2 Principles of Steady-State Converter Analysis 24 INTRODUCTION104 Fie, 22. Detention ofthe sikh supa oag Ce component by inning and dindng ty te (= DY, switching pei ory, o o DT, 7 power semiconductor devices, such as transistors and dies, which ane controlled to tum on and off as required to perform the funetion of the ideal switch, The switching Frequeney f, equal to the inverse of ‘he switching period 7, generally les inthe range of L kHz to | MHz, depending on the switching speed ‘ofthe semiconductor devices, The duty ratio D is the faction of time that the site spends in position | ad isa number between 2ero and one. The complement of the duty ratio, Dis detiaed as (1 = D “The suites re the de componeat of the voltage: the switch output voltage vf) has ad ‘component is less than te converter de input eV, rom Fourier analysis, we kno thal the de ‘component of v2) is given hy its average value (9), oF 4s illstrated in Fig. 2.2 the int value is j= 2 (ony, So the average value, or de component, of (2) is equal to the duty eyele times the de input voltage V The switch reduces the de voliage bya factor of D. ‘What remains i to insert a low pas iter as shawn in Fig, 2.3. The iter is designed to pass the «le comporent of (2, ha to reject the components of», at the switeing Frequency and its harmonies. ‘The ouput volts ly equa wo the de comport of 9,0 “The converter of Fig. 2.3 has been realized using lossless elements. To the entent that they are ideal, the inductor, capacitor, and switch do not dissipate poser For example, when the switeh is clase, its vol age drop is zero, and the eureat is 2et0 when the switch is open In ether ease, the power dissipated by the switch is zero, Henee, eiiencies approaching 100% ean be obained, So to the extens that dhe com ponents are ideal, we can reulize our objective of changing de voltage levels using lossless network L eel * RS wo Fig, 23° Insertion tie outa ‘he swihing harmonics pas only te de eompoaent oy 0 1Fig. 24 Buck converter de output wllage Vs. duty eye D ‘The network of Fig. 2.3 abo allows conto ofthe outpat. Figure 2 isthe contol characteristic of the converter. The output volta ctistc, Also, the output vol Feed systems are often constructed tha adjust the duty eycle D to regulate the converter ‘opt volt ven hy Eg. (2.3) is pled ss. duly eycle. The buck eonverter bes «less than or equal to the input voltage, since Invertets or power amplifiers ean also be bull, in whic the duty eyele wares slowly with ‘ime and the output voltage Follows The buck converter is just one of many possible switching converters, Two other comnosly “J comserets, which perform diferent voltage conversion funotions, ate illustrate i Fig. 2.3. Inthe boost convert. the positions of the inductor and site ae reversed. Its shows later inthis eater that the boost converter steps the volta e up: VV, Another converter the buek-hoost converter, ean either Increase or deevease the magnitude of the voltage, butt ‘inp ‘of any magnitude. may at magaitude das the input, oF iat have opposite polarity, But itis indeed possible to praluce any desired de ourput volag laity is inverted, So with a posit oltag, the ideal uck-boost converter can produce a negative tpt volts fist be suprising that de output voltszes can be produced that are greater using ps sive network of only inductors, capacitors, and emedied sich, In the above discussion, it was possible to derive an expression for the output voltage of the buck vonverter, Eg. (2.3), using sone simple sapuments ba ‘on Fousier analysis. Hossever, it may 1 be immediately obvious hows to directly apply these szaments to find the de expat voltage of the boost buck boost, o at tral comertets. The objective of this chapter is the development of a more g method for analyzing any sitchin svt [L. ‘The principles af iuezor y converter comprised of network of indictors, capacitors, and tesecond balance and eapacitor charge balance ate derived these can be used te solve forthe inductor curents and capacitor volt Of switching converters, A useful catly facilitates the analysis. Some ting the filter element values are also discuss spproxinn simple ple appraination, thas For sel 22 INDUCTOR VOLT-SECOND BALANCE, CAPACITOR CHARGE BALANCE, AND ‘THE SMALL-RIPPLE APPROXIMATION Lotus mote closely examine the inductor and capacitor waveforms in the buck converter of Fig. 2.6. Its posible to build a perfect loe-pass filer that allows the de compo the components atthe sitching frequoney and its harmonies, So the low-pass fiker must allow a least equescy harmonics generated by the switch 1 real the output. Hes ‘waveform te) appears as ilustrated in Fig, 2.7, andl ca be expwessed 2s 119 pass but completely removes fn practice the output voltag So the tual ouput vliage v2) comsiss of the ds re se component pls 8 small undesired ae comaipl Apoosa etal yeavefirm MDS V tpl Fe 27. Osu wipe saetoon eh y | ee consisting of de componeat Vand svitci ; pple is much smaller than th ore, the ouput voltage») swell approximated hy its de earsponnt¥, with the small ripple term ted wis 26) ‘This approsimation, known as the small-sgple approximation, or & greatly simplifies th analysis ofthe converter waveforms an is used thr Newt ing the inductor val noeted tothe input volts cen hy carripple approximation, out this book 1 us analyze the inductor current wavetonmn, We ean tnd the inckictoreurtent by integra waveform, With the switch in position 1, the #1 sie of the inductor is con “and the eireuitresuces to Fig. 2.84). The inductor voltage v2 is te As de ribod above, he output voltage v(f) consists of the de component V, plus @ small ae Fipple testa (5, We can make the smal ripple approsimation here, Eq. (2.6). replace rit} with its de compo ent So with the switch in position 1, the inductor voage is essentially constant and equal 10, ~ as shows in Fig. 29. By know induetor curtent can be found by use of the defini 1 the inductor voliage waveform, th voan ti 294 8 i ok » L P50. P55E* 2H = ' tun Ty | ve c wv Wg ke LN Huck eomerter ets (a wile the switch sin poston 1b) while the such is in poston 2W——Penopler of Suan Caer Aatns Fig, 29 Seay ste intact voltage waster, —— oF pr, | buck sone . Sitch positon 1 2 t Thos, during the frst ineeval, when 00) is appronia ely (VV), the slope of the inductor eurrent sussefonin is which follows by diviging Bq, (2.9) hy £, and substituting Es essentially constant while dl (28), Since the inductor volag witch s position 1, the inductor eusrent slo sto andthe inductor eurtentinereases Tine al, when the switch isin position 2. The io the iret of Fig. 2.8(b). [ts important toy eurvent and voltae: in particular, dhe polity’ of 28(h), So the inductor voltage din Similar angunnests apply du lef side of the inductors then connected to ground, le eamsistently detine the polarities of te ind fined consistently in Figs, 27.2 Ine sscone!subinterval “a aun Use af the smal ripple approximation, Bq. (2.6), leads to nape ¥ 22) So the inductor voltage is also essentially constant while the switch is in postion 2, as lusts in Fig 2.9, Substitution of Fi, 2.12) imo Eq, (2.9) and solation forthe slope ofthe inductor eurent yields Hence, during the second shinteral the inductor current changes with a negative andl essentially con stant slope ‘We can now sketch th 10, "The inductor current begins at som initia walue (0) Daring the fist subintersl, with the switch in pasition 1, the inductor current inereases with the slope given in Eq. 2.10) At time r= DI, the switch changes to postion 2. The rem hen eereases with the constant slaps given hy Bg. (2.18. Attime r= 7, he switch ehanges back 10 inductor curzom wavetorm (Fig Fig, 2.10 Steady-state inductor carent waveiem, buck comer i)position 1 and the process rep I of interest to calculate the inductor current pple Ai, As illuseated in Fig 2.10, the peak induetoe current is equal tothe de component J plus the peak-to-iverage ripple Ai. This peak eurrent Ns through not only the inductor, bu alse Hou Kpowledge ofthe peak current is necessary when specifying the ratings ofthese devices, Since we know the slope of he inductor curveat dung the fest subinerval and we uso knows es subierval the current increases by 2 (since A, i the peak rip ual tothe slope (che applied smiconuluctor devices that comprise the snitch thole cal about fad hence during the I-tospeak ripple is 24). So th ‘of the fest subinterval we ean calculate ce ripple magnitude. The jt) wa pile. the 9 inductor vol change in curent, 2, is © slvided by 1.) mes the lenge of the fist subinteral (DT j angen i) = openness very eu ead=[5~}lon) Solution or i, yields ‘Typical values of A, fe inthe undesirable to allow Ai, to bo ani ofthe semiconductor switching devi ‘of 10% to 209% ofthe ful-lod value ofthe de component ffs jase the peak curren of the inductor the ome too large; ding so would ine and would increase their size and cost. So by des inductor current ripple is also usually smal compared tothe de component J. The smull-ripple approxi matin j(71= Fs usually justifies forthe inductor ewerent The inductor salve can be chosen suc that & dested current ripple Adis ataine. Solution of Fg, (2.15) forthe inducsanee yells bi eis ue oF inductance in the bck eon This equation is eorammly used 0 Lis emily possible wo solve conserters exactly, thous use of the smal-rpple approximation For example, one could use the Laplace transform fo write of Figs, 2.8ta) and 2.8(b), Ose could then iver the transforms, match boundary conditions, wl fi the petindie steady-state solution of the cnet. Having dom cull then fn the de components of orm and the peak values, But this i 3 intractable, Besides, the extra work involved in siting eguations that exuetly describe the ripple is apponimiation is exsy bo apply waste of time, sine the ripple is small and is undesired, the small-ripph «and quickly yield simple expressions for de de eompoments of the converter waveforms 2.10 i erawm under ste ing in eyullibvium. Let's consider next what happens 0 the inductor cutrent when the converter i fist zed on. Suppose thatthe inductor current ad output voltage ae initially 2st, and at ce, is thon applied As shown in Fig. 2.11, {08 7er0, Dating the fst subinterl, ith the ‘witch in position 1, we know thas the inductor cuerent will increase, with x slope of (V,~ vf an with ‘initially zero Nos of vib since vis iiilly zero, ths lope is essentially ze. It ea be seen th there is amet inerease in inductor The inductor eurent waveform of Fi dy-state conditions, with the current aver the frst switeiing peso, Iecause 4(F,) i greater than (De SinTT rower? Fig. 211 Inductor curent wwvelorm during convener tron tse ‘lows to the output the output capacitor will ehagpe slighty, and y will increase slighlly. The process repeats during the second und succeeding switch nt increasing daring ch subinterval {and decreasing during As the output capacitor continues to cha peviods, vith dhe inductor eu cach suhinterval 2 ses, the slope during subinerval rive. Eventually, the points reached sshere the increase in inductor eurtent during subinterval I is equal tothe decrease in inductor current id, r)= in HE) ing subinteval 2 becomes more ne, luring subinterval 2. Thee is then so wet change in inductor curzent over a complete switching p andthe converter opertes in scaly slate. The eonverter waveforms are peal: jn From this point on, the inductor current waveform appears as in Fig 2.10, The requirement that, in equilbsium, the net change in inductor eurent aver one switching leads us Wo a yo Find steady-state conditions i any switching converter: the principle peril bt tndictarvole-secrnd balance. Given the detining relation of sa inductor i en iF Inegration over one complete switching period, say from =O tT This equation states that dhe net change in inductor current aver one switching period, given hy the le hand side of Pq 2.18). 1s proportional to the integral ofthe applied nuctorvolage over the interval stealy state, the initial and final values ofthe ict curren are equal, and imegral of the applied inductor val ence the left-hand side of must be Fg. (2.1818 zero, Therefore, in steady sated The Fight hand side of Bq, (2.19) has the units of vot that the tol area, oF net volt seconds, under the», (8) waxeform must be zee, 19) staves conds or flus-linkages. Equation Aw equivalent form is obtained by divi both sides of Eg (2.19) by the switching period 7; The righthand side of Eg, (2.20) is recognized as the average valve, ode component, of v0) Equation? ¥,-¥ Total area h 20) states tht, in equlibrium, the applic inductor vol The inductor vos mast have zero de compo in ig. 212, with the e waveform of F 1. The total The ave valu is therefore d= d=p[v,vjeo[ vy) By cquating (»,) 10 ze, and noting that D = 0 o=py, [pepe =ny, 4 Solution for ¥ yields vepv, Which coincides with the wesult obtained previously Eg, (2.3). So the principle of inductor vole-second balance allows us to derive an expression forthe de component of the converter output voltage. An pplied to any converet; One simply sketches the voltage waveform, and equates the average value 1 ze, This method is used lar in adsantage ofthis approach is is generaity—it ean be applic induc tw sole several more complicated comer Similar arguments ca be applied te capacitors. The defining equation ofa capacitor is ton of this equation over one sitehing peri yields fou (226 In steady state, the net change aver one switching period of the capacitor voltage must be zero, so that the left-hand side of Fy fore in equilibrium the integral of the eapacitor current over one switching period (having the dimensions of amp-seconds, or ch The 16) 1s equal 19 20r0, The aould be ze. is no net change in capacitor charg in steady state, An equivalent statement isThe aver ‘or de eorapontent, of th Equation (2.27), ealed the principle of eap be used to find the st 23 BOOST CONVERTER EXAMPLE The ducing ade output voltage great boost converter, Fig. 2.13(a), is another apacitor ew This should bean intuitive resul, Ira de cures is pied Wo at explo them th ill increase without bound. Likewise, if de voltage is applied to an svihout bow large balance, ean I-known switched: mode conve in magnitude than the de input volt must b zon in eur, ra will nere A pra that is capable of peo ial realization of the switch, using a MOSFET and diode, is shown in Fig, 2.131), Let us apaly the small-sipple approxima: tion ad the principles of inductor volt-second balance ad capacitor charge bance to find the steady state output vol With ‘itch in position 1 the in the network of Fig. 2.14(a3. The inductor vol und inductor cura for this converter hand side ofthe inductor is connected to Use of the linear ripple approximation,» = ¥ deals to Fig. 243 ose comers (a) with eal teh, (>) rial realization using MOSFET and diode oa, ses jpacitor eurent For this subinterval areFig. 2.14 Boost omsser isu) whi the sich is postin J (bie the switch iin position 2 With che yoiteh in position 2, dhe inductor is connected to the output, leading to the eiteult of Fig nductor val 2a) {Use ofthe small: ripple approximation, v= Vand = f beads 10 k Equations (2.29) and (2.31) are used to sketch the inductor volrage and eapwcitor current waveforms of Fig 25, Fig. 215 Boost comerter vo — vr —| vv tO ravi s— pr, +} or | VRFig, 246 Ds comerion sti £0) of the bent canner 02 awe D Tecan he inferred from the inductor vot han the input vokage V, During the fist subintorval, «(9 8 equal © the de inp sotage V, and postive volt seconds are applied cathe inductor. Since in steady-state, the total volt seconds plied ine val-seconds must be applied during the second sub se waveform of Fig, 215) thatthe de output voltage Vis gre ‘ser ne switching period must he 2670, 1 dnterval, Therefore, the inductor voltage suring the ssvond subinterva, (V, — V), must be negative, Hence, Vis greater than V, The tral vol secon’ applied 0 the inductor over one switching period are [vJor. +(e, vor, ex iy cai his expression to zera ud collecting terms, one obtains vaD+D)-vo=0 23 Solution for Vand by noting that (D + D")= 1, yields the expression forthe output vol 230) “The voltage com fom ratio MUD) isthe ratio of the output to the inpat val Equation (2.34) predts that dhe voltage camversion rai mio: f= bet 235) This equation is plotted in Fig, 2.16, At D ‘output voltae greater than the input voltae. There are, of course ints to the output valk produced by a praticl boost eave fn the next chaptes, component Hoa is found that she maximum output voltage of a pratieal boost eonverter is indeed Limited. Nonet large output vollages can be prstuced ifthe nonidealities ae sufciently smal Tipe de component ofthe inductor curgent is devi During the fist subinerva the capacitor suppl val the ind ny use of the principle of capacitor change the load current, and the eapwitor i partially current supplies the lod aad, additionally peri s fond by inte xd, During the secon subi evs the capsicitor. The net change in capacitor charge over one stein the ste) wavetoum of Fig. 2.1500),Fig. 217 Vivian of inductor cuent de ompomentFvth dy cycle, boost convener 6 4 a a 0 02 uk oe ust D (236) Yor, [1K )or, CColleting terms, and equating the result Zero Feds the steady p=p'|+s=0 Bby noting that (D + D and by solving forthe inductor ew ide component fone obtains So the inductor current de compone Eg. (2.34) to eliminate V sields *, Substitution of v wR This exqation is plotted in Fig. 247. Tees he approaches that the inductor eusrent becomes a This incor curent, which cineides with ‘than the Yoad current. Physically. his must be the ease: to the extent thatthe converter elements ae idea, the converter input and output powers are equal. Since the converter output voltage is greater that the ingut voltage, dhe input eurtent must ikewise be greater than the ouput euevent, lo practice, the inductor nt flows though the semiconductor Farward voltage deops, the inductor win istanoe, and power loss. As the duty cyete approaches one, the inh at hovomies very Tange and these component nanidealities lead 1 large poster losses, In eansequenee, the efficiency’ of the hoost 2ses rapidly at high duty eye 1 de int eusrent dhe boost converters converter de Next et us sketch he inductor current () waveform an deriv an expression forthe inductor ccuront ripple i, The inductor voltage waveform (21 has been slscady found (ig. 2.15), so we can sketch the inductor current waveform directly, During the fist subincerval, with the sich in position | the slope of the inductor curem is piven by Likewise, when the ssitch isin position 2, the slape ofthe inductor current waveform isFig, 248 Boost comener ind 18. Du tor current, 24%, seul to the slope malted by the Teng “The inductor eurrom waveform is shetched in Fg Solution for Ai, loads te chata given value of Ai, is obtained This expression ean be used 0 Seloct the inductor vale Ls Likewise, the capseitor voltage uff) waveform can be sketched, and an expression derived For ‘magaitude 4e; The eapacitor current waveform lf) Is piven in Fig, 2.15, aimestoncortiecs he output voltage ripple pe itor vai Daring the fist subinter uge waveform vt) is ae Ducing the second subintersa the slope i tity ttt bas y eC TOR ‘The capacitor vollage waveform is shetchod in Fig, 2.19, During the fist subimerval the change in capacitor vollaze, -24n,is equal to the slope multiplied by the length ofthe subnerval: nave g or 2.5 Solution for dv yiekés 219 Howst comerter onwe capacitor value C vo obtain a This sven ou voltage ripple peak agit dy 24 CUK CONVERTER EXAMPLE As a second example, consider the Cuk eonverer of Fig, 2203) This conve Siow funtion similar to the bck-boost converter: tea ether increase or decrease dle volige, and it inverts the polarity. A practical alization using a transistor and dnd iby sitive energy transfer As illustrated in Fig conngcted though L, 0 the input source while the switch sin position 2, and source ene C, When the switch isin postion I, this energy is released dough J to the load sci volt The inductor currents and * are defined, with polarities assigned somewhat arhiarily. in Fig. 2.204) In his seein, the principles of inductor volt second balance and capacitor se balance ae applied to find the de eomponcats ofthe inductor cureats and eapaitor volt voltage and During the fist subinterval, while the steht is in postion 1, the comerter urzent ripple magnitudes ae also Fou cult reduces 10 2.219), The inductor voltages and capacitor currents are » t, « | am [et iG he yp & Fig, 2.20. Cuk comer) with eal site, (by prstcl realization using MOSFET an die2s inp of ead Sate Comer Anayie Fig. 221 Cok converter ete: a) while switch sin poston 1, (0) while switch sin position 2, We neat assume dht the snitehing ripple my liudes in fy, (8, (and v.49 ave small eompaned © fore make the small-xipple approximation, their respective de components J, FV, ad V>. We can d and Ba, Qst8) becomes During the second subintorval with the switch in postion 2, the converter cic asin Fig, 2.21¢b), Phe inductor voltages un capacitor curren ae We ain make the smal-sipph Equations (2.49) and (2.51) are used to sketch the inductor volta 1 waveforans in and capacitor cure The nest step is to equate the de components, or average values, ofthe waveforms of Fig, 2.22toy J— or —| v9 » t | Fig. 222 Cuk comener waveform [pr fe fa) inductor votage v6. (6) inductor : setae vf (eh eapacio eure eth wow (dh eapsctr eunent ic fel ; w —— pi Dh to 2e10, of the steady-state conditions inthe convener. The results ne =v,+0(V,-¥)) =0[-¥-V] +8} fo) =D + 07,=0 label Yeo Solution ofthis system of equations for the de components of the capacitor vollages and inductor eure woot EUOR Von the duty cycle Dis sketched in ched in Fig, 2.2413) and 2.24), and the eupacitor C, The dependence ofthe de oust vot The inductor eurromt waveforms ate skFig, 223 De conversion catia 1 SD) = Viv, ofthe Cok eomerer My Db C v MiDr= =~ (Py voltage waveform » (fis sketched in Fig, 2.24(e), Dusing the ist subi val, the slopes of the wave: Fors ate given by wt aie est Equation (2.49) has been use here to substitute for the values of yy, ane i, during the first subiter val, During the second interval, the slopes af the waveforms are given by {a} inductor curent I(r), ¢b) inductor On. ! .cond sulbinte Fquation 2.51) was used to substitute For the values of», 3. ANd ig, Sing th al. During the first subnterval, the quantities 49h it, and wf) change by 2Ai,. — 28, and = 2, respectively. These changes are 4). multiplied by the sub. ual co the slopes given in Eg inesval length DP, yielding The de joaships, Bq, (2.53), can now be used to simplify these expressions and eliminate V,, Vand “These expressions cam be used 10 select values off, fay ad Cy stich that dested Values of switching Sgple magnitndes are obtained Similar segments cannot be ese to estimate the switching ripple magnitude in the output tw Fig, 2.224, the eurrent (9 is continuous: unlike weitor vollage #0), Aceardin 'snonpulsating Irthe switching ripple of) s neglected, then the capacitor the capacitor eurent jf?) ceurtent (1 doesnot contain an ac component, The smal ipple approximation then sion thal the utp sitehing Fipple Avis zo, oT estimate the magnitude ofthe out ‘pple present inthe inductor cur only sourve of ae eurtent diving the output capacitor Cy. A. erand in oth OF coutse, he output valage switching pple is not zs ‘wo must not neglect the jut so Fipple in this conver Feil ft, since this current ripple isthe ‘Similar converters is discussed in the neat sec simple way of doing tis the Cuk co 25 ESTIMATING THE OUTPUT VOLTAGE RIPPLE IN CONVERTERS CONTAINING TWO-POLE LOW-PASS FILTERS A case whete the sal pple approximation is mat use sin converters containing two- poe low-pass For these Fillets, suchas in the up of the Cuk converter (Fig. 2.20) othe buck converter FigThe buck eon vipp ripple, egandless ofthe ‘he ourpu filter egpactane, The problem i thal the only component of oupat eapacitor eure orcurtent rile, Hence, inductor curent ripple cannot be ‘ouput capacitor voltage ripple, and Fis, proximation preicts zero op vl ceases is that arising ftom the ind ue approximation is puto forthe inductor current ripple. The expacitor voltage ripple Folate tothe total ebuege contained in the postive portion ofthe it) waveform, Consiier the buck converter of Fi component and linear ripple of peak magnitude Ai, a8 shossn in Fig reat wavclorm san thea be 225. The inductor curent waveform i() contains a de 2.10. The de component 1 mast Yow emtigely thougth the load resistance R (why, while the ac switching ripple divides between the Toad resistance B® and the filter capacitor C, Ina well-designed converter, which the capacitor provides significant iltecing of the switching ripple, the capzetance Cis chosen large enough tht is impedance ithe switchig frequency is auch smaller than the loael impedance R, Hence nearly all of the inductor current ripe flows through the capacitor, and very litle Hows through he lod, As shown in Fig, 2.24 the eapaeitor eutrent waverorm (is then equal tothe inductor current wavetogmy withthe de rent removed, T urret ripple is linea, with poak value i Sted on th hetween the te zero-crossings of the capacitor eusrent 2. The waveforms ripple, of 2 When the capacitor eurret fr voltage ¥e10 inc Ys positive. charges dep e capacitor plates and the ses, The tage changes betwen its minimum and masini ext in is the peaketorpeak output voll in capacitor voltage can be elated tthe tot waveform, the capacitor vo symmetrical, and the total cha 1 contained in the postive igo Total charge Fig. 2.26 uur capacitor voltage al core selon, fir the buck convene in Fig. 22525 Eximotin ee Ou i Convers Contig TP Low Poe Furs 88 OF renat fu tinkage x ig. 227 Estimating inducor erent ties ‘ ripple when the indactor stage i TR T foe is contin, br br. i es in hat 1 tw Asillustrated in Fig, 2.26, the charge g isthe integral ofthe current wave For this example, th COvsing to the symmetry ofthe curent waveform, the integral can be expressed a the a height i, crossings aceur at the conterpoints ofthe DT. r P99 oda Sulpsctuion of Bg, (2.58) ino Bg, (2.59), and solution forthe vol ripple peak magnitude Av gills si 60) alll This expression can be used to sect value forthe capacitance C such that a given voltage Fipple Av is joblained. In practice, the additional voltage ripple caused hy the capacitor equivalent series eesistance (est) must also be includes. Similar arguments ean be applied to inductors. A example is considered in Problem 2.9, in which a to pole input filter is added to a buck converter as in Fig. 2.82. The capacitor voltage ripple © conclusion tha no a voltage is applied aeross the input sulting in zero input curent ripple, The actual ductor soltage wavelet i ential the ae portion ofthe input filter capacitor voltage, wih linear ripple and with peak value Av a illuteated ia Fig, 2.27. By use of the inductor relation i= £3, result similar to Bq, (2,60) can be derived. The der= ‘vation is left as a problem forthe student ‘sannot be neglected: doing so would leu 10 siker inductor,