly delivers the secondary no- oad voltage of the transformer, indicating that transformer and supply system have been deprived of any value of internal impedance. wle-Phase Bridge Connection In a single-phase bridge connection (fig. 11.4a) the short circuit path converts the diode network into two pairs of antiparallel-connected diodes. As a unit, each pair conducts in both directions and can be replaced by a straight connection, if we disregard any impedance of the individual branches. In the equivalent circuit of fig. 11.45 one pait is retained in order to mark the position of the diodes within the circuit so that later from the total current flow the current carried by each branch © can be determined. , Thus, with a voltage source without internal imped- P ance substituted for the transformer, and with all values of impedance F represented by L’and R, the simplification of a single-phase bridge connection with only common impedance results in the circuit shown in B fig. 11.45, 1 i "The current flow in such a circuit is illustrated by fig. 11.5. Since the | pair of diodes conducts in both directions, the current flowing through F the impedance of the circuit is the sam¢ as in a regular a-c circuit. tis ‘accompanied by a transient component that is a function of the instant E of closing the circuit, and it gradually approaches the steady state con- dition, The transient component is a maximum if.the circuit is being ‘losed at the instant of voltage reversal, the condition selected for fig. H 11.5. ‘The transient component increases the current flow through one H diode, represented by area A in fig. 11.5, and reduces the current load of f the other diode, corresponding to area B. Unider steady state conditions F both diodes carry equal load, represented by area C of fig. 11.5. ‘Computing the average of the rms value of the currents that are ac- ‘companied by transient components is quite difficult. Therefore, “transient factors” will now be defined and graphically presented; they simplify future calculations of this nature. 182 D-C Characteristies ‘Transient Factors ‘The current flow in a single-way circuit as well asthe load carried during. the first eyele by the diodes in a double-way circuit have been represented by crosshatched areas in figs. 11.3 and 11.5. Computing the size of these areas is rather complex but necessary for expressing the magnitude of the current flow, and many more times during the following investigations it will be necessary to determine the average or the rms value of such a current. Therefore, the most difficult part of the computations will be done just once, and the results will be graphically presented so that they ‘can be used for all future calculations of this kind. Fortunately, there is only one variable, the ratio @Z/R, provided that the results are presented in accordance with the following definitions: Transient factor for average values sew ditt, nen aie 119 f gv dt area C (fig. 11.5) wie [ad pees current ow none detion na reaetve Ratner oul nung 2m component, and ing tres te crest flow of enchant state” “steady state condition. For computing the rms value of the current carried by the diodes, we define, in accordance with the definition of the rms value of a current, Transient factor for rms values: Sema = [ fit / ftgee ai] (112) ‘here the integrals cover the same intervals as specified above, Both factors are functions of the ratio wL/R, and are graphically presented by fig. 1.6, In mathematical terms, these factors are: Sore = Lcd +8) (a1) 2eos ¢ Seas = [14 fee 1583 Short Cireutt Currents where tan ¢ = oL/R, and 4, shown in fig. 11.5, is defined in implicit form by sin d= sin g x evirsenoint (11.5) Figure 11.6 also shows the transient factor for crest values, firs Which represents the orest value of a current with a maximum transient com- i ‘ponent in relation to the crest value of the steady state current. This factor if ss )lU8CO ‘uke ig. 11.6, “Transient factors. 2 is well known and frequently used, especially for computing the maximum force between bus bars carrying a short circuit current of high magnitude, Example (Average and rms values of the current in a single-way circuit) Determine average and rms values of the current in the single-way circuit b of fig. 11.26 with = Veal oL = 810] R= 60] 154 DC Characteristics Solution. Assuming a short circuit across the diode, we compute the rms value of the steady state alternating current. 50 [V) Vie tay* + 6 lap Then, we assume that each second half-wave is suppressed, but still disregard the transient components, and calculate: Tema = SIAI B 1 fyg = 5% S{A]=2.25(A] and Ime = x 5 {A} =3.54(A] In the circuit under consideration a maximum transient component starts every cycle, as shown in fig. 11.3. Hence we read from fig, 11.6 for wL/R = 1.33 the transient factors, and compute the real values Taye = 1.3 X 2.25 [A] = 2.93 [A] and Inna = 1.2 X 3.54 [A] = 4.25 [A] 2-Pulse Midpoint Connection ‘The conversion of this connection into an equivalent circuit i illustrated by fig. 11.7. As the first step, it is recognized that only one point of potential is common for the two branches so that the currents in these branches do not affect each other. Hence these currents will not be changed if the two circuits are separated, as shown in section (6), and recombined with a different sequence of the elements in one of them, as illustrated in section (©), provided the polarity of the: diodes is Kept ‘unchanged in relation to the polarity of the voltage induced in the trans- former windings. Now, the two transformer windings coincide and can bbe represented by one voltage source. With the primary impedance transformed to the secondary side, this results in the equivalent cirovit shown in section (@), In this cireuit some value of impedance is commoin and some is in- dividual for the oppositely conducting branches, whereas the connections investigated before have been limit cases of this general condition, one ‘with only common and one, with only individual impedance. It is now of interpretation whether impedance is common or individual; rather, it depends on the linkage between the magnetic fields caused by the currents in the various branches whether voltage is induced only in the branch Short Cireut Curens 155 f that carries the current, or in others too, and ftom the arrangement ofthe F transformer windings it can be very well distinguished between common and individual values of leakage Inductance, es demonstated later in | this chapter with sample calculation. Besides, the veal currents ier i considerably between the two limit eases, as alcady noticed in Chapter 10 where it was computed thatthe short circuit currents n crits with | only common and others with only individual impedance difer by the & 2) | @ Fg. 1.7. 2-pule midpoint connection developed into equivalent circuit. F factor 3. This may accent the importance of distinguishing between Common and individual values. Unfortunately, the treatment of the eneral scheme, shown in fig. 11.7d, is extremely difficult. There are | intervals. during which the common impedance is in series connection With the individual impedance of one branch, periodically changing with intervals during which both branches are conducting, and transient com- Ponents that decay with different time constants are superimposed. » Even the graphie presentation of these relations is difficult since there are 0 many variables. Thus these linkages are the origin of most difficulties of investigation, and of a good deal of confusion. “In view of both the difficulties and the need for an answer, we present fn approximation which is rather unique, but leads to practical results Fand to some understanding of the phenomena that occur in such cireuits “and of their inluence on the short circuit currents 156 D-C Characteristics Approximation for Cir its with Common and Individual Impedance ‘The basic idea of the following treatment is to substitute an equivalent impedance for the parallel branches with the result that the voltage across these branches and hence the individual currents can be determined, For the fist step, the substitution of an equivalent impedance for the parallel branches, let us assume that these branches are connected to a source of sinusoidal voltage, as indicated by the dashed circuit in fig. 11.8, and let us determine the individual currents as well as their total. The individual currents are combinations of the steady state alternating current, shown in fig. 11.9a, and transient components that start at the instants of voltage reversal, as already explained for single-way circuits. Thus the individual currents are represented by the ordinates of the crosshatched areas in fig. 11.94. As can be seen, there are intervals, ¢.g., from to ‘fs, during which only one branch is conducting and the total current is {identical with the current of this branch, and there are other intervals during which, as a result of the transient components, both branches are conducting. During those intervals current is circulating in the loop of the two branches while the voltage source catries only the difference of the individual currents. This current carried by the voltage source is presented in fig. 11.96, marked by ®. If it were a sinusoidal current, an impedance could be defined that precisely substitutes for the parallel branches—for Short Cireult Currents ig. 119, Currents carried by parallel branches and by equivalent impedance, the steady state condition, as least, However, this current differs slightly from a sine wave. Therefore, we substitute for the parallel branches an impedance as an approximation. F For rough calculations we determine the value of the substitute from F the fact that the crest value ofthe total current, @), equals the crest value F of the current in each branch, which is the crest value of the steady state 158 D-C Characteristies alternating current times the “transient factor for crest values,” presented 85 fi fig. 11.6. Hence we compute Zagasy A Zina (11.6) where Zequy = equivalent impedance, substituting for the parallel branches Zesay = impedance of each branch ‘This approximation disregards the fact that the real current, (is less displaced in relation to the voltage than the steady state alternating current, or, differently interpreted, that the substitute must be more resistive than the impedance of each branch. For more accurate computations we assume that the sine curve that gives the closest approximation is the fundamental of curve (), and we determine the equivalent impedance in such a manner that it carries a current equal to the fundamental ofthe total current of the two branches, if connected to the same source of sinusoidal voltage. Computing this fundamental is quite difficult, but since there is only one variable, this is the ratio o£/R of the individual impedance, the results can be easily presented and can be used for all future calculations of this kind. We introduce the symbols 4, = crest value of component of fundamental “in phase” with the voltage across the parallel branches. 2, = crest value of component of fundamental 90° legging, Esa = sms value of sinusoidal voltage across individual branches, 4 = auniliary angle, shown in fig, 11.9, defined in implicit form by sin d= sin g x en tt448)and ay 4 = displacement angle between voltage and steady state alternat- ing current, determined by tan g = 2s with Risaiv crest value of steady —_/3 Bina state alternating current =~ 7,5, the computations result in a, = L2Einate Le 4 4 d)c08 4 — sin($ + d)c08(2$ + 8)} (LL.B) Zina 7 and 5, =< V2 Binate Le 4. 4 dpsin g — sin( + dsin2p +8) (11.9) Zinatv Short Cireuit Currents 159 FFor practical use, these results have been further developed into two factors that ditectly transform the real values, Iggay ANd Ray» AS Well as their resultant, Zui into the values of the SUBStitUte, Lene» Regs 4 “Recand Rie gilli Tygon teens, + 6 a) eabntol indy —> Fig. 1110, Factors for computing equivalent impedance, rive These factors are graphically presented in fig. 11.10, With their aid the problem will be solved by the following steps: [ 1. The equivalent impedance (inductance and resistance) will be deter- mined from the impedance of each branch, using the chart of fig. 11.10. 2. The supply voltage of the equivalent circuit (ig. 11.74) is divided between the common and the equivalent impedance; the voltage sioross the equivalent impedance is determined. This voltage repre- sents, as an approximation, the voltage across the individual branches. 3, From the voltage across the individual branches and the values of resistance and inductance of each branch the individual currents are determined, as explained for the single-phase single-way connection, ‘This will be demonsteated with the example following the next paragraph. 160 D-C Characteristics The transient situation in a circuit with common and individual imped- ance is illustrated by fig. 11.11. Certainly, the current in one branch will rise to the highest possible value if the voltage time-integral of one full half-wave of the supply voltage is available for this rise, that is, if the steady state trating erent — A across parallel | hy, i | Fig, 11.11, Transient situation in a eicuit with common and individual impedance. circuit is being closed at the instant of voltage reversal. This condition has been selected for fig. 11.11, and the instant of closing the circuit has been denoted by wf. During the following period, the supply voltage is divided between the common impedance and the impedance of the conducting ‘branch, and the last portion appears across the second branch which does not start to conduet before this portion reverses the polarity. This occurs at of; in fig. 11.11. (For this illustration it has been assumed that the ratio wZ/Ris not the same for the common and the individual impedances, in order to show that af, does not necessarily have to coincide with the instant of reversal of the supply voltage.) After the second branch has started, the rate of decay of the total current is balanced by both the rate of decay of the current inthe first branch and the rate of rise of the current in the second branch. Thus the second branch reduces the rate of decay of the current in the first one. In other words, it increases the total current load of the first branch. However, this increase occurs close to the end of the conducting period of the first branch so that the real current load is only slightly higher than the value computed under the assumption that during the entire conducting period the common impedance is in series connection with the impedance of just one branch, In fig, 11-11 area A represents the current flow until af, and area B represents the rest of the flow disregarding the increase caused by the second branch. Short Circuit Currents 161 Example (Computing currents in a circuit with common and individual impedance) f) Determine the average value of the current carried under steady state 8 by each diode of the circuit of fg. 11.12. Lamon 410) ©) 125 1v) ms) Alaa 610 Ray = 210) ig. 11.12, Circuit for sample calelaion, Solution. For (oL|R)saiy = 3, we read from the chart of fig, 11.10 Legut Zea 20,62 and = Lina Rinate and compute the values of the-substitute: ‘oLaquly = 0.62 X 6 [2] = 3.7210] Requte = Ll X 2{0] = 2.2[0} ‘Then we determine the current flowing through the common impedance in seties connection with the substitute: ——_ 5 sora V(4 [0] + 3.72 (Q})* + (1 (Q] + 2.2 (Q)* With this result the voltage drop across the substitute ean be determined: 150 [A] V@.7210)* + @2 [OP = 65 [V] B As the next step, we assume that the parallel branches are connected to a F voltage source of 65 V, and that the diodes are short circuited. This gives 162 D-C Characteristics ‘a steady state alternating current through each branch with an rms value of 65 1¥) Vo 1a} + 211 With the diodes, the current flow is suppressed in one direction, and we compute the average value of the positive or the negative half-waves of the steady state alternating current: 10.3 [A] 2 103 1A] = 40644) In addition there are the transient components, increasing this result by the “transient factor for average values.” For oL/R = 3 (individual impedance) fg. 11.6 shows fagg = 1.76, and we compute the average value of the real current carried by each diode: nyg = 1.76 & 4.64 [A] = 8.15 1A] IF the cirouit of fig. 11.12 is the equivalent circuit of a 2-pulse connection, the short circuit current in the d-c path of this connection will be twice the computed value since the two branches conduct in opposite directions only in the equivalent circuit; in reality, both contribute to the ccurrent in the d-c path. ‘3-Phase Bridge Connection Previously it was stated that the short circuit path, conceived as a solid connection without impedance, subdivides complex rectifier systems and thus simplifes their treatment. This will now be demonstrated with the 3ephase bridge ofineetion, Figure 11.13a shows this connection with the ‘transformer already represented by voltage sources delivering the secondary no-load voltages of the transformer, and with all values of impedance referred to the secondary side nid assumed to Bgcoimon forthe oppositely conducting branches. It now can be seen that the short circuit path converts the diode arrangement into three pais of antiparallel-connected diodes which conduct in both directions and-can be replaced by straight connections, as already explained for the single-phase bridge connection. Hence the currents in the three phases will be the same as,in.a regular 3ephase system with symmetrical load, consisting of inductance and ‘Short Cireuit Currems Fig. 1.13, Reduction of a S-phase bridge connection, resistance (fig. 11.138). For such a system we compute eee = ea GR — 241 (phase 1) dt di bao = bo — gk — St hase 2) AR—2BL (phase 2) di ey (phase 3) “PL (phase3) E Total: Bean = (ea Here + Oo0) — (te FOR =Htdit ty, (11.10) ‘The total of the three supply voltages is zero, because we assume a sym- cal system., The total of the three currents or of the rates of change of the currents is also zero because theré is no return conductor.. Hence at any instant. From this we conclude that a connection made Pbetween the points a'and'0 (dashed in figs. 11.13a and B) would not carry any current and therefore would not change the currents in the individual phases, But this connection would indicate that the three phases are tied together at only one point of potential so that the currents are not changed ‘each phase is individually short circuited, isolated from the other phases. nce such a system can be split into three separate single-phase circuits. FSince this conclusion was based only on the assumption of symmetrical D-C Characteristies (e) phase 1 Wome2 + coomes a1 ot Fig. 11.14, Determining the current in the d-c circuit from the currents in the individual ‘phases, only common impedance 20 anit 10 5 25 12 i. of steady stata vale —>- °% 30 760 270 360 ca Fig. 11.45, Average rte of short circuit current in d- path, only common impedance. Short Cireuit Currents supply voltages and the conditions dictated by the circuit configuration, the result must be true for any kind of current that is possible under these conditions including transient components [Less evident is the situation if the three phases are linked together by @ ‘common transformer, especially if the windings on either the primary or the secondary side are in delta connection so that the voltage drop across the impedance of each primary line is caused by currents of different secondary phases. However, it can be shown that in this case as well the phases can be separated, provided the secondary circuits are double-way circuits with only common impedance. Under the general condition of common and individual impedance the oppositely conducting diodes are not directly paralleled, and it would be incorrect to substitute a solid connection for them and to separate the | three phases. However, it was shown before that, as an approximation E such branches can be represented for the steady state condition by equiv- alent impedances, and if this is done the rectifier connection is again E identical with a regular a-c system with symmetrical load (fig. 11.138), where the three phases can be separated, Therefore we decide: We separate the three phases in any case, but realize that this is an approximation unless all values of impedance are common to oppositely conducting branches. With this background the currents in a short circuited 3-phase bridge connection can be determined. With only common impedance the currents Jeads and in the individual branches are the same as calculated f for the single-phase double-way circuit, illustrated by fig. 11.5. The eurrent in the d-c circuit is the total of the currents of all contributing branches. This is illustrated by fig. 11.14, where the transient components are taken {nto account by displacing the zero lines for the current curves so that the P heavily outlined areas represent the real current flow in the individual | branchés. Their total is the current in the d-c path, shown in section (d). ‘Quite obviously, the result must be the same if all values of the negative side are added. With a different instant of closing the circuit in relation to the alternating voltages, the “breaks” in the curve of the direct current are displaced, but ‘The basic character of the curve will be unchanged, and ‘an average curve can be drawn that distegards these small fluctuations. ‘This curve is a function of the ratio wL/R of the circuit and is shown in fig. 11.15. It should be noticed and memorized that the first rise is practically the same in all cases, reaching the steady state level approxi ‘mately 60° after the instant of closing the circuit, and that the “overshoot” can be computed from the transient factor for orest values, fu (fig 11.6). 166 DC Characteristics ‘The other extreme is a connection with only individual impedance. In this case a maximum transient component starts in each phase every cycle, as illustrated for one wye of such a system in fig. 11.16. It can be seen that the current in the d-c path approaches, without overshoot, a value (e) phase (e) phase 3 AL (tout 3 ete Fig. 1116. Determining the current in the d-c circuit from the currents in the individual phases, only individual impedance. that is the steady state value without the transient components times the transient factor for average values, fuyg- For connections with common and individual impedance an approxi- mate method was presented for determining the currents in the individual J branches, and the direct current is, also in this case, the total of all currents flowing into the d-c path. Hence the general case does not need further treatment. We will now apply this knowledge to a practical problem. Short Cirewit Currents ixample (Z-phase bridge connection) Determine for « 3-phase bridge connection the steady state short circuit ‘current in the d-c path in terms of J,., assuming no losses and (@) only common impedance © only individual impedance Solution. (a) The steady state alternating current (rms value) is F,foL F com this value we compute the steady state short circuit current in the 4 path with only common impedance v2 Be = ol, ratio avg value of haf sine wave to ims value of full sine wave |__(3) phases contribute Ta, snort otroutt = ‘The theoretical value J, is defined by equation (10.1): nen By = 38, = 3 BS (equation 6.3) 1, can be expressed by ae Jags S35 = a ‘and the desired ratio will be computed as ‘Only common impedance: 1 3 a, sontonoutt _/3 TG Ts . (1.11) th only individual impedance the average valve of the currents in Re individual branches, as well asthe current in the d-c circuit, is increased by the transient factor yg, hich equals for an inductive circuit without losses (ratio of the arcas (4 + B)/A in fig, 14.6). Hence the desired ratio is Only individual impedance: fant 5 (11.12) 168 D-C Characteristies Both results accord very well with those derived in Chapter 10 for infinite inductance in the d-c circuit, The result of equation (11.11) differs by only 5%, while that of (11.12) is identical with the one derived in Chapter 10, This seems to indicate that the inductance of the d-c circuit has very little influence on the short circuit current, so that the results of this chapter, although derived for the idealized condition of a bolted short circuit, supplement very well the information obtained in Chapter 10. Double-Wye Connection with Interphase Transformer Ina 3-phase bridge connection, two 3-pulse commutating groups operate in series; in a double-wye connection with interphase transformer, they are paralleled. In both cases the two wyes operate with 180° phase shit ‘The only difference is that in a bridge connection the phase shift is the result of the opposite polarity of the two wyes, while in an interphase transformer it is achieved by separate secondary transformer windings ‘Thus both connections are very similar, and the basic idea of analyzing the interphase transformer connectionis to convertit into abridge connection— without changing the currents carried by the diodes, of course. This illustrated by fig. 1.17. ‘Section (a) of fig. 11.17 shows the circuit under consideration, In series connection with the primary and secondary transformer windings are the primary and secondary values of inductance and resistance. The two windings of the interphase transformer, with leakage inductance and resistance, are shown between the center points of the two sets of diodes, although they are sometimes connected between the neutral points of the two sets of transformer windings. As the first step, the two commutating groups are separated. This does not affect the currents since only one point of potential, the short circuit path, is common for the two wyes, and therefore no current circulates between them, 12096 Matvia ely cone 0 % 6 3 «a1 /R find impedance —> 1 Fig. 1.20, Correction factor for the alternating line current. been computed, They are presented by figs. 11.21 and 11.22. Again, the curves for the average value can be applied to either the average value of the current in a single branch or to the current in the d-c cirout, since this, current is the total of the currents of all contributing branches. Looking at these curves it seems surprising that there is so much f difference between only common impedance and a small amount of F impedance being individual, But it seems logical if considered in the | following way: ‘The increase of the current in a single-way circuit (ig. 11.3) caused Eby the transient components depends only on the ratio ok R ‘na circuit with common and individual impedance (fig. 11.74) a similar [situation exists in the individual branches, and even a relatively small * amouint of individual impedance is able to set up the transient currents {that increase the currents in the individual branches and in the d-c circuit considerably if only the ratio LR of the circuit in which these transient components decay is high enough. It should not be disregarded, however, ‘that the short circuit path is part of the loops in which these transient #components decay so that the short citcuit path resistance influences the fate of decay of the transient components the more the smaller the value of [individual impedance. This leads to the not very encouraging conclusion e—theoretically, at least—some impedance is always individual, currents are highly affected by small values of short 116 D-C Characteristies 22 oni Tada, 18 ¥ 0, |__4 ceerection factor —>- | 208 odiual at I L ~ 12 ; ly conan 19 * 0 2 8 10 * {aL/R of Indie impedance —>- Fig. 1121, Comrection facto for the average value of the current in the individual branches. common, Thus calculations of the short circuit currents or, turned the other way, of the circuit characteristics from short circuit tests, will not bbe very accurate, unless all details are very precisely determined. In view of this the approximations presented in this chapter seem fully sufficient for nearly all practical calculations. In a DOUSLE-WYE CONNECTION WITH INTERPHASE TRANSFORMER the two commutating groups are paralleled instead of series connected. ‘This does ‘Short Circuit Currents 7 F not change the situation in each commutation group, and the ratios P graphically presented in fig. 11.19, as well as the correction factors, are defined in such a way that they apply to both circuits. Only the curve showing the rms value of the secondary line current needs a different interpretation. In the bridge connection the secondary line current is the total of the currents of two oppositely conducting branches. In the inter- phase transformer connection these currents are carried by different 20) ny ini ® D | i { 5 » | #4 7 only egmmon 39, 8 10 4 6 ‘WEIR of Indus mpadance —>= Fig. 11.22, Corretion factor forthe rms value ofthe current inthe individual branches. 0 2 transformer windings, and only the primary line currents show their com- bination. Hence this curve may be used for the interphase transformer ‘connection to determine the primary line currents—not disregarding the P ratio of transformation, of course. Compatison with the Diode Ratings ‘The current that can be carried by a diode is limited by the maximum temperature the junction of the diode is permitted to assume. Under the steady state condition the heat generated in the diode is in balance with 178 5 D-C Characteristies the heat transferred to the ambient, and the maximum permissible temper- ature difference between junction and ambient determines the maximum permissible heat flow, and hence the maximum current. If the current is suddenly increased, or started, the heat flow, and hence the junction temperature, follows with delay because the masses of the junction and adjacent parts absorb some of the heat in order to raise their temperature, Ifthe generation of heat is then stopped before balance between generated and dissipated heat is achieved, the temperature difference will not have reached the value that corresponds to the same current under the steady state condition, Thus the diode is able to carry more than rated load for 1 short period of time. ‘A diode in a short circuited rectifier connection is exposed to current pulses, so that in between these pulses it is given some time to cool off, and if it has endured the first pulse without damage, the next few pulses are usually not critical, especially in a circuit with common impedance where the transient component is initiated only with the first pulse. If, however, the short circuit stays for a considerable period of time, the diode may then be damaged because the temperature difference approaches more ‘and more closely the value required by the balance between generated and dissipated heat. In view of these relations two conditions are of special interest (@ The current pulse during the first cycle, © the steady state condition. Starting with the STEADY STATE CONDITION, the overload capacity of a diode is usually specified by the average value of the current that can be ‘carried for a certain period of time, based on current pulses with rec- tangular waveshape and a certain conducting angle, or on half sine waves. Under the short circuit condition, the current pulses are (nearly) half sine waves, so that with the knowledge of the previous investigations the ‘maximum permissible current, or the maximum time until the short cire rust be cleared, can be determined. If the comparison is made on the basis of average values, using the valuc that is shown in the diode specifi cation sheet for 120° conduction, the result willbe on the safe side because the ratio of rms/average value is higher for 120° conduction than for half sine waves. ‘More diffioult is the comparison for the CURRENT PULSE DURING THE inst CYCLE, since in this ease the average or the rms value taken over one full cycle does not closely enough represent the real duration of the heat input, Therefore we compute either the heat that is generated in the junc~ tion by the current pulse under consideration and determine the tempera~ ture rise directly from it, or we convert the real current into a de pulse ‘Short Cireult Currents F that represents the same current load to the diode, and compare this pulse with the short time ratings for dec. In the first case, we recognize that the voltage drop across the diode consists of a constant portion, and a portion that is nearly proportional E to the current, and compute the corresponding portions of the generated heat, W, and W,, measured in wattseconds, Assuming a maximum P transient component we find fame vale of teady 1. cotta portion va Sits aerating, * Joe % 37% of ola aon curent tiode a transient avg value of steady : . state alternating current °° conte total ampereseconds of current pulse (11.16) For computing the other portion, we determine from the slope of the voltage-current characteristic of the diode (d-c values) the “internal ance,” and compute: [i value of steady 1 “internal resistance total ampere-square-seconds of current pulse aun ‘The sum WY, + Wis the total heat generated by the pulse. The tempera ture rise still depends on the length of time during which this amount of heat is being generated because the time determines the extent to which adjacent masses participate in the absorption of this heat. Therefore, we E represent with At the duration of the major portion of the heat input, disregarding intervals that do not contribute considerably, and determine Hi the temperature rise of the junction W,-+ Wy, “transient thermal Ao = “1 & resistance” [°C/W] (a1.t8) F The “transient thermal resistance” is a function of the duration of the [> current pulse and is specified by most diode manufacturers, For supply frequencies of 50-60 cps a value of Ar = 10 msecis a good approximation. Tr the transient thermal resistance is not known, or if it does not seem P necessary to investigaté the situation so closely, we may convert the teal ‘current pulse into a d-e pulse that generates the same amount of heat, and ‘we then may evaluate the result with respect to the short time ratings for D-C Characteristics AL/R of comron ‘and nada Impsane = 10 4 065 Gas COT Gos TOR OT oO 020 Zep (oet-voltcommutating impedance) —>- ig. 11.23. Caleulaton chart for the transient short circuit curent ina 3-phase bridge (ora doubie-wye connection with interphase transformer. dc. Again, 10msec is a fair value for the equivalent pulse, and we compute magnitude of _ sms alue of steady Tome pulse ~ sate alternating current * Se Fo rasa H9) For f= 60 eps the value of the square root equals 0.915. ‘The rms value of the steady state alternating current can be easily deter- mined from the no-load voltage and the impedance of the equivalent and is graphically prescnted by the chart of fig, 11.19. The transient Short Circuit Currents 181 fact0r fine is shown in fig. 11.6. From both sources, and from (11.19) the chart of fig. 11.23 has been computed. It shows the magnitude of the equivalent 10-msec pulse in relation to rated direct current. Although this chart primarily refers to circuits with only common impedance, it may also be used for other conditions since it was realized earlier inthis chapter that with some impedance being individual the current flow during the first cycle is only slightly increased, ACC Short Circuit short circuit currents are sometimes initiated by internal faults, especially by diode failures, and if the diodes are protected by fuses, the points of ‘concern are as follows. (@ Is the lowest possible short circuit current high enough to blow the fuse in series connection with the failed diode? (©) Is the highest possible short circuit current low enough not to blow fuses or to damage diodes in other branches? Ifa high number of individually fused diodes are paralleled in each branch, the situation is usually not critical since many fuses stand against one. If, however, the number of paralleled diodes is small, a careful coordi- nation between the fuse and the diode characteristics is required, as well as 1a good knowledge of the real short circuit currents. As a guide for investi- gations of this kind, the short circuit currents in a 3-phase bridge connec- tion will now be investigated. Figure 11.24 shows the potential and current curves, assuming that the diode in the positive branch of phase 1 fails to block the reverse voltage at the end of the conducting period. From ant, to wy, phases 1 and 2 are short circuited, and the potential of point a stays on the mean value of the no-load potentials ofthese phases, as explained in Chapter 8 for commu- H tation., At of, phase 3 joins the short circuit, and point a stays then on zeto potential, because the three phases are symmetrically short circuited, This situation continues until, at «fy, the current of phase 2 is reduced to F zero again. During the following interval phases 1 and 3 are short P circuited and the potential of point is determined by these two phases. F The story can be easily extended. From the voltages between the points 1, 2, 3, and a we then determine the voltage-time integrals that represent the tise or decay of the currents in the corresponding phases. The areas representing these values, as well a the currents, are shown in sections § © and (6) of fig. 11.24. The total of currents @ and @) is the reverse current in the positive branch of phase 1. For rough approximations it D-C Characteristies 8 © Oo © “a Cr a Fig. 11.24, Situation in a 3-phase bridge connection with a-c short circuit, ccan be assumed, according to this illustration, that current @) is raised from zer0 to its erest value by the voltage-time integral of one half-wave of the line-to-neutral voltage. This is the same situation as under a sym- metrical short circuit with a maximum transient component. The current through the failed diode includes the contribution of phase 2 and is therefore somewhat higher. iInverter Operation From a rectifier operating with «> 90° a general picture of @ line-controlled inverter will be derived. Special attention will be ‘paid f0 the current transfer between the commutating phases. ‘Actectifier transmits power from an a-csystem to a d-ccireuit, An inverter F does the opposite: It is a load to the d-c circuit, and a source of a-c power. Depending on the structure of the a-c system, inverters can be classified into two basic groups: (@) Live-conTROLLED INVERTERS; @ sur-coNTROLLED INVERTERS. F Inverters of the first category are connected to and controlled in synchro- isi with an a-c system that is rather “stiff” in frequency and voltage and able to exchange with the inverter any amount of effective, reactive, and distortion power with practically no influence on the frequency or the ‘voltage of the a-c system. In the second category the effective, reactive, and distortion power E delivered by the inverter must be fully accepted by a relatively weak a-c ‘circuit. As a consequence, frequency and alternating voltage are variable F and assume the values that satisfy this condition. E- The investigation will be restricted to inverters of the first category since Htheit operation is similar to that of controlled rectifiers, whereas the foperation of self-controlled inverters is completely different and does not fit into the philosophy developed in this book. f Line-Controlled Inverter f certain degree of phase control the direct voltage of a rectifier connection starts to assume during certain intervals negative instantaneous ‘values, provided that the current is being maintained, This last condition 183

You might also like