ms CHAP 1: Basics of Control Systems 3» oe eee System ¢— oF Asie isan arrrangmnent or combination of ditferent physical coe Suchthat it gives the proper Olp to givensip- A kite is am example : Physical gis. ecoz itis made up of paper & sticks: Actassraom is 29 xa™P ofa physicalsys- Contro} s- The mean meaningof contyol is to requate, direct or command a sys $2 that Aesived objective is obtained. Plant $- THis defi ined as the portion of sys. whichis to be ontrolled oY vequiated- It isalso catiedas o. process. Input - the exc excitation signa) that is applied 10 a covntyol sye- to get a specified IP led Os dp. Ovtput (o1p)8- The actual vesponse that is obtained from a, control sys-due.to the application of the Jjp is termed as Olp- Controller ¢- Tris the element of the sys: itself, or may be external tothe Sys: It controls the plant oy the process. Disturbancess— The signal that has some adverse effect onthe value of the olp of @ SyS- lig called disturbance- Jf a disturbance is produced within the sys-,it iS Htermed an internal disturbance otherwise itis known asan extemal distubamre- Ht Control System §- A contol sys. iS a combination of elements arranged ina planned ‘manner wherein each element (uses an effect fo produce a desired olp. ‘reference iip wu) |__, ntrolted onto] System al ct)tives] « Acontiol sys. must have ti dip @)o1P Carways toachieve LiP&olp obje () control action. 4 Classi ficatlonof Contiol System 9° contro system pen loop control SYS" cloged loopcontvol SYS (oucs) ccics) positive ftedback negative fredback Control sys control SYS 4 Open Loop control System o- srmolated under the basic condition thatthe atisa conditional control sys. fo n ys is not subjected toamny tyre of disturbances: Control characteristics of guch sys ave independent of the olP of the sys: The olP isneithey measured norfeedback for comparison with thelP- ap + {ea ws [ai | (© Automatic coffer sewer Bx: @) Traffic light controller @) Bread toaster ) Electric wasting machine » Advantages 8- + Disadvantages 3- 1, simple & economic LT Inacurate 3. unreouisable 2. No stabi lity problem 2 The effect of pareneter variation &erterna nwise is more. NOTES §- + performance analysis is not applicable 40 O-t:CS: becoz they ave highly stable sys- Faithfulness of an open loop control sys- depends Upon the accuracy of Alp calibration.+ Closed Loop Control Systems o- {Control characteristics of the syS: depends upon the olpof tne SYS: “At isalso termed as feebback conto} sys. The ontiol action is actuated by an error signal’ C= 4P-olP) * The purpose of feedback is £0 reduce the error bjw the reference di P&SYS OIP- + Sonu ats) ip —¥ Gs) HIS) Exarnples 3- @) Electric iron —_(b) D¢ motor speed contro} (c) Radar hacking sys. @) Avtopilot SYS) Economic inflation G) Hunan respiratory S¥5- ) Amissile launching sys (divection of missile changes with the oration of moving nse) Advantages - (@) Accwate & Reliable. (b) Banduldthof the sys: can be ted with ve feedback ©) Reduted effect of parameter varlation () Reduced effect of non-lineavities Disadvantage s- @) The sys. is Complex and costly) SYS. may become unstable © Gein of the sys. redvces with eve feebback- Effect of negative feed back ¢- 1. Effect of parameter variations reduces. 2 The gainof sys. yeduces by a factor ( 1t1H) - 3. The bandwidth of the sys- increases. 4. Effect of intemal disturbances reduce - Remembers- b> Feed back is Not used for improving stability. » Anopen loop stable sys- may also become unstable when Ove eed back 1s applied: P Except Oscillators, m@ve feedback ,we have always unstable systems:—- + Comparison blw Open loop & closed loop control Systems 5~ Closed loop C'S: 4, Ag the error byw refeverre Lp &the ip is continuously rneasored ttnough Open loop C's- 4 ‘The accuracy of arn o-t: sys- depends onthe calibration of the dip- Any departure From pre-determined eebback oF dveto feedback, cLes calibration affects the olp- igmore accurate: 4 O-L- SYS ig simple to construct & cheap. | & CL SYS: ig emplicatedto construct & costly. % 1-sys.are generally stable n 3. cL-gyg.cam become unstable vrder certain conditions: (stability deperds on operation. sy$- components) , |4- Comparatively the performance Is betley 4. TE pon-lineaity is present, sys- rma thom o-t:sy$- if NON Jinearrity is present operation degenerates: The transfer function of a tinecrr, timne-imvaviamt differential equation sys. ig defined asthe votio of Laplace transform of the ofp to the Laplace fang form of the Jip. under the assumption that alt the mitial conditions ave Zeros. « Acys is sald to be linear if princi pleof superposition & principle of hovogeneity applies: © A differential equation is linear if the coefficient are const: or function only of independent variable: Open loop tranefer Function ¢- crs) RO) att) = Trmsfer function, T= GUS) dip eH) sip ctl) oe Ree 4 (Ampuse dip sues), TRE ((S) = Jmpulse Response, Tmpulse response of asys- , GS) = TF g= cl) NOTE 3~ * Tp. of a Sys. Is the Laplace transfovm of ts hmpulse vesponse + TF. ig dlways calculated with zero initial conditions. + while deriving TF. of a [i saith a . eee ee* Asis. with zeroinitial conditionsis ald fobe at rest since there is NO stored en ergy. Closed loop contro] System TF. 3~ — Rls) GIS) as) Transfer function, re. = OS) 1 GIOH(S) He) # Poles & Zeros of 0 Transfer Function 2 Let THR = GG) = US) -KC$#2,) (S424). 3 Ganfactorof FF: 615) ROS) “Tein (SP * TF numerator poly nomial wots $=-2,-2, ~~ ete ave called aS E705: * TF denominator polynomial roots S=-P.s-Pa -----ete ave called as poles * The value ofthe TF at azevo is 2ev0 and at a pole js c2- * Real zeros and poles will lie onthe ea) axis. * Poles and zeros ave always symmetrical w-rt: rea) axis. * Complex zerosand poles will tie as shown infig: * Response of the sys: depends onthe both the poles &zercs- * Characteristics of the sys: depends only on the poles. * Denoninater polynomial of the TF is called asthe chnacteristic potynomal, Chanacteristic equation cE = (stp)($4Pa) (S#B)--- = * characteristics equation yoots = poles of the sys pole-zero plot’ Plot obtained by locating all poles and zeros of a TF ins-plane is called pole- zero of a SyS- Order of a Transfer function : eine : . The highest power of 's’ present inthe c& Ce inthe denominator polynomiat of a. closed loop TF: of a sys: is called ‘oyder’ of a system: @ for to frsdt =1 0 for azo! -o NOTE: Unit Impulse function &)* —gey= { Lartace transform of impulse vesponse of a near time invaviant Sys-is its TF witt athe initial conditions assumed 40 be 2er9.= eS 1 | tt Properties of Transfer function g~ 1, TF ig defined only for o. ineer time ~imvarianl sy5+ or time variant systems: a Fes GS)= LOI LCatP) 3. an itial conditions of the sys: ave set fo zero" «TR isindependent of the dip of the sys. 7 . TE the sys: TF. has no poles Ov Zeros with Ove real paris, the sys 15 uimitnum phase system: Non. wrippimum phase functions ave the functtons which have polesov Zeros on tight hand side of s-plane- . 6. The stability of LTE sys. can be detevirined Aromits characteristic equation. ce 4 denornirator plynomial of c-L-FF to zero Ttis notdefined for nonlinear] at # Advantages of Transfer Function $- 4. Tt gives simple mathernatical algebraic equation a. Tt gives poles & zeros of the sys- divectly: 3. stability of the sys cum be determmed easily: ‘A+ The ofpof the syS; for any Alp ambe deter mined easily. a pisadvantage of transfer Function ¢- 1 lsapplicable only for LT system. 3, 1t.d0es not take initia! condition into account: 3. The inferral states of the sys- can not be determined: 4. amalygig of multiple 4p mottipleo|p (mmo) sys: is cumbersome . 5. Controllability and observability an not be determined: Shortcut Methodss - 4. To find close loop transfer function from open loop troncfer function; re our. = Humerator then, Jour. Seen Denominator Penominatoy+ Numerator LY a, Tofind open loop transfer function frorn close loop transfer function : If «CLEP = Numerator then, | O-LTP = Numerator Denomrimator Penomivator- Numerator[se 7 # DC Giain for Open oop 5= let aTe having p novof polesat origin be represented as, nego) nin const TF = R(S-5)(5-53) Sn EVE) ~ (&") form SPCS-Sa)CSSp}—(Fm)— Par gy. Then its, lcgain = Lim gf. KsrSx by (2) (8 S S90 SPs Spe-e- im Ee” Ee) # Procedure for Determining the TF: of a, contol SYS: 5 step 1: formulate the equations for the given sys- —— ifionsaszew. Step &: Take the Laplace txformn of the sys- equation, assummmg initial cond step3: Specifythe sys aip & the tip- stept: Take the ratio of Laplace transform of the olp & the Laplace t¥form of the ip: 4 Interacting & Non-Interacting systems $- R 7 ap. = Bol) - MSC 2 1 n “Consider, ap. - Esl) = AE = ret Ren g _1T3 EWS) R+Xe. SCRAI e TG=Rc > time const. * Consider, Assuming zero initial conditions, ow ah 9 Ins-domain, iD) 4 d c & Using kvcintoop 1, RILS)4 d [A4S/HyI) = it) ' fa ! using Hye in loopa, gi,(s) +2 (1y5)-1¢6)) =o} (ssi) =-Bol8) Re ckts in cagade After solving ire: ExtS) = T i foaee , : Se sttaeer af oP amaar he 3 time ¢ BS) SRerasRet! Fe 43st4! v Here , TF of each of the imdivideal Re cht ot Bot, His seen that overall TF of the a AC ckts connected incascade + product of two tL i itis A * | gan gan] but instend it is aha + This difference is explained by the fact that while desing the TF. of single Rc chts, itis a ssomed that theo}p is unloaded: However ,when the jp of ah ckt isobtained from the op of First, a certain amount of energy is drawn From the 151 ckt and Hence its original TF is no longer valid: Thedegree totwhich heoverall TF is modified from the product of indi vidval Tes depends upon the amount of tondingHence “Itcan be concluded that when a time const elementsave cascaded interactively, the overall TF of such anangment is notthe product of 4wo individual TF,dueto Loading effects. * Now consider, , 1 Vy($) - Le , Vy) x T ; Vx) = My) = : I vats) SRC aa BO %eae yo yg 19 wo ~ yd Vyts) SACHT ots) K overall transfer function [TF sig) = Gece 1-7 of such en two 64S are connected non-interactively, the overall TF OF ce of loading effect: So wht mentis product of a ‘individual TF,dve to absen arrang 3 _cHAP.3: BLOCK DIAGRAMS), A block diagram 15 a pictorial representation of the given SYS «+ for a.closed Wop sys, the function of comparing the different signals is . i He for the feedback purpose is indicated by takeoff paint in block diagram: © The cormection bjeo the blocks is s! + An Arrow is associated with each and every branch which along the branch: + Ablock diagram of a control SYS: explicitly shows 0. unilateral property: «A Block diagram cam bevsed to model linea as well as non-lineay: + Block piagram forgiven sys: 19 rot unique: own by ines called bramches of the bloc Alagran} indlcates of Flow of signal Tateaft print sonmingeimt glock diagain of a. Closed toop System o- AI ao 3 Forward path TF wg) > Feedback Path TF Gus) HIS) = 8 open top TF TEs) 9 S13 closed op TF RIS) Gus TIs) 1+ Gus) His) £6) «cay 5 4 Senor voto, a primnany feedback ratio CAs) ECS) 405) (9 = RDB) = RO) -HEOC(S)Multlple Jip- multiple o1p Systems b= ated independent tion Le olP ‘hen multiple Jipsare present ma tinear sys.,¢ach dtp can be tH OF the others. completeoip of sy& can then be obtained by evPerPOs! _ OFesponding to each dip alone ave added tagether: koser, FS! Assume vrsyz0 Cals) = o}pdue to Alsyacting alone Cats = _Suls)Grts) ae BB) [t Gu) (8) HS) Guts) RIS) aser He) Fase: assume Rls)=0 Gig) us) (5) = op due to US) acting alone Cl$)= —=——_. le Pee 1+ 605)1D 15) Now overall Tesponse, cIs)= Cals)+ Cyls) 3] cg) = bial? _ [t1is) RISD + i) [464(5) Gin) HIS) Procedvye 10 solve Block Diagram Reduction Problems ¢~ OS aes eee eerie stepi: Reduce the blocks connected in series. Stepa. Reduce the blocks connected in povallel- Stepa.. Reduce the minoy internal feedback loops. Stepa: As far as possible try to shift fake off points towards vight and summing points to the left- steps: Repeatsteps itot till simple form |s obtained. stepe: Using standard TF of simple closed loop sys, obtain the closed loop TF we of the overall system.Transformation Leomblring bcksin cagrode 3. Remaving a blockfrom a. forward path 2. pining blocks in posal] Eavt Block diagram 4» Elim naling « feedback loop S. Removinga block from O feedback (oop wo — © Rearranging = summing points 0 rs x40, ¥ =: moving aSorming 7) % yO z point before a block > 7 a : : 8. moving a summing ee om Pomt beyonda block Vt 9. moving takeoff point before ablock 0. moving atakeoff point beyond a. block pie ne movingataheoft point before a Summing point 5 “J tavng atakeoff pomt beyond.a summing point ~é< Ss isBOE yy CHAE: SIONAL Flow dignrtls (SFO) % ng of povtaving the praic equations: co of sys: variables: “A Sioral Faw Grraph (SFA) Tay be defined as a. graphical mea Jp “OP relationships blw the variables of aset of lineay alge * Hisanapproach which does not vequire any reduction process be * Diailabiity of a Flow graph goin formula.whichrelates the Jipaolp 1. Asrer applies only to linear sys- A 3 The epatione foruhlcha srevisdvoun must be algebraic uations 7 Me ‘m of couse & effect: 3. Nodes are used to Tepresent variable: Novmally.the nodes are arranged #707 left torignt following a succession of cause andetfect relations through the SYS: 4+ Signalstravel along branches only inthe divetion described bythe arrow oF ‘the branches: S+ The branch directing from node yy to Ys represents the dependence of the Variable yj upony, but not the reverse- & A signal Yq traveling along a branch blw nodes yp & Ys | gain of the Bramch, ax; So that a signal ay pis delivered Definitlons forrsFars - Tip node (Source), a node that has only outgoing branches- Olp node (Sink) + a node that has onlyincoming branches: Path: Tis any collection of a continuous succession of branches traversed inthe samme direction. ' Fosward Path’ It ls 4 paththat starts at andip node and ends atano|p node ard alorg Which No node is haversed more than once: Loop: Itisa path that originates and terminates onthe same node &atong which no other node 1S encountered more than ence: Pathgain: The productof the branch guins encountered intraversing a path is called the path gain. Forward Pathgain : 14 is defined as the path gain of a. forward path: Loop gain; Itis defined as the path gainof a toop- Non- Touching loop Two parts oF SFO are non- touching if they do nor shove a common rode. Self loop: Tt is a loop consisting of a single branch. is multiplied bythe ad node YjRules for Drawing SFa From a given Block Diagram ¢- “Hy A L.While drowing sF Gi froma, gvenblock diagiam,the adjacent summing points & Joke otf points (but not a takeotf point preceding a gumming point in the ivection of sF61) ave vepreserted by a node andthe block TF is represented by aline joining the vespectlve nodes: The divection of signal flow is trdicated bya now onthe tine. 2. TFinthe direction of sigral flow, afakeoff pint precedes a summing pont then Such points are represented by two seperates nodes witha hransmitiance of unity bw them Ovigina) Block Diagram Equivalent SFG wes___foisy }— 8) a8 8) at) BOTS tts i as) ee wR HIS) HIS) NUS) Lg HHS) Yt A) cre) Rts) +H) 1) Nts) 1 a ris) x ets) U8) 1 ts) R_ 624° Tous) =H) Gal) _ Ral) {etSignal Flow Graphs Algebra 3- , ofall 4. The valve of the variable represented bya node is equal 0 Lid ; the signals enteringihe node: for the sft of figute, the valve Ot et equn)tothe sum of the signals turmitled throvgh all heincorniT . thotis, Ys ay ¥n 493143 + 94144 + OGI4S- . i 2 The value of the variable represented bya node is txmnitfed through a branches leaving the node. Tnsfor figwebove, Ye= 044, .YyF ary » Yor 104 3 Poratlel branches ‘nthe same divection cormecting two nodes Gm be Lees bya. single branch with gain equal to the sum of the gains of the pavalle bramehes: . 4. A seties connection of unidirectional branches canbe replaced bya single branch with gain equal to the product of the branch gains: on aa 8 o——+ oF : eee % Ya Ys Mason's Gain Formula $- # the relationship blw andjp variable and ano}p variable of a SFG isgiven by the net goin bw the dip and olp nodes, is Hnown ag the Overall gainof thesys * tls much less time consuming to write down the dtp-olP yeratlonship by inspection from the original SFo by using j Overall system gain,or TF; WIDE, p= Path goin of KP forward path A = 4- (sum of loop gains of ait individual loops) : + (som of gain productsof alt possible Combinations of two non-touching cops ) ~ (sum of gain products of all possible combinations of tree nen-tovching oops) + Ay = valve of A for the partof the graph not touching with kth forward path *A Is called seordeterminant or chavacteristic function, since, azo isthe sys. characteristic equation NOTE cove must be taken while applying the gain formula tornsure that if is applied bw am ijp and anojp nodest0$- i ) WMS yy CHAPS: FEEDBACH cHaRACTERISTICS SY | * thefeedbad signa es this 5. the capably act as safeometig ‘mectanish | Beneficial effets of feedback 4. the controlled variable accurately follows the desired valve. 2. Effect on the controlled vaviable of external disturbances ofher than those assoclated with the feedback sensor ave greately reduced- 3. ERect of variation in conolley and process parameters (the forward path) on sys- performance isreduced to acceptable levels: These variations ocr due to wear, aging, environmental changes etc. feedback inthe control loops allows occurate contro! ofthe ojp even when process or controlled plant parameters ave not HnOWN accurately: 4. Feedback im como} sys: greatly improves the speed of its 7esponse: NOTE ¢- © The cost of achieving these improvernents ‘in sys’s performance through feedback are greater Sys- complexity » need for rnuch larger forward path gain and possibility of the gys-ingtability (Undesived | persistent oscillations of the ofp variable )- © Gienerally, negative feed ave bydefauit assumed wl that positive (eve) feedback is there: # Sensitivity & Let the variable in a control Sys- which changes its value be A such asolp &this change |s consideved due oparo meter variation of element H suchas gainor 4eedback then the control SyS- sensitivity is expressed as, hange INA A _ BAIA ~ 3A co | = ec oan 3° change ink Sq = de RS back isso extensively Used that closed loop contol sys itheve feedback unit it 1s specifically mentioned Sen: sh 3 sensitivity of variable A wrt: parameter K - * Tt isthe relative variation inone parameter w-v-t- other pararreter: * THis preferable that the sensitivity Function si should be minimum.—t oe Sensitivity of overall TF Ts) wrt: Forwerrd Path TF OHS) 6 SedSitivity of overall TF 71s) wrt: Forwerrd Path TF OND! The sensitivity fonction for the overall Functlon Tis wrt: variation 6, ste Yr. BT) Gus) AT “Tous ~ VOUS)” TS) 1 Open loop contro! systerns- ts) RIS) T= 4) = 9 {GI} © Fm 8) oT) - TS) - TL WG), dS) (eae) 36K) 1 & tet $0 Sy= 7 Das) 3. Clnsed loop contol system &- 9 ) — T1e)2 LHS) #4 THOUS) HIS) Sus) _1_____, BU). 6) = Fhreusynts ce) GS)“ Fit catsonisy]* 1S) Gus) ROG HES) 30, ol (lousy Higy]> 3 |g = 1 Cit us) Hess)? “Gy Ty Guts) HIS) NOTE: Closed loop sys: is less Sensitive to variables inthe forward path TF Gus), compared to the open loop sys- Genxitivity of overall TE Tis) wert Feedback Path TE HIS) $-Linclosed loorsy Sengitivity of overal) TF Tis) w-ret- Feedback Path TF HS) T. S19/Ts) = HO Ts) H TRO Hs) Te) DHS) Tee) = _ 6418) amy 2 - Gate)? 14 04S) HIS) OHS) Tiros}? NOTE: The closed toop control sys. iSmove sensitive 10 variations in feedback path parameters than the variations inforwnd path patameters, there fore the specifi cations of feedback elements in a cloged looP Contvo! sys: should be more rigid as comparsed to that of forward path elementsH Effect of Feedback on stability §- 1 *S2 O if Get then opottpesys: = forany fivite Lip. mstable sys. WOH * Feedback can cause asys that is oTigmally stable to become Unstable: Certainly, feedback ig a double -edges sword, when itis improperly used, it Cm behormfvr. * The advantage of incorporating feedback is that itcanstabilize an unstable sys. © the bottomline isthat feedback com improve stability or be parmnfol to stabi lityif it isnot properly applied - Remember $- « primary purpose of using Feed back isto reduce the parometer variations: The parameters of sys- may vary environment etc- © the closed loop sys, has higher bandwidth than oven Ieop sys, this implies ‘increased speed of response: « The effect of feedback on noise and disturbance depends greatly on those extraneoo Slgnals which occurin the sys. No general conclusion cm be reached, butin many itvations feedback can reduce the effect of noise and disturbance on syspevfoimaye censitivity of the SyS 1 withage, with charaing # control over System Dynamics by vse of feedback 3- Letus consider an ele mentary sys. shown in Fig: cage1: OpenloopTF of Sys, RG) cs) eH, TI6)= kK HS or sa forimpulse iP. R= 3, op C9 A >} fevene] cngem closed 1oop sys, 8) = KM, fir ADE 5 aye fo RS) SHAR sank + for eve valueof tk effect of the feedback istoshift the pole. Gvely to S=~(44K), aNd so the Hime const reduces: AS Ht sys dynamics continuously becomes faster i-e- transient response decays more quickly: be Control on the eftectsof the Disturbance Signals by the use of feedback when R(s)=O ip “distorblig sgnt RS) ono og) if ae TH) bats) 1 cs) OOHCo) - ots) TH) “Te gutsy Grat8) HO) iF [Hsroutsroatso] 2 then, Sox aL { nis De) GH? by feedback. Thus, if ous)is made sufficiently lage, effect of disturbance canbe ed PY HF Effect of noise (pisturbance) Signals 3- i 1 oH Gig toe CeCRHCy cae Guta ois by N HonGigh [toy 0H “CN > eHect of noise signal onolp, soit mustbe winnmised and best way todo this is to increase Gy sertothe Se the effect of noise cenbe reduced by ting the forward pathgatn eartiev?? Noise. Remember s— To conclude, the use of feedback has the advantages of relucing sensitivity. ‘improving tronsient response and minimizing the effects of distorbance signals in contro] sys. On the otberhand, the use of feedback increasés ‘the no-oF components of the sys, thereby increasing its complexity further it reduces thegain of the sys and also intodvces the possibility of instabilityee ee ae SD CHAPG : Mathematical Modelling of Conol Sys. >> 7a SOE oie nese eee ere The get of methematical equations, descnbing the dynamic cheracteri stics ofa SYS. iscaled mathematical model of the sys- Hemechamial systems ¢- AN mechanical sys. are divided ito two parts: Translational Motion s- Ac to Newton's law of motion, sumof forces applied on rigid body (01) SYS. moustbe t velocity and £ forces corgumed to produce, displace men etementsof the sys- The following elements ave of tranglationa) motion systems: output 5 linear displacement exo linear velocity (v) equal tothe Sum 0 dominantly acceleration in various ‘involved into analysis Here, Input > force (F) (@) Mags Element §- eee Inertia fore, (b) Damper Element ¢- a x1 pamping Force, = damper const oa i ¥ m2 f v2 © Spring Elements - m1 ' eping tee om ae GE 2, Mechanical Rotational System o- Rotational Motion s- This is the motion about a Fixed axis: In such gys.,the force gets eplaced by a moment about the Fixed ori (-e. (Free distanre -romfixed axis) which is called donque: Here, Input }Torque(T) ovtput > angular displacement (6) or Angulax velocity tw) ait mermassof block fR) Inertial element ¢- Inertia. Torque,©) Torsional TPisional Damper Element 3~ Pamper nque = 498 = fcaray) tT 440 = §(wy-w,) 45-4 Fg. ©) Toysiona) spring ee 2. Spring Torque, 3} torsion aim} stfiness Tew Ext 8 Ex = 7 x ltr \y $ The yh Mo oth f A Tanipside T= Ta tat = Jd94 548 46 Fe Fiafatts = mdeasde ge dt? dt Mechan)ca} Rotational Systems Mechanica} Tramsiation of System HF Electrical Systems 3- 1. Series pic circuits 2. Parallel Ruc circuit + ep 9 L R 7 2 7 y D er 1M cp ee YFV4V24¥3 = val Rs 1 fiat Te H+] 41, = ca ye Lf uth Ve eben dat To chby 1 db 4 & ait a t FE Aralogous System o- Electrical equivalent of mechanical sys: is called aralegoos.ys- iv Analogy: ech emica) Elechial ] Tarslation | Rotation | Fove-Votage analogy [| fae- current cralgy | 4 $ R YR m a L . K K % A x 8 rs : 3 Tt +—— P v zHt Nodal method for Writing Diff Eq” of Complex techamlcal System 5- Steps: Noof nodes = No-of displacements stepa: Take an additional node which is arefevence node’. Stepa:_ connect shemass and inertia. ass elements anys Be the PAEiPle nade &reference node- Stepts. connect the spring and damping elements either blw the orirelle nodes or bw principle nodes and reference depending on Apeiy position: steps: obtain the nodal diagrarn and wnite the describing diff. eq” at each nde. NOTE $- ' if > one mass element = order of differential equation =2 vn mags element 3 order of differential equation = 2” iP + Apaint or node blu damper and spring mustbetaenas adisplacement. * Each mass must be taken as a. displacement. Glee Tin A gear tain ig. a mechanical device which transfers energy from one port of the Svs toother part without any loss: N= novof teeth on the circumference of geay wheel ‘T= Yadlvs of the geay whee] cm) T= torque (N-m) : = angular displacement (radios) [ E— to 4 an — = oy sent wh purpose of Greats ¢- ee : «# Gems are mechanical devices which ave used as inlennediate elements bl electrical motors (shaft)and load: © Glens are used for stepping up or stepping down either torque orspeed- © Gears ave mechanically analogous to electrical transformers: Nore 3- Mass ox mertial mass elements and spring elements ave Knownas consecutive elements becoz they dre analogous to inductor qpacitor in electrical analogy.¥ Servo Motors 5 ive position, The servo system ig one in which theo|p issome mechanical ei Velocity or acceleration: the motors used in servo sys are cal motors. Requirements of good sewo motor 3- Nequrrements of good sewo ‘motoy d 4. Linear yeationstipbw etectil ca} control signal and the rotor'speed over sP° Vey awide range. 3 Tt shoud have linear torque-speed chavacteristics- % Tertia of rotor should be as low as possible. 4. Ts vesponse should be as fast ag possible. 5 It shovidbe easily veversible: Types of Servo motors 3- Servo motors. Ac Seno motors —_D¢ Servo motors Special Servo motors Avmature Field Controtled Controlted, # Armature contolled Dc Servo motors $- Fo Da aetenst. fala t supa) ; \ O1P On Jon Fm O Block Diagrams $- Dynamic equations 3- elt OS i jako i Dynamic equationin' [corresponding equ nal Om(s) time domain in s- domain JvetS) © = 4) (Gla Fa) (53m * Sm) Narey= Rasa vada ats) = ney Sta y= Rit = Ryton | EyLSI= SK yO) = HetDfS) fe) Relating wmis) & vats) ‘at ia Rect S14) nt Sm) wnt) T= Rrda Ty) = Kr Tals) Tne: he Sains, 48m |156)= Syn Ost Sint) rid T= ap Haden [Tal SInhm(+ Sn) Cy->fach ens, Ty> Torque+ — : suitable for large rated motors T+ hag gmail time const: + Field controlled bc sewomotoy &- asin yg * Preferred for small rated tot motors t ee 7 Sie Ly " TE has lenge flme const: Is $ men Dynamic equations + Bock diagrams * @ Relating mts) & Vs (5) (b> Relating wm) & vg(6) Dynamic equation ‘m time dornain Vpe Rpg yds ters RyayS)+ SHI) r = Hgts) Tne KGat Tolsd= HHS Tenens fn Om in=inee ed Tt8) = $m Owls)# Sh yt) _ alS)= STyym'S) 44,4) (aya SLE ST) L T= indi mum .¢ motor is open loop while ammature controled 0c motor \s NOTE! Field controlled 0 armature controlled DC motors ae preferred over closed loop sys Hence Field controlled type © AC Servo motor : The AC Sewo motor is basically two phase inductlon motor: The olP power Power of Ac servo motor varies from fraction of watr to Few hundred watts - 4 Two Phase AC ServoMotOT &- Here mh (slope of torque seed el jo chavacteristtcs) ay PY eal uz Te srystalingtore | cal wore” Ve yp > No load Speed - Dmnamic Equations: Reference writing (ananicapatn Block Diagramns + ‘ime domain " pee eq | (eRe Oni) At " = Yet) =m Onis) - T(S)= $m Gyn(S)# HVE) s(stn) > (b) Retatng wy s)& vets): T=Tnd Sng fg dln [Ty Sn S64 cn Tm) wat) at S40 Orn) -Torque-Speed chancteristics: Wrqve- Speed Characteristics . Torque The Tovque- Speed chavacteristics of a Huo phase Luge induction Ynotoy ,rnainly depends on x x ratio. ; Ce ih wally Bs” “S(itT§) peed + Generators 9- se *Stherately Excited De generadurt Excited DC generator o_ : i Euan ne Ep kag ty 4 hey , A he Vee Aa ats 9 yore 9545) a C= Kady. > ets)=HgZH(s) Hg + Generator const+ * Seherately Excited De generator cormmected +0 load pete Ke) gh vets) oy mn ot AB 5 au: Wo) ws) aasts) Rete) syed pve a Yere- Ral, SVU9)= Bt5)- Rahs) Tachometer &- Techometer ig a speed traneduces used os feed back element in corivol sys- 9__L AL 9 om) SAGE 29 whee, sho # Taher const NOTE e- * Servomechanicm is electromechanical sys whose dipis electrical vottage and o|p is mechomical pasitton or uts time derivative: [+ Field control pe servo motor iSused for [ow power applications while armature Control DC seryorotor }Sused for high power appli cati ons: Ac servomotor is generally used for lowpower application. Tovque-Speed characteristics Is linear fot Ac servomotor wheceas it Is nonlinear fof general purpose induction ynotor: fox the fast restonse of the seryomotorsys, the inertia of therotor should be Hept ‘to mnimum: This 1s achieved by reducing the diameter ond increasing the length of the rotor:ae CHAP: Time Domain Analysis of Control Systems Sita SAT i ieeememre oe aoe TIME RESPONSE Time response of « control system means as t2 how a wittime when a specified Jip test signal is applied: tne Hime Tesponge ofa conto system ig divided m0 PoTHS: i Tronsient response lb Steady state response Transient part of time response reveals the nature of response& its speed. Steadystate part of time response veveals the atcvracy of a control system. Total Tesponse of a system, system behaves in acovdarge| faba foncion: Impulse function’ Step function: Ramnp function: qui) -03:8#0 a ne =AzvIN. - Puy set vin= fed 10 in Saat ae 2 ee Every transfer function ‘representing the control sys has certain type & ordey: The steady state cralysis depends on type of system: Ithe type of the system 150 piained trom open loop transfer fonction [oLTF) GO)HIS)- ‘The novof open loop poles occuring ak origin determines the trgof guste: Let WI HI= RCHSTa fy PFO > fovT¥pe-0 SYS: a Brits) pet > For Type-d sys Trongient state analysis depends on order of the Sys The order of sys: ig obtained from closed loop transfer function Cee) setae j a) the highest power of characteristic equation Le. 14 GU) HIS <0, determines he order of contvo| system: ‘HOTER Notoral time const. of response onlydepends on poles of the S¥S- aaa STEADY STATE. EPROP PROMS RST ATE ITF the actual olp Of a control sys: doring Steady gs) 6) [states deviates from the veference dip (ire. desiredorp), the sys-is said to posses a steady State evroy: = RS). T+ CHS) H(S) Steady state evror, \3 (intime domain) > (in ¢- domain) eg= dim. 4) egg = lim SEU) es 5 SAIS) O55 = ao Tees* Steady state error depends on two factor; { Type of dip applied we. Al) tly Type of system ce. sists) fore * As the steadystate error ig anindex of accuracy of a control s¥S there IT hood be minimum as far ag possible. crevic EAROR coctricien aprioN ERROR COEFF POSITIONAL ERROR CoE _ ais)? 2 FIP. Risy= & (step dip) HHP. RIS=A (Rempdip) Perabelic , = * Static positional error +static velocity error « static axcelarat? oe fel Toh s)HS + |steady state error, + steady Shale error + steady state E7707 , A ess = eq STEADY STATE ERROR FOR DIFFERENT TYPE OF STE nD Typeof | Enoy constants Fa et [4] system Pil ip [A te Rv Ka. PEt) ame f el 0 KO 0 th o i t > k oO A © ° A K 2 co aeee Co geee ra sen acin 2 ° A po o ssa $0, 48 S¥$ gain increases, Ogg decreases. ti] For Linear Time Tovariant (LT) Sys, the max-type no- is 2- Beyond type 2, the system exhibits non linear characteristics. Steady state error isvalid only for closed loop stable system Steady state evro7 ave calculated for the closed loop gyg: by using OLTF» By using the method described above, es of any linear dosed loop sys. subgectad, toanijp with order higher than the pavabolic function, canalso be derived Aheoretically.if necessary. The terms position evvor, velocity evroy & accelavation error mean steady-state deviationsin the o[p position. A finite velocity error implies that after transients, have diedout, thedlp & olp move at the came velocity but have a finite position difference: The evyb consts, He, Hy &Ha describes the ability of a s¥s to reduce or elivminate Steady state errors: . ‘As the type of systern becomes higher (ice. ting no: ofintegrations) progressively MOTE Ee are eliminated:md) > DANA C(onrenemnsr) Eon coEePCENTS 2 Dynamic error coefficients give the steady state error performance ofa feedback contiol syctern 10 amy amie Up function Ets)= _R)_ tet Fs : PEGS) HIS) Trims) 8° 8) RISFLS) lamas Retctionchip bw static & Dynamic Enor coefficients 3- v ogee Roe tie ake ae The static & dynamic ervor Consts are ‘Tversely ‘elated to each other, however they need not to be divect veciprocal values becoz dynamic error Const. ave function of error series: >> TRANSIENTs STATE ANALYSIS » Deena when subjected to an Lip & depends on order “KilS) aps O04 = of control sys- +E TIME RESPONSE OF A ZERO ORDER CONTROL SYSTEM %~ dependent: tf TIME RESPONSE OF A FIRST ORDER CONTROL SYSTEM &- cs) = ets) Transfer functlon, oye aH lp for the sys (8) = he, RO pt | cree 4 Response to Unit Impulse Input function $- ae Rai Cree] "cS)= atu Time responce expression, t ° TOT gr ar Response to Vrit step Input Function $- fis) = [reper] CS)= ae S(stt!) Time response, [c= (i-€ 7) vw) We dowt need anytime domaln analysis cl as a becoz thedip,olp chavacteristics are lineayly 6k} (8) bo Hdtals with the nature of response of a System, yy) iniMibn’ ee cts)* lower the time const. ,faster is thedime response ofa control 518 ae * |The step response will not have any overshoot far any combinationot $15 Parameters. ‘Bl Response to unit Ramptjp Function 2- aac a gS dip mnet sae, ows 7 Si (ste) , MeNeSHONEE, [Em=(d-T4 TET IUD Prot, ets wh-cit) 4 fewd= (1-Té) ow Stead: =i -Te T)= ystate error, |e. tim, (7 Té T)=T . velocity [The time vesponse shown infig, indicates that during steady state ee Poa Matches uitth dip velocity but lags behind thesjp by timeT & positional err exists inthe system. 4 IP FUNCTION, TIME RESPONSE exe. | REMARKS | unit Ramp 3, line t-t+Te"T —— [oitferentiate unit step ¢ ans -e% unitimpuse 1° fans 2@7 ‘tegrate é TIME RESPONSE OF SECOND ORDER CONTROL SYSTEM S- ee eee Transfer function, £19) = _Wr® Seren + vt op aNatoral Frequency of oscillation & 3 damping Ratio. gun 3 Damping factor * [Damped frequency, ww§= wy re © Effect of Damping on Nature of Response 3- RO) Ihavacteristic equation, s+ agung +Uy"'=o0 . Toots, — 54,53 = AQWnt VA GWnl? Hut =-guytUnTZT = - sunt oy #6 O4GAT 1 TEs Week eeellpelyimainary (One erate [Rel equal A Gre Reonequal & Be Natvreof Joscitationwith (onst-Freq, | amped osciiatlon (Critica pure exporentiag [Purely exponential sowe Resgonse | & amplitude : ‘ioiSh. Cassificatior} Undamped underdomped lcritically damp ed overdamped ow 1) oF valve cial valve. Figure(G] Pole tocations for pifrevent cases fer a Second Order System Si W, a n oer Gams) _ DP 4 en. LN Pi . 4Re (+e) feat pole for “ad (-1m@) J+ The Root locus ( locus of close loop poles in s Plane) of 2° order system obtained by varying the damping ratio (@)in & Semi civcvlay path with radius py and breakawwoy point 45 S=~dn onGve Teal axis: I> most of contol systems are designed for 04¢41 becoz the response canbe, aralysed with move no: of performance specifications: *lAs gincreases From oto 1,then the frequency of oscillation reduces: «|for all @vevalue of ¢, systems are stable: Pole owny from origin ave called incignificant pole and does not affect the stability lof the system. [Esa Time Response of a a™ ordey (GaMeRaapAR) Ci Sys- subjected toUnitstep ti Fs using: Asinwt4Beosut= Jaryg> cin [t+ 408(%)] up sre= 1 (stepitn), RI)=E = a Fane orp of sys. cis) RIS)_Wn 4. es “re Faget OP jens 2 ez sinfuyt+ tar (= ) wrt Sitedysta cts Ln = Se wpb + Or! t |ct= 1-2 sinfwytt 6) | fe (st + wy (1-8), Stn + d= gun qmpmgfactor|actoal |damplng coeff. y= 3 4! 2 5 Grqon + a + dampingrati = PSIG ay g lcigyeb- StQUN_ - BW, citlaidamping = Wn eget MS Greonhah |. time const of underdamped Response, erie v a gun a= 1-6 cant ~ Sp. cP snyt Y wd = Wn fee ant ; es [re casup + gsinyt] 1(>| Tine ao Specification of order control System oh DELAY TIME (ta) &- ” Lt isthe time required for the response toveach Sox fae EI » of the Final valvein first attempt. » [Rise TIME (+r) 3- Irt is the time vequived for the response to vise from odtoo7, (onterdomeed), : sx to 3% (critical damped ), oy to Gor, (over damped) of the fina) valve- Fy At the firstinstamt when time response Acceptable Sol; ugty +6 = 1 reaches joo of the desived value, cti)= 4 - ent, © t= TES] scot ree el y TS Te Sin(waty +9) d typ gant ‘ : am As, £7" finite, *. Sinfwyht6)=0 bei ze f 3 [Peak Time (te)s- F 5 “ltt isthe time vequited for the response to reach the first peak of the overshoot ee. lat t= 4p, clt)= first moxima, nanaabee “| det Bes at 131, 3,5.. for overshoos a n2 AG FOT undershoots MOTEGH n= 1 3 fist overshoots & n=2 4 first ndershoot, n=3 4 2° overshoot Sy [PERK OVERSHOOT (MP) ¢- The max Ove deviation (First peak) of the olP wit its desived steady state valve is Known Qs max: (or peak) overshoot “aT ratnp= Clte)= CC) y1007, = C(4A)-! xt00| Imp = CAFR xI00 a ; er vests nef 635 JretMp = @ Vig for 1 reak overshoot ees + For ordershoots usedto measure relative stability of a control Sys. lavge mp 1s usually undesitable- 7 . For SimestS.npmty occur at alater peakA if SYSTF hes an odd no-of zereg inthe vant half lof s-rlane, a Ove under sheot may even accu»» TEBAITIE RESPONSE OF ZMovdey CONTROLSYS: FoR UNIT STEPI/P Function - (ii oominannT Poves oF TRASFER FUNCTIONS 3 Settling time (s)$- [for aMonder underdamped ctre¥s) THis the time Tequived for vesponse so vise & each tothe tolerance band: for ay tolerance bard, tg= 4qu= 4 gun for sn tolerance bard, | t= 30= fC in Tntermsof no-of cytes, uy =angy 547 cydefsee on for sy. tolerance bond, [Rexfg= 214 cyceg] For ax tolerance band Wn For o desirable transient response of a adorder sys, [034g 40°F ‘The max overshoot and therisetime Conflict with eachother-Tnother words, both the Toa overshoot & the risetime can not be made smraler simottaneously rf one of them \smadesmaller, the other Necessarily becomes largey The equation derived above for obtaining the ay. tp, mp & 4s are yolid only for the std-second #4 ordey sys-defined by ct) - _wnt RIS) Se Ag unst wR If a order SYS: ‘involves azero oT 2 zev0s .the shape of the unit step response conve will be quite different from thestd- and order time Tesponse curve: Ite time peviad of the oscilation before reachingthe steady state [focinton” a) No-of oscillation before reaching steadystate, | N= Teettiing Togivation + Sustained (undamnped) oscillation «+ Giticaly damped oscilation * overdamped Oscillation = 4) (@0) feed) ame ema] aver a GREET) Foranalysis & design purpose » itisimportant to sort out the poles that have a dominent effect onthe transi ent Tesponse & hence called the domnent poles: Dorninent poles are used to control the dynamic performance of thesys~ Insignificant poles ave vsed for the purpose of enguring that controller TF canbe reealised by physical components: : ihe. Relative dominance of close loop poles 1s determined by the vatio of veal parts of the eLpoles aswell ag by the Telative magnitude of the ‘yesidve evaluated atthe ct poles. The magnitude of residues depends on both,the ant ceo 40 Close loop poles & 2erOS: cece sane ITF magnitude of real part oF apoleisatleast stole times insiitigne Ithat of a.dorninant pole ora pair of @mplexdominant pel tole unst cl Poles Rtheve ave nozevo ‘nearby then the Pole maybe —} times _ regarded as insigni Ficant pole as far asthe transient = “wesponige Is concerned -Ig Notes vue ed REMEMBER; # *ITime response of Higher order contvol syS+: using partial Pole representation: 2h Z Seto aw tw gt a ow ‘lmmpue Response, R. fk T Y Here s=1_: dominant pole- . The hovizontal distance froma, closed pole to the $w-axis deter eS oes Fime of transients dueto that pole - The smaler isthe distance, fhe longer !S 8 The closed op poles that have dominant effects on the transient response behaviour ave called dominant close loop poles: Approximation of Higher Order systems as Lower ordey systems: Reducing the order of a s¥S-TF [1 delete a cextain insignificant pole of ath) Convert the high order TF in time const: form before eliminating ine Poles K His) = RH oe to(tt A) After neglecting S=-10, = —H20__ (i818) T¢ the type of system ‘mcreass, the error decreases accuTacy MNCTEASES bot Ihe stability decreases. : the settling fraction techmiques @2¢r| insignificant ISevies RLC CRE Parallel Buc CHE | Translatory System CHAR EQ”: CHAR EA: Char Eq”: Char Eq”: Sst deco sto es* 50 sf s+ <0 ste ist f =o Une wre wne{E une JE eRe Cake & che ade some Practical Example of a° order system s- foraronal Sistem gs |4 i 3 STABILITY ANALYSIS of CHAP LINEAR CONTROL SYSTEM » 4 44] concert of STABILITY S- ° Mi ; citations setaypina system in consequence to small change inthe sys Jip in iitiolcondifions oy in S46: pavamneters ove damped oot wt time,” the system is called stable: 8180 Stability (Boonded Ze Bounded olp stability] A Gnear Time tnvariant (LTL) S16 1s sald to bestable if the following conditions ve satisfied: . : o (0)] when the system ig excited by bounded Jjp, the olp is also bounded ie: BIBO Stability! Alternatively, a sys. is stable if dts impulse vecponse approaches to zero as time approaches Lo infinity (co). ; {bi Tn the absense of dip, olp must tend to zero irrespective of the initial conditions. This stability concept 1s Know AS asymptotic stability oF zevo Lup stability: , Unstable systems- Tor ungtablesystem, oscillations are ‘increasing in mnagnitude- margival stable system ¢- [limitedty stabe ov critialystable System] IA UTI SyS-Issaid to be Tnarginal stable if for the bounded Jip, the otp maintains the const, omplitude and frequency 7+ ‘represents the verge of stability/instability. «|Absolute Stable Systemne- -t Syatem ig stable forall the value of sistem gain (10 from 0 £0.00: 3) The term abgolute stability is used in relation toquolitativearalysis of stability 3} The absolute stability can be determined from the \ocation of roots of characteristics equation in s-plane. Conditional stable System ¢~ - A system is said to be conditionally stable WF 0. S¥S pavarneter if it isstable only within a certain range of valve of the stid pavameter CK) : Relative Stability o- 3 Inpractical sys-,it ls not sutficiant tolnow that the sy$- is stable but a stable sys. mnvst meet the specifiations on relative stability which is a quantitative measure of how fast the +ransients die out in the sys+ >| wet Nine __{ Rerative stabity for various a ecations inthe Splene »eines bility may be measured by relative sailing times of each voot 0% pair df Bots « The setting time of a pair of complex conjugate pels 1s inversely proportional to the veal part (eve) of the v001S ‘This result 1S equally valid £ov veal voots: As a. vools (or apaivof roots) ynove further cong fromthe vimginary aig (as shownin fig;), the vel ative stability of sys: Improves: The max overshoot damping ratio & gatn-mardin, pase margin are measure +o relative stability: H | staaiurry ANALysis IN $ DomAiN $- Stability depends on locations of closed 100P poles poles are the roots of the characteristics equation of the Closed 100p poles = Roots of the chavacteristic equatton (ce): Tfall the roots of the chavacteristic equation have -pegative Te parts, the gystemn 1s stable: Tf any root ofthe CE has a positive veal part, there isa repeated woot on the ju-ayig ,the system is Unstable. IT the gjctem is stable except forthe presence of one ot move Non repeated roots on the 3u- aris the system is marginally stable: Both syst & sys~2 are absolutely stable- at zt gysv2 ig velatively more stable than sys. i Mt Condition for stability 8 * * ieee ae he nece scary (but not sufficient) condition, we f Hor stability of a linear ys: ig that alt he coefficient of its CE be real & have thesame sigh Further, none of the Coeff: shovit be zero: ‘ = The @ieness of the coefficient of the CE is necessary & cofficient condition for the stability of the sys-of 15t& 2°4 order sys the @veness of the CE ensures the Gveness of the reolvoots but does not engote the eveness of Teal Parts of complex roots for 3°4 & higheY ordet yg $o.it snot the sufficient condition for 39 ahigher order systems: 3) ROUTH HURWITZ (RTH) CRITERION » pia eel laa eee amd the closed 190P system: ar +f [Purpose 9 To Find: me Closed loop system stability: | oroF closed toop pole in rght Sie ov left side of the s-plane. | Range of K for conditiomally stable system: .| valve of k vequited for marginal stability andthere by determing the Frey, of oscillations Penne“NOTE: BK gt . i) ) a0 # 4. 2 3 atterratg: Replacing ¢ 2” SteoT:| Form the auxiliary equation Als)=o by using the coefficients fromthe vow just S00} Take derivative, 88) 20 3 gives the pair of poles hing on imaginary ayis. To Find a closed loop sys. stability by using RH criteria, we need CE» Routh Array 3- General ntherder characteristics equation’s, ayi"+as"raye™2.-----+ 0,86 apo Routh array: [a a Qe Mm [a a3 as gra |tardets — MAHORE ee . ay “ay sn for claged oop sygtem tobe stable, all the coefficient inthe first colurnn Must be @ve and there shovid not be any missing coefficient tnthe first colvrm: Té there ig any sign change nthe first colviwn of Rovth Array then sys: is unstable. Novof sign changes = no-of roots tying inthe Tight cide of the s-plane: Rovth array” gives no information ‘regarding thevalve of the voots and al not distinguish bjuo real & complex roots: Inagritude & sign of everyterm In each Tow & colvenn depend on first column element, therefore we always see first column term. ihe presence of zero(o) in tSt colin of Routh's avray lends to following conclusions: Equa] roots creo!) with opposite sign. Pairof conjugate roots on hmaginary axis: Difficulties & Limitations of Routh Array o- eee ee ee The Rovtharray is applicable to LTI system only- Tt determines poles in LHS ov RWS Of s-plane but not their exact location: Whenever any one coefficient iszevo in TStcolvmn then replace zevo by smallest Ye const [€] and continue the Rovth array ,finally substitute €>o & then check for the sign change: apply Rivth criterion on the modified equation intermsof 2. The no: of z-rools tuith ve veal parts will be same as the no-Of s-roots with Qve relat This method works tn most but not inal the cases: When all the elements in any one Tow of the Routh array ave zero, tt Conduct the following steps: : mus Is does preceding the rows of zeros. Replace the rowof zeros with the coefficients of AAI). 5 Continue with Rovth's tabulation in the vsual man ney with the newly formed row of coefficients replacing the vow of 2er0s——_] Step tore: - = Ht] Relative stabitity Analysis using Rovth Array o~ Trlerpret the change of sign. of the coefficients inthe Zst coumn of the Routh's tabulation inthe usual manner. The order of avrillary equation Ats)0)- The Ce of the sys: under study is then Ynodified by Shifting the ovigin of s-plane 20 g<-e;, te by the substitution s =s,-5, Tf pew ce ins, satisfies the Routh's critevion, i1 implies that all the roots of the origina] ce are syPlne —CHAP: g » Root Locus » wares NOTE: + Ruled’ Root Locus dlagrarn (Rib) i$ the graphical method of analysing & designing a.sys. RLo isaplot of Loci ofvootsof the CE wile sys.gain K is Varied from 0:t0.c, 7 api ¥ 1s varied om -@ to 0,the diagram iscaled as Inverse OF Complimentany LD. . li¢ mové than one parameter’ vavied,the corresponding diagram is called as ‘soot Contovy diagram: Root Locus (RL) > K+ 0L00 + Complete RL} K>-@ toto Complementary RLY K>-et00 © Root Contour >> multiple pararne}¢ variation] When H (sy6-gain) 1g varied From 00 c0,the plot is called pivect Root Locus, hile is varied from ~«o 40 o,the plot obtained is called Tnverse Rest Locus: ANGLE CONDITION 8 iit is used for checking whether certain points lie onvoot locus oF not and hence the validity of root locus for closed loop poles for a, print o lie on root locus, theangle evaluated at that point must be an ‘odd tmuttiple of +16" qe 0uade~ {mpoIniTODE CONDITION 8 Fi 1s uged for finding the system gain) Kok ary point on root Locus ,| |aus)Hts r+ can algobeused for finding the voots of Ce (or the close loop poles) corresponding! 40 a. given value of thegain: Angle condition & mnagnitude condition ave not independant and one implies the other: The valve of S that tulfill both of these conditions ave the roots ofthe ce oy the closeloop poles: CONSTRUCTION RULES OF ROOT LOCUS &- [construc h Root locus is symmetrical about veal oxig-L e -axis of the s-plane] /As the poles and zev0S of the O-L-TF ave eitheY veal or complex conjugate or Combiration of both, these ave always located gyrmmettically about therea| aris: Root locus starts (at Keo) from the open loop poles & terminates (at K=) either ‘on open loopzeros ov ‘ynfinity (co). . We only need to construct the upper aif of the root locvs and dvaw the mirror image of the Upper halt inthe lower half of s-plane- RULE2:| Let P= Nos of Open loop poles , & Z= No-of open loopzevos IF PZ IF ZP then no: of branches of RL= P + no-oF branches OF RL=Z se no-of bran ches terminating at zexos = 2|* Vo-of branches terminating ox poles = P [& no-of branches terminating at co = pag }* Novof branches termindlingat o = 2-P—_| Ruves: _ g Eg; fuss: $+ 2008 (P42) ways lies in \ying In plex conjugtte wo Apointon the veal ayis lies on the RL if the no-of open loop pale: Onthe Teal oxig to the right side of thls point 1s odd: ie. ALO these regions to which the no-of open loop poles plus zero ( P+2) ! Vight side ave odd- Here all the poles & zeros either real oF com} ave counted: tisye Hsay(ste), FFB Stst4)(S18) pezed Angle of asymptotes (6) ¢- : . lines The (p-2) branches of the RL terminates ak oo along certain straight know as asyiptotes whose angles are given by. Oz (2ati)160" |g o,,2,.-.--vpto (P-2-1): P-Z P-Z= Norof asymptotes Centroide— Ttisthe point of intersection of asymptoteson the veal axis: Tt is always tie at veal axis. Tt may or maynot be the part of Yoot locus: 7 ‘Teal part 0 centoid= Eateries — 2 creme sees PZ Break away points 5- ; ; They ave the points where multiple vools of CE occurs (points at which the RL gets break)- . Procedure to find out Greak away point: reoceavve OO Construct cE, tt GUSHIS)J=0 (0)] write Rinterms ofs dp (iti) wv) (c)} Find Gk (0) Find a The roots of geo give Break Away (8:A) points General Predictions abovt Break Away Points: Hine branches of RL eather approach or leave the B-A-puints at om angle of 18°74, Where n = no: of branches approaching or leaving the G:A- point - IThe complex congugate path for the branches of RL approaching ov leaving aba point Is a circle in certain (ases only: . Whenever theve are two adjacently placed poles ov zevos lie on teal axis (with a sectionof yea) avis bw them asa part Ru) then there exists a BA:Point blw them: - Whenever there is qzer0 on red! axis. ay left side of that zero If these jsno pole or zero on Teal axis with the entire section of yea) axis to the lettside of thakzero asa lpart of RL then these exist a GA: point to the left Side of that zero: