by Ton Wempe AN EXPERIMENTAL SEGMENT SPECTROGRAPH BASED ON SOME NOTES ON FREOUENCY ANALYSIS OF SPEECH SEGMENTS ='===============:=: SI]MMARY

Size: px
Start display at page:

Download "by Ton Wempe AN EXPERIMENTAL SEGMENT SPECTROGRAPH BASED ON SOME NOTES ON FREOUENCY ANALYSIS OF SPEECH SEGMENTS ='===============:=: SI]MMARY"

Transcription

1 r PROCEET}F{GS F R O M T H 8 N S T l U T E O F P T T O N E T C S C E N C E S O F T T E U N V E R S T Y O T A M S S g R D A M A U $ T E R D A M :

2 AN EXPERMENTAL SEGMENT SPECTROGRAPH BASED ON SOME NOTES ON FREOUENCY ANALYSS OF SPEECH SEGMENTS ='===============:=: by Tn Wempe S]MMARY Crnrnnly used dsplays f sgnat segment spect.ra lke '1ne spectrar mstly ffer pr readablty f varus speech sgnal parameeers. Sme deas fr btanng spectra wth greatly mprved readablty are descrbed. Experments usng hardr^rare were carred ut fr mplementatn f these methds. Fr estmatn f naccuracy n practce, artfcal fspeech sgnals' wth knwn parameters were used. Sme f ther spectra are shr^m as examples. Addtnalllr, spectra f sme natural speech segments are dsplayed whch resulted frm the methds as descrbed. The cnstructn f a segment spectrgraph based n ne f the methds was fnshed n 1976 and has been n regular use snce.. NTRODUCTON n general, the analvss f segments f speech sund s nt fundamentallv dfferent frm a descrptn us ng a set f unversal

3 - 45'- parameters. Hwever, parameters sutable fr the nvestgatn f certan segments f speech (whch s nn-statnary) maybe unsutable fr ther segments, dependng n the lcatn and duratn f the selected segment. Besdes, the chce f parameters depends n the am f the analyss. A partcular set f parameters may result n an analyss whch culd be cmpletely satsfyng fr a mathematcan but nt necessarly fr a physcst. T meet the acustc phnetcan's requrements fr an analyss we have t answer the questn: what d we want L measure? Or n ther wrds: when selectng a speech segment, what are \^/e lkng fr? When speakng f frequency analyss we naturally cncentrate n the mre statnary speech sectns n rder t gve the term tfrequencyr sme physcal relevance. S we are partcularly cncerned wth the vced segments. The determnatn f frmant frequences n the descrptn f ths class f speech sund stll plays an mprtant part. Defnng frmants as the maxma n the system functn f the vcal tract cnfrnts us at nce wth the prblem f extractng frmants frm the speech sgnal. n the speech prductn mdel the exctatn surce (g1tta1 pulse) drves the lnear flter system cnsstng f the vcal tract flter and the radatn flter. The frmants as defned abve, are system parameters, whereas, frm the utput, nly sgnal parameters can be extracted. Ths wuld nt be a prblem f the exctatn surce culd be cnsdered t delver a sutable test sgnal r, at east,3 sgnal wth knvrn parameters. The fact that ths s nt the case s the man reasn fr the dffcultes n frmant extractng. Frm a sgnal analyss pnt f vew the prblem desnft exst: f a frmant s undetectable wng t the prpertes f the ex-

4 -46- ctatn surce, the sgnal present cntans nsuffcent nfrmatn fr ext.ractng that frmant. Therefre we cntend urselves wth frequency dstrbutn dagrams whch are vald fr the selected speech segments. Lke many thers, we wll call the peaks f these dagrams rfrmantst fr the sake f cnvenence... Althugh theretcally frequency analyss s applcable t any length f the sgnal segment, t s mstly shrt segments we are dealng wth n rder t be able t make a mre detaled analyss by means f many cntguus shrt tme frequency spectra alng the tme axs rather than a lng tme average spectrum f a large sgnal prtn. A rapdly varyng frmant culd change wth a velcty f 30 Hz/msec (accrdng t an example frm Markel, 1972), whch means that the frequency shft between tw sequental perds f the fundamental frequency f a male vce culd be abut 300 Hzl n ur effrl n makng the analyss mre detaled, we wuld lke l shrten the segments t be analysed. Hwever, f we cntnue t shrten the sgnal tme ntervals, the term tfrequencyt becmes gradually mre meanngless. Accrdng t the uncertanty prncple the naccurac.y f the determnatn f a frequency cmpnent s abut 1 /t tl, where T" represents " the length f the segment analysed. n cases f rather steady parts f vwel sunds, t seems therefre advsable t take segments f lnger duratn n rder t be able t measure frequency peaks wth greater accuracy. Because f the perdcty f the sgnal wthn the steady segment, we rnght just as well stre ne F'-perd n a sgnal memry, whch s afterwards scanned repetvely s that a cntnuus perdc sgnal arses f any desred duratn. The frequency dman representatns f perdc sgnals hwever, are tlne spectrat wth cnstant lne dstances equal t FO Hz.

5 -47- A spectrum f ths knd s t be cnsdered as a sampled versn f a cntnuus spectrum f ne F'-perd. t wll be clear that as far as the accuracy f frmant extractng ges, there s n pnt. n takng segments f lnger duratn than ne F'-perd. The frequency cmpnents n the frequency dman representatn f sgnals are f nf nte duratn rnrhereas the sgnal segments are defntelv nt s. f we thnk f a sgnal segment as beng the result f a multplcatn f a cntnuus tme functn wth a rectangular 'wndwt functn f unt ampltude and f the same duratn as the sgnal segment, the spectrum f the segment s fund by cnvlvng the ndvdual spectra f the multpled functns. The spectrum f a rectangular twndw' functn s a (snnft'/nft') functn and has zer cmpnents at multples f /TO. The result after the cnvlutn s a spectrum wth addtnal maxma and mnma ('1bes'). Because the readablty f frmants can be pr wng t the ccurrence f these t lbes t, a number f dfferent twndws t have been appled (Gauss, Hannng, Hanrnng) n rder t elmnate these 'lbes' r reduce ther ampltudes. n addtn varus (cmputerzed) smthng technques have been develped t smthe ths'f^-rpple'r t cnvert the fspectral u " FFT-prgram nt a cntnuus graph. lne'utput f the Frm a physcal pnt f vew we shuld prefer erhaps the use f tme-lmted snusdal cmpnents, each wth a length equal t the speech segment, rather than cntnuus nes. Apart frm the ac,curacv f frequency determnatn, the dfferent lengths f the segments then have n further cnsequences as regards the shape f ther spectra. Anyway, ur am was t try t develp relatvely whch shuld: smple hardware l. apprxmate the ampltude spectrum f ne F.-perd wth p-

6 tmum readablty f the frmants and wth knwn naccuracy. nfrmatn abut the frmant ampltudes and bandwdths are thus ncluded; handle segments f varus lengths (.e. dfferent F^perds) be easy t perate. at ptmum reslutn; 2. FRST APPROXMATON The nsertn f tme (T") nt each perd (TO) f the fundamental frequency f a perdc vwel sund was the frst dea n apprxmatng the cntnuus spectrum. The effect f ths s that the dstances between the spectral lnes s decreased because the mdfed fundamental (tot) equals l/(t^+t^). The mre T- s ncreased, the better the cntnuus u a a spectrum apprxmatn becmes. T get sme nsght n the shape f the cntnuus spectrum f ne F^-perd f a vwel sund, we defne a smplfed speech segment U ' as a damped snusd, whch s started at t=0 and truncated at t=t^: U -Af (l) g(t)=e snllt 0 < t < T Usng Euler s relatn g(t) changes t: c(r)= +[. (cr-jr) t -(cr+jurr) rl J 0 < r < T ts Furer transfrm s: +v" (-cr,+j1-j) t G(trl) = dt,1. ; (--jurr-j) t - l e J.f

7 - l, - Evaluatng leads, (2) c(ur)= 7Jl L t: -[+5(ur1)J r -[+j (+r)] T 1-e l-e +j (-urr) 61+ j (+1) Ths can be wrtten as the subtractn f tw functns; (3) c(trl) =ce (ur) - c*(), where: G" () =,T [ r - r - l L*j (r-rr ) + j (ur+1) J c*() = g - (+ rrl) T - 2j jrr l e l_+j (ur-r ) _jrr _ e l 6x+j (ur+1),l 0r, n trgnmetrc frm: (4) c t'r,r\= - \'g \uj,/ ()1 (5) G (trl) = e m' - t'cr+ r,r\ T (cl+j) sn trlrt+trtr cs trtrt (cr+ jtl) 2+ur! The latter functn becmes zer f Tn tends t nfnty, s that G^ (t) represents e ^ the Furer transfrm f an untruncated damped s nusd. Frm the superpstn prperty f the Furer transfrm t fllws that: g(t)=ge(t) - Cm(t) whch causes S -m (t) t be: 8*(t) =e (See fg. 1) -nf s nl t T < t <

8 -50- ge(r) g(t)=e"(t)-sr(t) Fg.1

9 - 5l The functn f g"(t) wth gn(t) can be thught f as a tme-shfted versn -ClT *-" an ntal ampltude f e and a phase-shfted rgn. The ampltude spectrum f G"(trl) s determned by: (6) lc^ E (ur) l= (r)t V('*r?- u2)2+4u2u2 Ths functn expses maxlma at (7) (l)=u) = + m u?-a2 (B) The magntudes f these maxma are 1/2a. T derve the 3 db bandwdth f the peaks we slve the Jl _ l *l-- 2a\12 whch delvers: n ul?-cl2+2urr '"1 As lng as 2q,2<<L - Z*r, we may wrte: n 6f +6x2+2rrl = + (r +) = +(trrr-)

10 whch means that ujh-uj= 2A (9) r: B= 9 Hz The naccuracy f the apprxmatn n cases f practcal values f trtr and, s rather sma1l: fr example n the unfavurable case f fl=300 Hz and c=00 n, the devatn amunts t less than 32. The shape f lc.() s utlned n fg. 2. The shape f the functn C*() depends n the phase anqle LrT^. - U f rto = nn(where n s an nteger): ( t0) lc,n, {) l=. -CT "; c" t,) and n ths case lcr(ul) s srnply an attenuated versn f lc.tl l. tf 0rTO = n(n+!): (u) lc,r(ur)l = qf lc"', () lc*r() nly equals lcr, (ur).r 2 - l.2, vz at the maxma f lc*, (ur). Fr values f where u)2>>a.2 the l c*, () ncreases lnearly wth 6 db/ ctave. specrrum c'., (r) cmpared t mlncreasng uj, accrdng t Maxma ccur at (1 2) max r: u2 max (u?+za2)2 - d 2 Fr practca 1 values f tr and 0, we may neglect the term 4clq n

11 -53- respecr f (u?+2a2)2 s that '.,2 w max tl?+cl2 and (13) max!+2 The heght f the max j-ma rs:,^.., -, f Gm2 (tr') l*"*= \ / ^ V zutr t!r'r*a'-r, J Agan we may wrrte: c*, {,^,) lfu" 2@?+2a"2)-zu 4a2 ( 14) Hence: c*, {r) l,n"* = - 2a Whch s equal t the maxma f lc"(ur) l. The 3 db bandwdth can be fund by slvng the equat ln: a"'*tj' (cl2*?-t.r2 )2+4a2u2 4trr 6l+4a2 -ujl whch delvers: 'lr= zp'-'?-'* f +3p 4-4u-rf p 2 ^1= 2P'-'?-a2- t lf+:p4-4'?1 2 ( 5) where ul+4a2

12 -54- Let be fr= *+y and ull= x-y, then: -, 'r 0"-0,=\/**y -\/"-y N V ' V (u1-trr) z= zx-\ r *2-y2 s that (0trl.-ul,.*h *1,)2= 'Y 4l 2-2u?-26y2-2\ -w - v t -t Substtutng f (15) and sme mnr manpulatns yeld: (t \r-01 )'= +r,\f',*?- zu?-zaz Fr practcal values f tl.lr and 0, apples: 4a2ul( -\f*4$ << (ur?*a') ' Ths al1ws the smlfcatn: (rn-t )'= +r r! r', *F - 4w2r- 4a2 Agan we wrte: (t1-r)'= $l*fu', ** 4a4-4wl-4a2 and neglecr -4,4: (rrln-rrlr)2= +fu*zu2)-qu?-+a2= 4u2 (16) Hence, 0h-01= 2a r B = 9Hz S, n practce the bandwdth f the peak s case f lc^crll. (See fg. 3 fr urlnes f ' g ' ' Other values f the phase angle trlrt' result smewhere between these extreme curves. the same as lc,n, () and n ampltude n the c*, () spectra lr.

13 -55 ta-- Fle.2 rt6.3 -.//4: _. -tr- -'-'-' l6mr(ar)l / a,",t ]'-'.- \ lc(ar)l = t le c(c.r) -G F6. q tt\. t' t'-l' lc(r^r)l- t lg3() -6r0. F. s \ ----

14 T fnd the arnpltude spectrum f g(t), each frequency cmpnent f Gr(trl) has t be subtracted frm the crrespndng cmpnent f C"() whle accuntng fr the relatve phase shfts. Accrdng t the tme shft prperty f the Furer transfrm a tme shft f TO secnds crrespnds t a phase shft f -TO radans (hence the fact. "-jt n equatn 5). Thus when subtractng Gr(ul) frm G"(), rnaxmum devatns frm lc"(ur)l ccur when crrespndng cmpnents have equal r ppste phase. Ths accunts fr the tf.-rpplet n the spectrum f g(t). (See fgs 4 and 5). The rperdr f ths rpple s 2r/tO rad/sec. The lwer the factr e-t, the lwer the rrpple ampltudet. The methd used here may appear rather unsphstcated, but t has prbably the advantage that t ffers a detaled nsght vsually. T test ths tme nsertn n practce and experence the effect f the'f^-rpplet n the readablty f frmants, a smple artu -- fcal vwel generatr was made whch prduced tw damped snusdal vltages as frmants and was ftted wth adjustable TO and T-. The resnance frequences as well as the bandwdths were made a presettable. The utput \^/as cnnected wth a swept sgnal narrw band spectrum analyzer (General Rad wave analyzer Lype 1900A). Bref descrptn f the artfcal vwel generatr (see fg. 6). A pulse generatr s made by means f tw cupled mnstable mu1- tvbratrs OS and OS2, each ne trggered at the end f ts ppnent's cycle. Cncdng wth the begnnng f the T'-cycle a very shrt pulse s generated by OS3 (T6= l25psec). Ths pulse exctes tw resnance crcuts smultaneusly, GC and GC2, whch start prducng ther damped snusds. The resnance crcuts each cnssts f a gyratr, ts tw gates cnnected wth capactrs. These crcuts are equvalent t an LCR-crcut. A gyratr used as a resnance crcut has the advantages f beng stable, beng easy

15 F\ (5 u-

16 t cnstruct and has ts resnance energy cntaned n nly tw capactrs. (See chapter 4 fr a descrptn f the gyratr prncple). At the end f a T.-cycle, a T.-cycle s started durng whch the capactrs f the gyratr devces are shrt-crcuted, causng the utputs f these devces t becme zer. The result s a wave-frm srner,hat lke f g. 7. The effect f the exctatn pulse n the purty f the damped snusds s neglgble: the ampltude spectrum f ths pulse beng prprtnal wth sn lt6/t6 has ts frst zer at ul= 2nlr6 r at f= l/t6= 106/125 Hz = 8000 Hz: s that n the vcnty f the frmant frequences the rsurce spectrumr s almst f1at. The crrespndng spectra f sme f the varus sgnals wth dfferent parameters we measured n ths way are shr^m n the fgs 8 thrughl4. The utputs were prduced by a chart recrder (General Rad type l52lbql) whch s mechancally cupled wth the narrw band sdectrum anat.vzer alreadv mentned. The tme-nsertn prcess causes the average vltage f a sgnal t decrease wth a factr equal t T6/T0. n rder t acheve cnvenent dsplays, the utput vltage f the spectrum analyzer was ncreased accrdngly n cases f sgnals wth tme-nsertn. Naturally, the sgnal t nse rat f the measurng system decreases wth the same amunt; n practce the S/N rat f selectve measurements hwever, s very hgh (here B0 db) cmpared t the S/N rat f taped sgnals. f ne bears n mnd that durng the tme T the nse level at the nput f the sdectrum analyzer d can be kept very lw, eventual- cmpnents due t nse n the spectrum can be cnsdered as beng entrely caused by the rgnal sgnal tself. Frm the fgs 8 thrugh 14 the cnclusn can be drawn that the tme-nsertn prcess s a valuable methd fr accurate measurng f frmants frm sgnal segments wth relatvely 1w fnal ampltudes (.e. F^-perds f male vwels). Fr ths last class f sgnals the U ' frms 7 and 9 are qute applcable.

17 U z u z!l g td E u. N rn ll e\ N F\ tl N ts N tl \./. g (d d.rl r+t s \ a{,t Af tl +J q) h r E e l!! n tj l-< 4 l f;h; r Nr HsNdSSu t! HHh '; u t tr HeR '-g g: H;l - ( -\ - Ī \5 (.n

18 -60- u v z u2 ld f r? L tr E N N l +J tr OJ aa OJ.Fl +J d..p 'r{ F t\, (1 ll b0 (H lr q-.1 Fl (d - rl 5.:s (t s, qj aa rj q-l tf.lj a 0) (0 r F O T J O F t! J O z 80 Nr 3SNOdSSU uj lrj l c z H =" (! F \ O tr Zl - l < ] J. t r f t ), L,. l u.t u J (! - 6 = F a - =utr rll :.J O : l uj :-4 t :1.tr t z-l t l l ;l l T ; : : l l :.'t', -:-- t,.-j-..,. -.-.'_')^

19 -61 C) Y = z uj f l! cr l! N ts tn ll N N -r tl N F c\ tl H E r--{ b0 t+{! +J q) (t) b0 80 Nr SSNOdSSU r t r F t u = J O z tr ('! r tr ul G u.l ; - u, G F O E.J ( r O z = = UJ u l ( L F 0) rt z E =tr l U ft J a z

20 Y = z lrj f uj (r l! N \ \ tl - d. rj t'r 0) 'r"{ c).rl p {J F q-.1 tr q-1 r--f (d - b0.rl 0) {J r+-{ tr lj a c) b0 'rl go N 3SNOdSSU E F O ttl! ;! J O z tu ul k ( E F c E u r - (r() = -r' =)l:, - t 1,-'. z - :, l\ : ' tr lul l l6 l uj ll t ' ul t! u)- = F c E * =! (\ tr (')!! f LlJ :{ a =

21 -63- C) Y = z UJ f, ul (r l.l. t.l O rn tl ftl N F \ tl fr{ N c{ l (d F, b0 (') l+l! lj c) U) b0 go Nl :lsnodssu E F O UJ t! J O 4 z E z ul t llj ;l (L = ( E F O E U J z = = uj l.! O. F a f 1 z F tr- = n N n! u l a (t T-t z 1, : 1 {l r ' : t 1_

22 -64- = z u.l l g u.l E. L N ts -$ tl d. +J tr q) 'rl c.) +J d. ]J.Fl F b0 'rl LH tl q{ r-.l (0 d. 0.F{ CJ - +J q-l tr 4J a qj a0 l r t r F O t! ;! J ( f z l,, l l l, l l ' : r l a 5--l*-**. uj l r r l l : l a l ; " l l r, l l, lur '---1 l,, r : l'' : '-', t.t-;,:;.-:-..t O N SSNOdSSH UJ ut F E...1 uj llj a O r._l z ; cc. =! N r"t U J (7) T-t = J - l, 1,...r--*.-..-.^ b0

23 , ANALYZER *-$ BANDWTDTH tr3!10 ---^*---;"* - f *1."'" -*; RECORDER charr speed N,/MN -- c0 = uj a z (L a UJ cc L**.*. *._...: RECORDER WRTNG SPEED --",!1 fl 10.. L t=tzv,- -.; ln/sec -!-.' '..* -.'' - -,M*'4q!'lzlt4.@'.<.!r -,;*, r * B, D D '...: :ns +tr:1 +t! :1 +! RECORDER CHART SPEED N/MN t- m = uj Q) Z l. (.t) : r! : cc. RECORDER ]WRTNG SPEED : *;*_;!1 [t _-.*,tr3!20 - rn/sec. - - Fg. Spectrum f s gnal 5F, F =24OHz, F,=4O0Hz, F,=7 \OHz Fg. la Spectrum f sgnal 5F wth rme-nserrn, FL=l4.4n,

24 ANALYZER --*;-_-t BANDWDTH -^ L_l ru -50 t---*,, : t ;;;;;;;; CHART SPEED N/MN Q).-.-. "...", ,.".-...-, U.,. : J ,^.-...._. -_.*-...- RECORDER 'Wntf tng SPEED n1!t tr3 n20 N/SEC -..;...-'..-.".- -,.,-. m r] = u 6) z {-] (r t r T Tr- T r T.ANALYZER F'*'**; BANDWDTH l 3 n10, t 1, _".*-,. -. RECORDER CHAHT SPEED N/MN = u a z - u) rl tr...***r***---, -..*...*,* RECORDER WRTNG SPEED - L D10 : l]3 a20. N. SEC ) Fg. 2 Spectrumf s gnal 2l'1, F =160H2, F,=540H2, Z =86OHz Fg.12a Spectrum f sgnal 2M wth tme-nserrn, =l4.4hz f,f

25 u3 D10 '. l l,.-,.---. ",..-- -,-l..: RECORDER "**r speed, _: NMN _;--_" r. r : f" " -_ WRTNG SPEED *,---nt lll f 3 _2 rn'sec -,* -""^_*1" m = ul a z - a J (r E, U t- L! l f f t l T 1 T T - f* :.., **-.,.:;**:-**..-,. -.. : RECORDER CHART SPEED *-^:"** N/MN r'e ccfdt***"- *...-*-,",,..--*::*.f*"-. ^* t r t "--.--: :-.-.-;:- RECORDER l c0 z : lfl U). z.: llj j 't WRTNG SPEED tr1!10 Fg. 3 Spectrumf Fg. 3a Spectrum f sgnal 7F, F F,=580H2 =Z4OHz,, Fr=93OHz s gnal 7F w th tme-ns er tn, F t-=1 4.4Hz

26 Y = () z JJ f uj TL N 6l tlc{ Fll N co C.l lt N \t C-'t tl rt, --l cd a'j.f{ ct) (H tr +J q) b0.fl 8C Nt 3SNOdSfU

27 N $ \f tl r t +J tl c) u) d 'rl c.) F -. +J 'r F b0 tt-.{ q-.1 F'l (d d..rl a.) -. ]J q.{ tr ]-J 0) L l $ b0.rj Y = C) z uj : U E. tl 80 Nt SSNOdSSH

28 The F'-rpple f spectra frm sgnal segments wth cnsderable fnal ampltude (fgs loa and l4a) culd ccur t such an extent that the readablty f the frmants rnruld be rather Dr, especally when mre than tw sgnfcant frmants are present. Furthermre, the lmtatns f the methd fr frmant extractng frm lne spectra usng fhe prnc.ple f a weghted average f maxma and ther neghburng cmpnents (Ptter and Stenberg, 1950) are clearly dsplayed partcularly n the fgs l3 and l3a. The hgh-frequency behavur f spectra f the type as shvrn n the fgs la and l2a s caused by the fnal ampltude step present n the crrespndng sgnals. n practce, when gatng a sgnal segment, ampltude steps can easly be avded b1r takng care that the segment starts and ends al zer-crssngs, s that the spectra f fgs lla and l2a are n fact academc nes (and wth them the functn Grr(ur)). 3. SECOND APPROXMATON The cnnnnly used methd fr elmnatng r attenuatng the rnterval rpple' n spectra f tme-lmted sgnals s multplcatn f the nterval wth sme rrndw functnr prr t the transfrmatn nt the frequency dman. Hwever, f ths s appled t the shrt ntervals f the F.-perds, large sgnal prtns f these are greatly attenuated, especallv n the ntal parts, where the sgnals have ther majr ampltudes. Frm chapter 2 we knr,,r that the rrpple ampltudet s prprtnal t the ntal ampltude f the mdfcarn functn g*(t). The bvus way t attenuate the -0Tn " ntal arnpltude (whch equals e rf'-rpple' s decreasng ths tmes the ntal ampltude f g(t)) by multplyng the sgnal wth an expnental func tn: (17) s, (r) = "-"' -n [O<t]

29 - 71 Nw, culd be thught f as the sum f the natural and artfcal dampng: g(t) = e - (crrr+c") TO s].n (r), t t The ampltude f the mdfcatn functn alters t: -(cl ' +a )T n a' Or^,. e lg^(ur) f " culd be gven an ndvdual value fr each segment t be measured, t wuld be pssble t keep the factr -(a ' n+a a )t ' e cnstant fr all dfferent segments. f -(cl - n +a )T a - e = k, then the spectrum ccurs between (l+k) lc.(rrr) and (l-k) lc.() l. e tg"rthmc ampltude scale causes a cnstant rpeak-t-peak rpple value': (r8) ur = 20 g j* uu Chsng k = /20 yelds the rpple t be wthn db. -(d +c[)r t a' " (19) Therefre " =* f fr mst vwels ct' s valued at 250 and fr TO s wrtten l/f', then frrn (19) the rule-f-thumb can be btaned: (20) 6x =3F^-250 U Of curse the accuracy f the dspl-ayed frmant decreases wth ncreasng cl. anc frmula (20) therefre ffers a cmprmse between rpple smthng and acc.uracy f the frequency measurement. Fr a twrse caset example, assume! F^ = 240 ltz, then: Ct = 0 +d = 3F^ = 720 n a u

30 Substtutn f ths nt (7) tgether wth the fl l = 300 Hz as used befre gves: ur---=ff=f42 max 'wrse' value r: f = Hz. max The errr s abul 7.52, whch s lw cmpared t the famlar frmant extrac.tn errrs f ths type f sgnal. Besdes, because the devatn frm f, s knwn, a crrectn s pssble f des red. Of curse the 3 db bandwdth s cnsderably ncreased. n ths example t amunts t abut 230 Hz (B = 9). Tr' ' Evdently, t check ths methd n practce, the same test set-up as descrbed abve s applcable, f the dampng f the snusds s adjusted accrdng t: J cr = 3FO = -.F Sme spectra btaned frm measurements as carred ut n ths manner are dsplayed n the fgs 15, 16 and 17, whereas the spectra f the crrespndng sgnals wth 'nrmalt values fr s. (.e. wthut addtnal dampng) are depcted n the fgs 1/s, l3a and l4a. As a result f applyng ths fexpnental wndw', the readablty f the ffrmantst s mprved cnsderably. Usng the cnvlutn therem t can be stated that multplcatn f the tme functns results n the cnvlutn f ther ndvdual spectra. Thugh ths s nt exactly Lrue fr ampltude spectra, the errr culd be sgnfcant nly n the vcnty f zer frequency (Randa11, 1977), The prcess f a measurng flter scannng the entre frequency range can be cnsdered as the practcal analgy f the prcess f cnvlvng the nput spectrum wth the measurng flter characterstc. Then f the flter pass band curve culd be made equal t the spectrum f the expnental wndw, ths methd wuld gve results very smlar t the expnental wndw methd.

31 '"*- BANDWDTH n3 [10 Lr 50 t- RECORDER CHART SPEED N/MN.."_".,* '*_-t = ul U> v 0- ) ul tr RECORDER WRTNG SPEED _-,r, n3 ttt /ctr^ --t l 10!20.A\AL T 4trN *.*.**; BANDWDTH tr3 n10 Tl 50 : : : "-.t "..,.-" RECORDEH CHART SPEED lnrmln.----.'- ; ,.-. c0 = ul a 7- (-') uj (r RECORDER WRTNG SPEED tl r -l 10!3 Llz N,'SEC : : Fg. 5 Fg. 6 Spectrum f the sgnal frm fg. 12 (sgnal 2M) r^rth tmemultplcatn, cr"= 200 nsertn and exnental wndw Spectrum f the sgnal frrn fg. l3 (sgnal 7F) wth tmernultplcatn, cx,"= nsertn and expnental wndw )50

32 ! ( ) 0 J ^ Y q O (( - d z. l J \ v.rt z B a JJ - d d F l llj t d r.r- cd.'r h n ^ v ( / ) { J r l t \ t F. B.d "(, q { t 'r d E ^ H r'-r q.r (d "l r--r P A / tr) t a Q ' X AJ CJ d. d. (f.t $ ^ d. F -. d q lr.,j +J tf c ) 0 J ( n ^ 4 (.1) ' Fl : b0.rl rr, t-

33 t " The ampltude spectrum f the functn e to < tl s equal t: A smple secnd rder resnance crcut wth a bandr,qdth f matches ths functn apart frm the asynnnetry when pltted C -nz 1T na lnear frequency scale. Althugh ths methd seems qute realsable, we thught t easer at the tme t apply the expnental wndw methd snce n that case we culd use the narrw band spectrum analyzer agan whereas we therwse had t cnstruct a spectrum analyzer wth a cntnuus adjustable bandwdth. The tme nsertn prcess and the expnental wndw methd were als appled t natural vwel sgnals. Fr ths purpse the fllwng measurng set-up was made (see fg.l8 ): Frm the vwel sgnal t be measured ne perd f the fundamental frequency was slated wth the ad f the tprecsn Gatet 1976) and stred n a dgtal sgnal memry (1000 x B bt). The memry length was made 20 msec whch s suffcent t stre any F^-perd f vwel sgnals. Thus after the sgnal perd zeres U ' were stred. (Wempe, The cntents f Ehe memry were repettvely scanned at the same rate as durng the recrdng, cntrlled by an external tmng crcu t. The bnary equvalents f the sequental samples were cnverted nt ther rgnal values by a multplyng dgtal-t-analg cnvertr. The DACrs reference nput was cntrlled by a devce whch -cr t prduced the functn e ts,a was made adjustable. The tmng -cl^ t 4 crcut caused the functn e t start frm ts rgn (t = 0) at the begnnng f each memry scannng cycle. Besdes, ths crcut als actvated an analg swtch n such a way that the analg utput f the DAC was cnnected t the spectrum analyzer nly durng ne ut f three scannng cycles, whch caused an addtnal tme nsertn.

34 Ea Ft{ F r.j r lr> r/l< t- E e < ; 5S s 5 ar.u L U Lttr '.5L z - E g F \ J '9r OU J F z 3 <vl lj U r./'t! j U A Z < lrj!!(j - r ) ;. 9 E A E l! e LJ JJ g U F Fe. 8

35 " r t ' t ' t +F- -r +rr+ _. FREQUEN 0 l : u J FREQUEN Fg. 9 Fg. 9a Spectrum f male y] wth tme-nsertn, F=95H2, Spectrum f the same vwel wth tme-nsertn and wndr^r multplcatn, ^= 50 Ft =14.4H2 expnental

36 Y = UJ f n [l E l! N l N O c.l tl -, +J tl q) a4,.fl q) 'rl.u -. P qj r-4 (d (+-l tl.lj c\l.fl Fr : j t '".^'** ' T E. F O t! N ; L J O 4 z ; z < < c [r f] : _ -_j -_j_- BO Nt 3SNOdSfH : ^ ; l ',. : u l : : U t,! t t r J : n - - N, ' tf tlj.^.) v / t - a - s t^_-.--.,..*--:.* -",..- uj! u ul - a. ; a cc z : _ - = r : u l = 'ul F _ r!.(r Btr

37 d..f{.u tr q) a d..r{ U J J d a c{.r. + J b 0 ( 0 'F O tf-l. rl r--r l r r J q-.1 tl. cd..1 E F QJ r+ +J 'rl.lj q-{ tr d. t r X lj c) 0J. c\ 0 tr{ f r L a N ; t l z < _ : < (D L_.1 Lt ' l : t l ec Nr SsNdsSH u l, H :!! :!*"*"...-. ' -. v) r > : ; t l t z ; u. - r (r O.! n : t l l l ^ - N cc 6,, : -* -.* tu l- J Ur.; ' 1. r t z u) : - 7 r ( J - ( 9 - cc =L :

38 -80- Y 4 z LU f lr.l E L N ll N -\ c! tl a..1 t{ qj d. 'rl c).rl.lj -.lj.fl GJ r-"r CS qj q-.1 q-l t lr ]J a.) 6l b0.fl Er r{r SsNdstu -**1 ( F O llj N = -l > 5 ' J O z, l,-' :!: tll e ul ; -, 4 6 *!. 4. l J. h ; F. '() ( l(r z.l"*'*--1*' : l, t. l 1 ; u J l t t : l ^ - N : c c O ul!! l!.-- ';E O u), t Z \t ; 2. O = - c D - '; =utr l

39 - 8l -f jj_ Y = z uj f E u- -,.r{ +J tr qj U) d. q) O! s J (.d c! B < J N t r.rt. -{.lj l O, q-.{ 'rl.lj r A J B r--r O CJ q-r CJ.ll ttl 'rt F Fr 'Fl 'lj tr d.! X lj qj aj.d c,)(s (0 c{ : r" 1 l EO N SSNOdSU tll l* ur '- 1 r n Z ; ;(r F \ l O E z r.! l r n r u ) u.t O - - ' r ( ' ' - t u! ) E z ' v F cr =tr, J j, ' : N : u r.r.l "-.,j a, l. l \! z l t c ' - l f r l..-:,,"..^.:.":.

40 (The resultng fundamental perd T' was slghtly mre than three scannng cycles wng t a tme delay between cnsecutve scannng cycles. The fnal Ff therefre was equal t 14.7 Hz). The effect f the adjustable factr Oa n the tme functn f the speech nterval was made vsble n an scllscpe. Cancellng f the multplcatn functn was pssble wth the swtch S. Frm the sgnals v/e measured n ths manner a few are dsplayed n the fgs 19 thrugh 2l. The fgs 19,20 and 2l were made wthut the use f the expnental wndw, whereas the effect f the expnental wndw s shr,an n the f gs l9a,2oa and2la respectvely. 4. THRD APPROXMATON The cnsderable amunt f tme necessary fr plttng a cmplete spectrum and the ptmum adjustment f cx," smetmes beng dff cult were the majr dsadvantages f the prevus methd. Althugh bth prblems can be practcally elmnated by respectvely adaptng hgh speed analyss and determntb " autmatcally frm the length f the segment, at ths stage an exper.ment f a dfferent methd fr frequency analyss r,cas started. The fllwng rnay explan the prncple f ths methd. Cnsder the practcal case f a measurng flter respndng t a tme-lmted sgnal prtn. Frm a tme dman pnt f vew the j-nfluence f the sgnal length n the respnse f the flter nly starts after the mment the sgnal segment ceases. n fact the sudden change at the end f the segment cntrbutes t the flter respnse nly frm that mment n. The'frmant maskngt sde lbes n the cntnuus spectrum (r n ts practcal apprxmatn) are therefre slely caused by ths sudden change at the segment end. Nw the dea was t mdfy the tme nsertng methd frm chapter 2 'enablng' by the measurng flter nly durng the ccurrence f the sgnal segment wthn each perd f the flter nput sgnal. 'Dsablng'f the flter was ntended t be acheved by wthdrawng the energy frm t s that

41 -83- the flter respnse shuld start frm zer at the begnn.ng f each sgnal perd. Usng narrw band flterng mples that n ths way the flter utput vltage s kept cnsderably belw ts steady state value. The spectral nfrmatn hwever, can be extracted frm the varus peak vltages f the flter utput at dfferent center frequences. Real zatn. Fr the practcal mplementatn f ths dea the cnstructn f a specal purpse spectrum aaalyzer was necessary wng t the unusual requrements. The measurng prcedure culd be rughly descrbed as fllws: Frm a sgnal, an nterval wth 20 msec length s recrded nt the sgnal memry n such a way that the fundamental perd r ther segment t be analysed frms part f the memry cntents. Durng repettve replay f the memry cntents, the wanted segment can be selected by means f cntrls whch cause the passng f the sgnal va an electrnc swtch nly durng the ccurrence f the segment nvlved. (en scllscpe serves as a mntr necessary fr adjustment f the cntrls). The sgnal passng thrugh the electrnc swtch s fed t the nput f the measurng flter f the analyzer. After each replay cycle f the memry, the central frequency f the measurng flter s ncreased ne step. The cntrl vltage fr the electrnc swtch serves als as a cntrl sgnal fr the'enablng'and the fdsablng' f the measurng flter. (lsablng ccurs by shrtcrcutng the vltages f the flter va electrnc swtches). tmemry cmpnents wthn the The flter utput s cnnected t a full-wave peak rectfer whch s reset befre each new replay cycle s started. The utput vltage frm the peak rectfer s fed t a strp chart recrder thrugh a lgarthmc cnvertr. After a number f replay cycles f the memry, the flter has swept ver the frequency range f nterest and the plttng s cmpleted.

42 -84- Bandwdth. n rder t attan an utput vltage f sme mprtance frm the nterrupted flter, ts respnse tme shuld nt be t lng cmpared wth the segment duratn (T),.e. narrw band analyss cannt be appled. Hwever, ths desntt mply a lss f reslutn: the hghest pssble reslutn s lmted by the nherent 'pu1se bandwdth' f 1/THz. Obvusly durng ne spectrum plttng a cnstant bandwdth s requred. Althugh the chce f the bandwdth sn't crtcal here, t shuld preferably be made prprtnal t /T n rder t avd the spectral ampltudes beng dependent n the segment length. As the bandr^rdth s cnstant, a lnear frequency scale s justfed. (The pssble reasns fr applyng a lgarthmc frequency sweep are cmpletely gne when analysng shrt sgnal segments). Tunng culd be dne by scannng the memry cntents at dfferent speeds and leavng the flter unaltered. Althugh ths culd be realzed very easly and cancels the prblem f makng a tunable flter, t mples a measurement wth a resultng cnstant percentage bandwdth nstead f a cnstanl bandwdth. The prblem f stepwse crrectng the bandwdth beng cmparable wth the prblem f rnakng a tunable flter, we decded t slve the latter, partcularly n accunt f the taylr-made prpertes f the gyratr band pass prncple. Ths may be explaned by the fllwng descrptn f the gyratr band pass flter prncple: A gyratr (Tellegen, l94b) bascally cnssts f tw mutually cupled 'transcnductance amplfers' (see fg, 22 ). The utput current f an deal amplfer f ths srt s prprtnal t the nput vltage, whereas the nput mpedance s nfnte. One f the amplfers s an nvertng type whch thus causes a negatve feedback. Each amplfer utput frms a termnal (gate) wth regard t the cnrnn cnnectn. Let bth amplfers have equal transcnductances (g). f ne gate s laded wth an mpedance 2,, the effected mpedance (Za)

43 Fg. 22 q l V2 Fg. 23

44 _ 8 6 _ at the ther gate can be cmputated as fllws: A current flwng thrugh the effected mpedance Za causes a vltage v ccurrng acrss t: Z, = v/. The upper amplfer utputs a current, = g.v, whch s flwng entrely thrugh 2,. Hence v, = g.v.zl. The lwer amplfer utputs a current ^=-g.v, =-g2.v.2,. z - l - l As the amplfer nput current s zer t can be stated that 2=-.Hence: - g'.v.z, = - and as Z. = v/l: " l E t (21\ \-./ Z "t = l/p'7., ul f fr Z, a sngle capactance (C) s used, (21) changes t: C / = 1lt\ - t Jw g ' whch behaves lke a self-nductance: a (22) = -"- g2 t wll be clear that a smple resnant crcut can be made by cnnectng the left gate wth a capactance as well. ts resnance frequency results frm: (23),.=G=fl$=t assumng that the tw capactances have equal values. Ths resnance frequency can easly be adjusted by varyng g, whch culd be accmplshed by means f a varable attenuatr at each amplfer nput. n ths way, lnear tunng s very smple t realze. f Zl cnssts f a capactance and a resstance (R) n parallel, Z, changes t: 2 _ R t - Fl,r,cR

45 , 6rr.l! L- = - = -. grzr + trlcr grr Nw Z* can be cnsdered as cnsstng f tw cmpnents t cnnected n seres, namely: a resstance r = + and a g,r self-nductance accrdng t (22). The applcatn f ths means that the flter bandwdth can be determned by ratng the resstrs cnnected acrss the capactances. A schematc set-up f ths s shwn n fg. 23. The fllwng equatns are vald:, = 8'v, 2= E(v. - vr) v, = r'z, v, = fz, Sme substtutns result nt: v g'22 (24) 2= t t+g2zl and v g z 1= t t+g2zl ^7nere,l = R ;JA,CR. 2 1 Wrtng,, = E and, fr 15 changes ths t: (25) Zr = c1;1;; Substtutng (25) nt (24) yelds:

46 uj2 r 62+ (a+j) 2 r, slghtly mdfed: (26) v 2 = g v. C (cl+j) 2 n a smlar wav can be fund: (27) v l = g V, C ]-.+ju) u.r2+(+) 2 r Apart frm the t"r* these results are equal t the cmplex f; spectra f damped snusds wth ntal phases f O and { radans respectvely, as calculated n chapter 2. f the nput vltage f the gyratr flter s dvded by g, ts ampltude respnse s prprtnal wth the ampltude spectra f the damped snusds. Therefre what s fund n chapter 2 cncernng the maxma and bandwdth s cmpletely vald fr the flter respnse. Mrever, f ths tlzpe f band pass flter s used, the 'naccuracy' f the spectral peak lcatns as calculated n that chapter s decreased and even cmpletelv crrected when the flter bandwdth equal s 2cl. f the sgnal gatng s carred ut accrdng t the fnal remark n chapter 2, ptmum crrectn ccurs when v, s cnsdered as the flter utput vltage. Wth reference t the functnal dagram (tg.2a) the wrkng prncple f the segment spectrgraph can nw be descrbed n sme detal. After the recrdng f a part f a sgnal nt the sgnal memry ('transent recrderf) as already descrbed, the sgnal s replayed repettvely at a cnstant rate. The analg utput (Y) f the

47 b 0 + J t r O.Fl ( / ) q J +J A 'F{ $ r r F l { $ 0 t l 0! ]. ( d +J A.rl. l t t O t r A! F l q ) ( d $ t r r r d H V tr.r.r{ 0J.lJ A A F l v b O l J t r. F r a. / l B F r 6 lrl g c d a u ll f.t \ t J a -9 t-g sg td <a G2 l'cl 6 9 Y = t,a F ( J d z l r ja U &!, f- E Z E rrr ;6 r J d a F t! - a g l e c l - j ( n u J l- 2 6 ul ul vt : a &, Z Al lj', l-c * u 33 8! E e a J3 9 v', H gp : F(/' g 0 0. rg l{ A lj tl J J U 0l llr e 6 ul l- J Z b. l E 6 ta a t- d a E A t v d F = \ J a. v d ; ( 4 v A ] g fe ;.$ N b0.rl

48 transent recrder s attenuated by a dgtally cntrllable dvder (usng a multplyng dgtal-t-analg cnvertr) and then led t the nput f the gyratr band pass flter. Tw multplyng dgtal-t-analg cnverrrs serve as dgtally cntrllable attenuatrs, makng the central frequency f the flter prprtnal t the same dgtal number (A) as used fr the nput dvder. The utput vltage f the gyratr s taken frm v, va a buffer arnplf er. Ths results n flterng f the utput frm the transent recrder wth a flter respnse whch s prprtnal t equatn (6) where 0l s prprtnal t A and cx, s cnstant ( = *T). Durng ne cmplete scannng f the transent recrder the number A remans cnstant. The replayed samples n the transent recrder are cunted by a sample cunter. The cuntng result s cmpared wth a number (Nl) whch s preset wth the ad f fthumb wheelt swtches. When bth numbers are equal, a flp-f1p s set, carrsng an analg swtch t cnnect the utput f the transent recrder wth the scllscpe nput. n addtn the flp-f1p enables a prgrarunable cunter whch als cunts the utput samples. After a preset number (NZ) f samples, ths cunter generates a pulse whch clears the flpflp causng the scllscpe nput vltage t becme zer. The desred fundamental perd, r ther segment, can be selected vsually therefre by adjustng N, and Nr. As any sample number belw 1000 can be preset, the adjustment reslutn s 20 psec (the cmplete cntents f the memry beng 1000 samples f a 20 msec sgnal nterval). The nverted utput f the flp-flp cntrls tw analg swtches smultaneusly whch keep the tw capactances f the gyratr flter shrt-crcuted except durng the replay f the selected segment. The number (A) whch determnes the tunng f the flter s frmed by a bnary cunter (9 bt) and s ncremented by ne at each pulse frm a tmng crcut. The same pulse serves t restart the transent recrder scannng cycle and the sample cunter. The peak

49 - 9l vltage f the gyratr flter utput s btaned by a full-wave peak rectfer whch s reset befre the next tunng step. A sample-and-hld crcut changes the vltage drectly frm ne measured value nt the next ne, thus preventng unnecessary large excursns f the wrtng pen f the chart recrder, whch s cnnected va a lgarthmc amplfer. The peak rectfer reset cmmand and the sample-and-h1d cntrl are generated by the trnng crcut. Ths crcut ensures a tme lapse between cnsecutve tunng steps whch s suffcent fr plttng the spectrum accurately n spte f the relatvely lw wrtng speed f the chart recrder. The tmng crcut s drven by the same clck frequency as the ne whch cntrls the utput rate f the transent recrder. Durng the tme 1apse, the flp-flp s nhbted n rder t prevent the gyratr flter frm respndng t mprper nput sgnals. The replay velcty f the transent recrder s made prprtnal wth the length f the selected segment by makng the utput clck frequency equal t the prduct f a reference frequency and the number N, (acheved by usng the 'phase-lcked-lp' prncple). The reasn f ths may be explaned by the fllwng: Reducng the replay velcty by a factr k causes a decrease f the sgnal frequency cmpnents by the same factr. As the tunng steps and the bandwdth f the gyratr flter reman the same, the result s equvalent t flterng wth ncreased bandwdth and tunng steps bth by the factr k. Hence the flter respnse tme and frequency reslutn s matched wth the selected segment length. The peak vltages frm the flter utput therefre are ndependent f the segment length. Because the trnng crcut s slwed dwn at the same tme, the tunng tme s rased by the same factr. The result s that the sweep speed remans cnstant. As the chart speed f the recrder remans unaltered as we1l, the frequency scale f the btaned spectra s ndependent f the segment length. Furthermre the frequeney range ncreases wth the same factr n cnsequence.

50 n rder t avd any prblems caused by spurus capactances, the frequency sweep f the gyratr flter has been shfted t a lwer area tgether wth the utput clck frequency. Owng t the lmted wrtng speed f the chart recrder the tme nsertn between cnsecutve tunng steps s large enugh t a11w fr replayng at slwer rates wthut the cnsequence f ncreasng the analyss tme. The actual value f k s chsen t be 5 when a l0 msec sgnal segment s selected (.". N2 = 500). n that case the utput clck frequency equals l0 khz (whch accunts fr scannng the cmplete memry n 100 msec). Therefre the reference frequency equals l0 k[z/500 = 20 Hz. The reslutn bandwdth and tunng step wdth at a l0 msec segment were made 50 Hz and 162/" Hz respectvely J (the number f tunng steps wthn ne bandwdth was chsen t be 3). That causes a lo Hz bandwdth and a 31 / " tt, tunng step wdth 5 fr the gyratr flter as n ths case k = 5. The chart speed beng 2 nm/sec and a frequency scale f 200 Hz/ cm cause the requred tunng speed t be 40 Hz/sec. f the tunng step wdth s 1621rH", J the tunng speed requres 40/162/3 = 2.4 steps/sec. Thus the utput clck frequency (10 khz) must be dvded by 4167 t bt.an the requred step frequency. Cmputatn f sme prpertes f the respnses f the nterrupted flter system t the smplfed artfcal vwel sunds as used befre ddn't seem a verv smple matter and therefre we decded t check whether r nt ths system wuld at all functn satsfactrly n practce. At a later stage the theretcal behavur culd be wrked ut. The fgs 25 thrugh 29 shw sme utputs btaned frm ths nterrupted flter system at dfferent nput test sgnals. The utputs frm a few natural vwel perds are represented by the fgs 30 thrugh 32. t can be cncluded that the 'sde lbest have cmpletelv dsappeared and the frmants are dsplaved cnvenentlv.

51 L -93- "lf l-t f-.1 t^ l l t l T T 1 'l {- L t-- h t--. t tr, h T - -T-- tr- -T J L r-r t l l : t -l fe t_, l t l l f---t \ J J J -1 f + J t 1 J \, \ l E f \ ' t / T l L L \ L 4 E \ :-\ 4 tf.r r tll _l tt l! \ \ --f j l l j Fg. 25 A: Spectrum f 2O perds f l000hz snusd (T=20msec) B: Spectrum f 2 perds f l000hz snusd (T=2msec)

52 -94- ) E.$ tl H (u F 0,, l{ cl u rt (a N rn N t}{ t lr A ' r q..l a }{ +J C, A tn!?,-l:l-

53 -95- Fg.27 r_ l t, Spectrum f the sun f Fr=1450H2, T=lTmsec 2 danped snusds, Fl=l2OOHz, Fg. 28 Spectrum f the sum f F,=40OHz, T= Oms ec 2 danped snusds, E r=250h2,

54 -96- :T-]-*T-T* r- t t t t t l l --l--r- -l l L r l. l l l l t l t l --+ +#--l] sas--l _t _J l_ l l l 1,1 1 l +"-t Fs.29 A: Spectrum f the sum f 2 danped F2=l500ttzr T=8msec snusds, Fr=850H2, B: As A except T=3msec

55 -97 - r f 1 l l, r l r l - ", l :--f q) t Fr --l E ] -;----l : l t :----r ---'l ] t s l t l-_'l a f > t '_ - l l t --- Fl d l r r *rr r l l, F;- \ l r-r N f f t, t t,j lj') 1 r l r l \ r l l r : l l \. t :'+--F; tl l ', l r- 1,, llt \ l l + { ; - - A / +J --;- A a : t n :-:-- ' (a t). r.l t' l\; )- \ l :1- _L a,l r l U \ -l t ( -r L 1 ',,t tr { 1 l T a : t t l l l L, l : l ) r L --l- --L- T U u 0. F F tr e

56 -98-, E Or.f, tl ts ) ql trl c) tl{ r+{ E a lr TJ & u)

57 r \ l l"-- ^ L ts"-* T L ctl3 a l t-- T- t-- J t ] 5 t u \ 00flz t, Ī l A 3 \ jl t \ t l _0_ { j -t l< :\ OmnlScrlhe* charr rype Ec-r Fg.32 Spectrum f fenale [], T =5.3 msec

58 Even at a segment length f ne r tw perds f the tfrmant frequencyt, ts detectn s pssble, althugh ts spectral peak shfts smewhat t a lwer value. As ths shft s a functn f the segment length, the measure f a nt aceunl. crrectn mght be taken 5. CONCLUSON The expnental wndw analyss and the nterrupted flter analyss as descrbed prve t be valuable methds fr the analvss f vced speech sgnal segments. The latter methd cmpletely cancels the nfluence f the fundamental frequency n the spectrum (the well-knrnm tsde lbes') f ne perd f the fundamental frequency f the sgnal s selected. Addtnallv the frequency reslutn s lnted by ether the segment length r the ntrnsc frmant bandwdth and nt by the analyss methd. The segment spectrgraph devce based n the nterrupted flter methd therefre can be cnsdered as a useful tl fr frmant extractn f speech sgnals per perd, hgh-ptched sgnals ncluded. Thus fr example relable frmant trajectrv measurements culd be carred ut. The requrements as mentned at the end f the ntrductn therefre can be cnsdered as fu1f1led. Naturally, the applcatn f the devce culd be extended t ther speech sunds such as frcatves and furthermre t sgnals utsde the speech fe1d. Acknwledgments. rd lke t thank all peple wh, n ne way r anther, helped t carry ut ths prject and n partcular Drs.A.C.M.Retveld fr hs stmulatng c-peratn n an early stage f the experments and Mr.A.W.van Maanen fr hs mprtant rle n the cnstructn f the segment spectrgraph.

59 - l0l BBLOGRAPHY l. Brch, J.T. and Olesen, H.P. On the frequency analyss f mechancal shcks and sngle mpulses Brel & Kjar Techncal Revew n Engelsn, M. Spectrum Analyzer Measurements Tektrnx nc Lynn, P.A. The Analyss and Prcessng f Sgnals Macmllan Markel, J.D. Dgtal nverse flterng, a new tl fr frmant trajectrv es tmatn EEE Transactns n Aud and Electracustcs 1972, AU-20,129-t37 5. Nerp, D.J.P.J.van, P1s, L.C.W. and P1mp, R. Frequency analyss f Dutch vwels frm 25 female speakers Acustca Y7.29 Hett Papuls, A. The Furer ntegral Mc Graw-Hll 1962 and ts Applcatns 7. Ptter, R.K. and Stenberg, J.C. Tward the specfcatn f speech J.A.S.A. v1.22 n 6 p Randall, R.B. Frequency Analyss Brel & Kjer 1977

60 -tz- 9. Randa11, R.B. Hgh speed narrw band analyss usng the dgtal recrder type 7502 Brel & Kjer Techncal Revew n event 10. Tellegen, B.D.H. The gyratr, a new electrc netwrk element Phlps Research Reprts n 3 l94b p. 8l l. Wempe, A.G. A system fr gatng segments f taped speech wth great precsn PrceednBs 4, nsttute f Phnetc Scences Amsterdam, 1976

Introduction to Electronic circuits.

Introduction to Electronic circuits. Intrductn t Electrnc crcuts. Passve and Actve crcut elements. Capactrs, esstrs and Inductrs n AC crcuts. Vltage and current dvders. Vltage and current surces. Amplfers, and ther transfer characterstc.

More information

Feedback Principle :-

Feedback Principle :- Feedback Prncple : Feedback amplfer s that n whch a part f the utput f the basc amplfer s returned back t the nput termnal and mxed up wth the nternal nput sgnal. The sub netwrks f feedback amplfer are:

More information

Circuits Op-Amp. Interaction of Circuit Elements. Quick Check How does closing the switch affect V o and I o?

Circuits Op-Amp. Interaction of Circuit Elements. Quick Check How does closing the switch affect V o and I o? Crcuts Op-Amp ENGG1015 1 st Semester, 01 Interactn f Crcut Elements Crcut desgn s cmplcated by nteractns amng the elements. Addng an element changes vltages & currents thrughut crcut. Example: clsng a

More information

55:041 Electronic Circuits

55:041 Electronic Circuits 55:04 Electrnc Crcuts Feedback & Stablty Sectns f Chapter 2. Kruger Feedback & Stablty Cnfguratn f Feedback mplfer S S S S fb Negate feedback S S S fb S S S S S β s the feedback transfer functn Implct

More information

EE 221 Practice Problems for the Final Exam

EE 221 Practice Problems for the Final Exam EE 1 Practce Prblems fr the Fnal Exam 1. The netwrk functn f a crcut s 1.5 H. ω 1+ j 500 Ths table recrds frequency respnse data fr ths crcut. Fll n the blanks n the table:. The netwrk functn f a crcut

More information

CHAPTER 3: FEEDBACK. Dr. Wan Mahani Hafizah binti Wan Mahmud

CHAPTER 3: FEEDBACK. Dr. Wan Mahani Hafizah binti Wan Mahmud CHPTER 3: FEEDBCK Dr. Wan Mahan Hafzah bnt Wan Mahmud Feedback ntrductn Types f Feedback dvantages, Characterstcs and effect f Negatve Feedback mplfers Crcuts wth negatve feedback Pstve feedback and Oscllatr

More information

CHAPTER 3 ANALYSIS OF KY BOOST CONVERTER

CHAPTER 3 ANALYSIS OF KY BOOST CONVERTER 70 CHAPTER 3 ANALYSIS OF KY BOOST CONERTER 3.1 Intrductn The KY Bst Cnverter s a recent nventn made by K.I.Hwu et. al., (2007), (2009a), (2009b), (2009c), (2010) n the nn-slated DC DC cnverter segment,

More information

PHYSICS 536 Experiment 12: Applications of the Golden Rules for Negative Feedback

PHYSICS 536 Experiment 12: Applications of the Golden Rules for Negative Feedback PHYSICS 536 Experment : Applcatns f the Glden Rules fr Negatve Feedback The purpse f ths experment s t llustrate the glden rules f negatve feedback fr a varety f crcuts. These cncepts permt yu t create

More information

Design of Analog Integrated Circuits

Design of Analog Integrated Circuits Desgn f Analg Integrated Crcuts I. Amplfers Desgn f Analg Integrated Crcuts Fall 2012, Dr. Guxng Wang 1 Oerew Basc MOS amplfer structures Cmmn-Surce Amplfer Surce Fllwer Cmmn-Gate Amplfer Desgn f Analg

More information

Chapter 7. Systems 7.1 INTRODUCTION 7.2 MATHEMATICAL MODELING OF LIQUID LEVEL SYSTEMS. Steady State Flow. A. Bazoune

Chapter 7. Systems 7.1 INTRODUCTION 7.2 MATHEMATICAL MODELING OF LIQUID LEVEL SYSTEMS. Steady State Flow. A. Bazoune Chapter 7 Flud Systems and Thermal Systems 7.1 INTODUCTION A. Bazune A flud system uses ne r mre fluds t acheve ts purpse. Dampers and shck absrbers are eamples f flud systems because they depend n the

More information

SIMULATION OF THREE PHASE THREE LEG TRANSFORMER BEHAVIOR UNDER DIFFERENT VOLTAGE SAG TYPES

SIMULATION OF THREE PHASE THREE LEG TRANSFORMER BEHAVIOR UNDER DIFFERENT VOLTAGE SAG TYPES SIMULATION OF THREE PHASE THREE LEG TRANSFORMER BEHAVIOR UNDER DIFFERENT VOLTAGE SAG TYPES Mhammadreza Dlatan Alreza Jallan Department f Electrcal Engneerng, Iran Unversty f scence & Technlgy (IUST) e-mal:

More information

Chapter 3, Solution 1C.

Chapter 3, Solution 1C. COSMOS: Cmplete Onlne Slutns Manual Organzatn System Chapter 3, Slutn C. (a If the lateral surfaces f the rd are nsulated, the heat transfer surface area f the cylndrcal rd s the bttm r the tp surface

More information

IGEE 401 Power Electronic Systems. Solution to Midterm Examination Fall 2004

IGEE 401 Power Electronic Systems. Solution to Midterm Examination Fall 2004 Jós, G GEE 401 wer Electrnc Systems Slutn t Mdterm Examnatn Fall 2004 Specal nstructns: - Duratn: 75 mnutes. - Materal allwed: a crb sheet (duble sded 8.5 x 11), calculatr. - Attempt all questns. Make

More information

Wp/Lmin. Wn/Lmin 2.5V

Wp/Lmin. Wn/Lmin 2.5V UNIVERITY OF CALIFORNIA Cllege f Engneerng Department f Electrcal Engneerng and Cmputer cences Andre Vladmrescu Hmewrk #7 EEC Due Frday, Aprl 8 th, pm @ 0 Cry Prblem #.5V Wp/Lmn 0.0V Wp/Lmn n ut Wn/Lmn.5V

More information

Shell Stiffness for Diffe ent Modes

Shell Stiffness for Diffe ent Modes Engneerng Mem N 28 February 0 979 SUGGESTONS FOR THE DEFORMABLE SUBREFLECTOR Sebastan vn Herner Observatns wth the present expermental versn (Engneerng Dv nternal Reprt 09 July 978) have shwn that a defrmable

More information

Section 3: Detailed Solutions of Word Problems Unit 1: Solving Word Problems by Modeling with Formulas

Section 3: Detailed Solutions of Word Problems Unit 1: Solving Word Problems by Modeling with Formulas Sectn : Detaled Slutns f Wrd Prblems Unt : Slvng Wrd Prblems by Mdelng wth Frmulas Example : The factry nvce fr a mnvan shws that the dealer pad $,5 fr the vehcle. If the stcker prce f the van s $5,, hw

More information

PT326 PROCESS TRAINER

PT326 PROCESS TRAINER PT326 PROCESS TRAINER 1. Descrptn f the Apparatus PT 326 Prcess Traner The PT 326 Prcess Traner mdels cmmn ndustral stuatns n whch temperature cntrl s requred n the presence f transprt delays and transfer

More information

ME2142/ME2142E Feedback Control Systems. Modelling of Physical Systems The Transfer Function

ME2142/ME2142E Feedback Control Systems. Modelling of Physical Systems The Transfer Function Mdellng Physcal Systems The Transer Functn Derental Equatns U Plant Y In the plant shwn, the nput u aects the respnse the utput y. In general, the dynamcs ths respnse can be descrbed by a derental equatn

More information

Lecture 12. Heat Exchangers. Heat Exchangers Chee 318 1

Lecture 12. Heat Exchangers. Heat Exchangers Chee 318 1 Lecture 2 Heat Exchangers Heat Exchangers Chee 38 Heat Exchangers A heat exchanger s used t exchange heat between tw fluds f dfferent temperatures whch are separated by a sld wall. Heat exchangers are

More information

Transient Conduction: Spatial Effects and the Role of Analytical Solutions

Transient Conduction: Spatial Effects and the Role of Analytical Solutions Transent Cnductn: Spatal Effects and the Rle f Analytcal Slutns Slutn t the Heat Equatn fr a Plane Wall wth Symmetrcal Cnvectn Cndtns If the lumped capactance apprxmatn can nt be made, cnsderatn must be

More information

III. Operational Amplifiers

III. Operational Amplifiers III. Operatnal Amplfers Amplfers are tw-prt netwrks n whch the utput vltage r current s drectly prprtnal t ether nput vltage r current. Fur dfferent knds f amplfers ext: ltage amplfer: Current amplfer:

More information

Exercises H /OOA> f Wo AJoTHS l^»-l S. m^ttrt /A/ ?C,0&L6M5 INFERENCE FOR DISTRIBUTIONS OF CATEGORICAL DATA. tts^e&n tai-ns 5 2%-cas-hews^, 27%

Exercises H /OOA> f Wo AJoTHS l^»-l S. m^ttrt /A/ ?C,0&L6M5 INFERENCE FOR DISTRIBUTIONS OF CATEGORICAL DATA. tts^e&n tai-ns 5 2%-cas-hews^, 27% /A/ mttrt?c,&l6m5 INFERENCE FOR DISTRIBUTIONS OF CATEGORICAL DATA Exercses, nuts! A cmpany clams that each batch f ttse&n ta-ns 5 2%-cas-hews, 27% almnds, 13% macadama nuts, and 8% brazl nuts. T test ths

More information

Conservation of Energy

Conservation of Energy Cnservatn f Energy Equpment DataStud, ruler 2 meters lng, 6 n ruler, heavy duty bench clamp at crner f lab bench, 90 cm rd clamped vertcally t bench clamp, 2 duble clamps, 40 cm rd clamped hrzntally t

More information

The three major operations done on biological signals using Op-Amp:

The three major operations done on biological signals using Op-Amp: The three majr peratns dne n blgcal sgnals usng Op-Amp: ) Amplcatns and Attenuatns 2) DC settng: add r subtract a DC 3) Shape ts requency cntent: Flterng Ideal Op-Amp Mst belectrc sgnals are small and

More information

element k Using FEM to Solve Truss Problems

element k Using FEM to Solve Truss Problems sng EM t Slve Truss Prblems A truss s an engneerng structure cmpsed straght members, a certan materal, that are tpcall pn-ned at ther ends. Such members are als called tw-rce members snce the can nl transmt

More information

EE 204 Lecture 25 More Examples on Power Factor and the Reactive Power

EE 204 Lecture 25 More Examples on Power Factor and the Reactive Power EE 204 Lecture 25 Mre Examples n Pwer Factr and the Reactve Pwer The pwer factr has been defned n the prevus lecture wth an example n pwer factr calculatn. We present tw mre examples n ths lecture. Example

More information

A New Method for Solving Integer Linear. Programming Problems with Fuzzy Variables

A New Method for Solving Integer Linear. Programming Problems with Fuzzy Variables Appled Mathematcal Scences, Vl. 4, 00, n. 0, 997-004 A New Methd fr Slvng Integer Lnear Prgrammng Prblems wth Fuzzy Varables P. Pandan and M. Jayalakshm Department f Mathematcs, Schl f Advanced Scences,

More information

CTN 2/23/16. EE 247B/ME 218: Introduction to MEMS Design Lecture 11m2: Mechanics of Materials. Copyright 2016 Regents of the University of California

CTN 2/23/16. EE 247B/ME 218: Introduction to MEMS Design Lecture 11m2: Mechanics of Materials. Copyright 2016 Regents of the University of California Vlume Change fr a Unaxal Stress Istrpc lastcty n 3D Istrpc = same n all drectns The cmplete stress-stran relatns fr an strpc elastc Stresses actng n a dfferental vlume element sld n 3D: (.e., a generalzed

More information

ANALOG ELECTRONICS 1 DR NORLAILI MOHD NOH

ANALOG ELECTRONICS 1 DR NORLAILI MOHD NOH 24 ANALOG LTRONIS TUTORIAL DR NORLAILI MOHD NOH . 0 8kΩ Gen, Y β β 00 T F 26, 00 0.7 (a)deterne the dc ltages at the 3 X ternals f the JT (,, ). 0kΩ Z (b) Deterne g,r π and r? (c) Deterne the ltage gan

More information

FE REVIEW OPERATIONAL AMPLIFIERS (OP-AMPS)( ) 8/25/2010

FE REVIEW OPERATIONAL AMPLIFIERS (OP-AMPS)( ) 8/25/2010 FE REVEW OPERATONAL AMPLFERS (OP-AMPS)( ) 1 The Op-amp 2 An op-amp has two nputs and one output. Note the op-amp below. The termnal labeled l wth the (-) sgn s the nvertng nput and the nput labeled wth

More information

Conduction Heat Transfer

Conduction Heat Transfer Cnductn Heat Transfer Practce prblems A steel ppe f cnductvty 5 W/m-K has nsde and utsde surface temperature f C and 6 C respectvely Fnd the heat flw rate per unt ppe length and flux per unt nsde and per

More information

V. Electrostatics Lecture 27a: Diffuse charge at electrodes

V. Electrostatics Lecture 27a: Diffuse charge at electrodes V. Electrstatcs Lecture 27a: Dffuse charge at electrdes Ntes by MIT tudent We have talked abut the electrc duble structures and crrespndng mdels descrbng the n and ptental dstrbutn n the duble layer. Nw

More information

4DVAR, according to the name, is a four-dimensional variational method.

4DVAR, according to the name, is a four-dimensional variational method. 4D-Varatnal Data Assmlatn (4D-Var) 4DVAR, accrdng t the name, s a fur-dmensnal varatnal methd. 4D-Var s actually a smple generalzatn f 3D-Var fr bservatns that are dstrbuted n tme. he equatns are the same,

More information

Linear Amplifiers and OpAmps

Linear Amplifiers and OpAmps Lnear Amplfers and OpAmps eferences: Barbw (pp 7-80), Hayes & Hrwtz (pp 63-40), zzn (Chapter ) Amplfers are tw-prt netwrks n whch the utput ltage r current s drectly prprtnal t ether nput ltage r current.

More information

Waveshapping Circuits and Data Converters. Lesson #17 Comparators and Schmitt Triggers Section BME 373 Electronics II J.

Waveshapping Circuits and Data Converters. Lesson #17 Comparators and Schmitt Triggers Section BME 373 Electronics II J. Waeshappg Crcuts and Data Cnerters Lessn #7 Cmparatrs and Schmtt Trggers Sectn. BME 7 Electrncs II 0 Waeshappg Crcuts and Data Cnerters Cmparatrs and Schmtt Trggers Astable Multbratrs and Tmers ectfers,

More information

Faculty of Engineering

Faculty of Engineering Faculty f Engneerng DEPARTMENT f ELECTRICAL AND ELECTRONIC ENGINEERING EEE 223 Crcut Thery I Instructrs: M. K. Uygurğlu E. Erdl Fnal EXAMINATION June 20, 2003 Duratn : 120 mnutes Number f Prblems: 6 Gd

More information

A Note on the Linear Programming Sensitivity. Analysis of Specification Constraints. in Blending Problems

A Note on the Linear Programming Sensitivity. Analysis of Specification Constraints. in Blending Problems Aled Mathematcal Scences, Vl. 2, 2008, n. 5, 241-248 A Nte n the Lnear Prgrammng Senstvty Analyss f Secfcatn Cnstrants n Blendng Prblems Umt Anc Callway Schl f Busness and Accuntancy Wae Frest Unversty,

More information

CIRCUIT ANALYSIS II Chapter 1 Sinusoidal Alternating Waveforms and Phasor Concept. Sinusoidal Alternating Waveforms and

CIRCUIT ANALYSIS II Chapter 1 Sinusoidal Alternating Waveforms and Phasor Concept. Sinusoidal Alternating Waveforms and U ANAYSS hapter Snusdal Alternatng Wavefrs and Phasr ncept Snusdal Alternatng Wavefrs and Phasr ncept ONNS. Snusdal Alternatng Wavefrs.. General Frat fr the Snusdal ltage & urrent.. Average alue..3 ffectve

More information

Module B3. VLoad = = V S V LN

Module B3. VLoad = = V S V LN Mdule B Prblem The -hase lads are cnnected n arallel. One s a urely resste lad cnnected n wye. t cnsumes 00kW. The secnd s a urely nducte 00kR lad cnnected n wye. The thrd s a urely caacte 00kR lad cnnected

More information

Physic 231 Lecture 33

Physic 231 Lecture 33 Physc 231 Lecture 33 Man pnts f tday s lecture: eat and heat capacty: Q cm Phase transtns and latent heat: Q Lm ( ) eat flw Q k 2 1 t L Examples f heat cnductvty, R values fr nsulatrs Cnvectn R L / k Radatn

More information

CONVEX COMBINATIONS OF ANALYTIC FUNCTIONS

CONVEX COMBINATIONS OF ANALYTIC FUNCTIONS rnat. J. Math. & Math. S. Vl. 6 N. (983) 33534 335 ON THE RADUS OF UNVALENCE OF CONVEX COMBNATONS OF ANALYTC FUNCTONS KHALDA. NOOR, FATMA M. ALOBOUD and NAEELA ALDHAN Mathematcs Department Scence Cllege

More information

CHAPTER 3 QUASI-RESONANT BUCK CONVERTER

CHAPTER 3 QUASI-RESONANT BUCK CONVERTER 27 CHAPTER 3 QUASI-RESONANT BUCK CONVERTER Hstrcally, prr t the avalablty f cntrllable swtch wth apprecable vltage and current-handlng capablty, the swtch-mde DC-DC cnverter cnssts f thyrstrs whch pertans

More information

State-Space Model Based Generalized Predictive Control for Networked Control Systems

State-Space Model Based Generalized Predictive Control for Networked Control Systems Prceedngs f the 7th Wrld Cngress he Internatnal Federatn f Autmatc Cntrl State-Space Mdel Based Generalzed Predctve Cntrl fr Netwred Cntrl Systems Bn ang* Gu-Png Lu** We-Hua Gu*** and Ya-Ln Wang**** *Schl

More information

Lesson 5. Thermomechanical Measurements for Energy Systems (MENR) Measurements for Mechanical Systems and Production (MMER)

Lesson 5. Thermomechanical Measurements for Energy Systems (MENR) Measurements for Mechanical Systems and Production (MMER) Lessn 5 Thermmechancal Measurements r Energy Systems (MEN) Measurements r Mechancal Systems and Prductn (MME) A.Y. 205-6 Zaccara (n ) Del Prete We wll nw analyze mre n depth each ne the unctnal blcks the

More information

Reproducing kernel Hilbert spaces. Nuno Vasconcelos ECE Department, UCSD

Reproducing kernel Hilbert spaces. Nuno Vasconcelos ECE Department, UCSD Reprucng ernel Hlbert spaces Nun Vascncels ECE Department UCSD Classfcatn a classfcatn prblem has tw tpes f varables X -vectr f bservatns features n the wrl Y - state class f the wrl Perceptrn: classfer

More information

Physics 107 HOMEWORK ASSIGNMENT #20

Physics 107 HOMEWORK ASSIGNMENT #20 Physcs 107 HOMEWORK ASSIGNMENT #0 Cutnell & Jhnsn, 7 th etn Chapter 6: Prblems 5, 7, 74, 104, 114 *5 Cncept Smulatn 6.4 prves the ptn f explrng the ray agram that apples t ths prblem. The stance between

More information

( ) = ( ) + ( 0) ) ( )

( ) = ( ) + ( 0) ) ( ) EETOMAGNETI OMPATIBIITY HANDBOOK 1 hapter 9: Transent Behavor n the Tme Doman 9.1 Desgn a crcut usng reasonable values for the components that s capable of provdng a tme delay of 100 ms to a dgtal sgnal.

More information

Water vapour balance in a building moisture exposure for timber structures

Water vapour balance in a building moisture exposure for timber structures Jnt Wrkshp f COST Actns TU1 and E55 September 21-22 9, Ljubljana, Slvena Water vapur balance n a buldng msture expsure fr tmber structures Gerhard Fnk ETH Zurch, Swtzerland Jchen Köhler ETH Zurch, Swtzerland

More information

Department of Electrical Engineering, University of Waterloo. Introduction

Department of Electrical Engineering, University of Waterloo. Introduction Sectin 4: Sequential Circuits Majr Tpics Types f sequential circuits Flip-flps Analysis f clcked sequential circuits Mre and Mealy machines Design f clcked sequential circuits State transitin design methd

More information

The Operational Amplifier and Application

The Operational Amplifier and Application Intrductn t Electrnc Crcuts: A Desgn Apprach Jse Sla-Martnez and Marn Onabaj The Operatnal Amplfer and Applcatn The peratnal ltage amplfer (mre cmmnly referred t as peratnal amplfer) s ne f the mst useful

More information

Sections 15.1 to 15.12, 16.1 and 16.2 of the textbook (Robbins-Miller) cover the materials required for this topic.

Sections 15.1 to 15.12, 16.1 and 16.2 of the textbook (Robbins-Miller) cover the materials required for this topic. Tpic : AC Fundamentals, Sinusidal Wavefrm, and Phasrs Sectins 5. t 5., 6. and 6. f the textbk (Rbbins-Miller) cver the materials required fr this tpic.. Wavefrms in electrical systems are current r vltage

More information

Dead-beat controller design

Dead-beat controller design J. Hetthéssy, A. Barta, R. Bars: Dead beat cntrller design Nvember, 4 Dead-beat cntrller design In sampled data cntrl systems the cntrller is realised by an intelligent device, typically by a PLC (Prgrammable

More information

Bipolar-Junction (BJT) transistors

Bipolar-Junction (BJT) transistors Bplar-Junctn (BJT) transstrs References: Hayes & Hrwtz (pp 84-4), Rzzn (chapters 8 & 9) A bplar junctn transstr s frmed by jnng three sectns f semcnductrs wth alternately dfferent dpngs. The mddle sectn

More information

AP Physics Kinematic Wrap Up

AP Physics Kinematic Wrap Up AP Physics Kinematic Wrap Up S what d yu need t knw abut this mtin in tw-dimensin stuff t get a gd scre n the ld AP Physics Test? First ff, here are the equatins that yu ll have t wrk with: v v at x x

More information

14 The Boole/Stone algebra of sets

14 The Boole/Stone algebra of sets 14 The Ble/Stne algebra f sets 14.1. Lattces and Blean algebras. Gven a set A, the subsets f A admt the fllwng smple and famlar peratns n them: (ntersectn), (unn) and - (cmplementatn). If X, Y A, then

More information

Chapter 6 : Gibbs Free Energy

Chapter 6 : Gibbs Free Energy Wnter 01 Chem 54: ntrductry hermdynamcs Chapter 6 : Gbbs Free Energy... 64 Defntn f G, A... 64 Mawell Relatns... 65 Gbbs Free Energy G(,) (ure substances)... 67 Gbbs Free Energy fr Mtures... 68 ΔG f deal

More information

Regression with Stochastic Regressors

Regression with Stochastic Regressors Sectn 9 Regressn wth Stchastc Regressrs Meanng f randm regressrs Untl nw, we have assumed (aganst all reasn) that the values f x have been cntrlled by the expermenter. Ecnmsts almst never actually cntrl

More information

ELG4139: Op Amp-based Active Filters

ELG4139: Op Amp-based Active Filters ELG439: Op Amp-baed Actve Flter Advantage: educed ze and weght, and therere paratc. Increaed relablty and mprved perrmance. Smpler degn than r pave lter and can realze a wder range unctn a well a prvdng

More information

Problem Set 5 Solutions - McQuarrie Problems 3.20 MIT Dr. Anton Van Der Ven

Problem Set 5 Solutions - McQuarrie Problems 3.20 MIT Dr. Anton Van Der Ven Prblem Set 5 Slutns - McQuarre Prblems 3.0 MIT Dr. Antn Van Der Ven Fall Fall 003 001 Prblem 3-4 We have t derve the thermdynamc prpertes f an deal mnatmc gas frm the fllwng: = e q 3 m = e and q = V s

More information

Section 10 Regression with Stochastic Regressors

Section 10 Regression with Stochastic Regressors Sectn 10 Regressn wth Stchastc Regressrs Meanng f randm regressrs Untl nw, we have assumed (aganst all reasn) that the values f x have been cntrlled by the expermenter. Ecnmsts almst never actually cntrl

More information

A/2 l,k. Problem 1 STRATEGY. KNOWN Resistance of a complete spherical shell: r rk. Inner and outer radii

A/2 l,k. Problem 1 STRATEGY. KNOWN Resistance of a complete spherical shell: r rk. Inner and outer radii Prblem 1 STRATEGY KNOWN Resstance f a cmplete sphercal shell: R ( r r / (4 π r rk sphere Inner an uter ra r an r, SOLUTION Part 1: Resstance f a hemsphercal shell: T calculate the resstance f the hemsphere,

More information

Physics 2010 Motion with Constant Acceleration Experiment 1

Physics 2010 Motion with Constant Acceleration Experiment 1 . Physics 00 Mtin with Cnstant Acceleratin Experiment In this lab, we will study the mtin f a glider as it accelerates dwnhill n a tilted air track. The glider is supprted ver the air track by a cushin

More information

Big Data Analytics! Special Topics for Computer Science CSE CSE Mar 31

Big Data Analytics! Special Topics for Computer Science CSE CSE Mar 31 Bg Data Analytcs! Specal Tpcs fr Cmputer Scence CSE 4095-001 CSE 5095-005! Mar 31 Fe Wang Asscate Prfessr Department f Cmputer Scence and Engneerng fe_wang@ucnn.edu Intrductn t Deep Learnng Perceptrn In

More information

Concurrent Adaptive Cancellation of Quantization Noise and Harmonic Distortion in Sigma Delta Converter

Concurrent Adaptive Cancellation of Quantization Noise and Harmonic Distortion in Sigma Delta Converter Internatnal Jurnal f Cmputer Engneerng Scence (IJCES) Vlume 2 Issue 11 (vember 2012) ISS : 2250:3439 https://stes.ggle.cm/ste/jcesjurnal http://www.jces.cm/ Cncurrent Adaptve Cancellatn f Quantzatn se

More information

The two main types of FETs are the junction field effect transistor (JFET) and the metal oxide field effect transistor (MOSFET).

The two main types of FETs are the junction field effect transistor (JFET) and the metal oxide field effect transistor (MOSFET). Mcrelectrncs Chapter three: Feld Effect Transstr sall snal analyss Intrductn: Feld-effect transstr aplfers prde an excellent ltae an wth the added feature f hh nput pedance. They are als lw-pwercnsuptn

More information

Difference Equations

Difference Equations Dfference Equatons c Jan Vrbk 1 Bascs Suppose a sequence of numbers, say a 0,a 1,a,a 3,... s defned by a certan general relatonshp between, say, three consecutve values of the sequence, e.g. a + +3a +1

More information

6. ELUTRIATION OF PARTICLES FROM FLUIDIZED BEDS

6. ELUTRIATION OF PARTICLES FROM FLUIDIZED BEDS 6. ELUTRIATION OF PARTICLES FROM FLUIDIZED BEDS Elutratn s the prcess n whch fne partcles are carred ut f a fludzed bed due t the flud flw rate passng thrugh the bed. Typcally, fne partcles are elutrated

More information

Revision: August 19, E Main Suite D Pullman, WA (509) Voice and Fax

Revision: August 19, E Main Suite D Pullman, WA (509) Voice and Fax .7.4: Direct frequency dmain circuit analysis Revisin: August 9, 00 5 E Main Suite D Pullman, WA 9963 (509) 334 6306 ice and Fax Overview n chapter.7., we determined the steadystate respnse f electrical

More information

Chapter II Circuit Analysis Fundamentals

Chapter II Circuit Analysis Fundamentals Chapter II Crcut nalyss Fundamentals Frm a desgn engneer s perspecte, t s mre releant t understand a crcut s peratn and lmtatns than t fnd eact mathematcal epressns r eact numercal slutns. Precse results

More information

A method of constructing rock-analysis diagrams a statistical basks.

A method of constructing rock-analysis diagrams a statistical basks. 130 A methd f cnstructng rck-analyss dagrams a statstcal basks. 0T~ By W. ALF~.D ll~ch).ra)so.~, ~.Se., B.Se. (Eng.), F.G.S. Lecturer n Petrlgy, Unversty Cllege, Nttngham. [Read January 18, 1921.] D R.

More information

Novel current mode AC/AC converters with high frequency ac link *

Novel current mode AC/AC converters with high frequency ac link * vel current mde AC/AC cnverters wth hgh frequency ac lnk * Dalan Chen, e, Jan u, Shengyang n, Chen Sng Department f Electrcal Engneerng, anjng nversty f Aernautcs & Astrnautcs, anjng, Jangsu, 006 P.R.Chna

More information

A L A BA M A L A W R E V IE W

A L A BA M A L A W R E V IE W A L A BA M A L A W R E V IE W Volume 52 Fall 2000 Number 1 B E F O R E D I S A B I L I T Y C I V I L R I G HT S : C I V I L W A R P E N S I O N S A N D TH E P O L I T I C S O F D I S A B I L I T Y I N

More information

Comparison between Back-to-Back and Matrix Converters Based on Thermal Stress of the Switches

Comparison between Back-to-Back and Matrix Converters Based on Thermal Stress of the Switches Cmparsn between Back-t-Back and Matrx Cnverters Based n Thermal Stress f the Swtches D. Casade, Member, EEE,G. Grand, Member, EEE,C. Rss, A. Trentn, L. Zarr Dpartment d ngegnera Elettrca, Unverstà d Blgna,

More information

Fast Acquisition Digital Tanlock Loop with Adaptive Time Delay

Fast Acquisition Digital Tanlock Loop with Adaptive Time Delay Fast Acqustn Dgtal Tanlck Lp wth Adaptve Tme Delay NAWAF AL-MOOSA, SALEH Al-ARAJI AND MAHMOUD AL-QUTAYRI Cllege f Engneerng and Infrmatn Scences Etsalat Unversty Sharjah UNITED ARAB EMIRATES Abstract:

More information

_J _J J J J J J J J _. 7 particles in the blue state; 3 particles in the red state: 720 configurations _J J J _J J J J J J J J _

_J _J J J J J J J J _. 7 particles in the blue state; 3 particles in the red state: 720 configurations _J J J _J J J J J J J J _ Dsrder and Suppse I have 10 partcles that can be n ne f tw states ether the blue state r the red state. Hw many dfferent ways can we arrange thse partcles amng the states? All partcles n the blue state:

More information

Spring 2002 Lecture #17

Spring 2002 Lecture #17 1443-51 Sprng 22 Lecture #17 r. Jaehn Yu 1. Cndtns fr Equlbrum 2. Center f Gravty 3. Elastc Prpertes f Slds Yung s dulus Shear dulus ulk dulus Tday s Hmewrk Assgnment s the Hmewrk #8!!! 2 nd term eam n

More information

Lab 11 LRC Circuits, Damped Forced Harmonic Motion

Lab 11 LRC Circuits, Damped Forced Harmonic Motion Physics 6 ab ab 11 ircuits, Damped Frced Harmnic Mtin What Yu Need T Knw: The Physics OK this is basically a recap f what yu ve dne s far with circuits and circuits. Nw we get t put everything tgether

More information

Is current gain generally significant in FET amplifiers? Why or why not? Substitute each capacitor with a

Is current gain generally significant in FET amplifiers? Why or why not? Substitute each capacitor with a FET Sall Snal Mdband Mdel Ntatn: C arables and quanttes are enerally desnated wth an uppercase subscrpt. AC arables and quanttes are enerally desnated wth a lwercase subscrpt. Phasr ntatn wll be used when

More information

Chem 204A, Fall 2004, Mid-term (II)

Chem 204A, Fall 2004, Mid-term (II) Frst tw letters f yur last name Last ame Frst ame McGll ID Chem 204A, Fall 2004, Md-term (II) Read these nstructns carefully befre yu start tal me: 2 hurs 50 mnutes (6:05 PM 8:55 PM) 1. hs exam has ttal

More information

Approach: (Equilibrium) TD analysis, i.e., conservation eqns., state equations Issues: how to deal with

Approach: (Equilibrium) TD analysis, i.e., conservation eqns., state equations Issues: how to deal with Schl f Aerspace Chemcal D: Mtvatn Prevus D Analyss cnsdered systems where cmpstn f flud was frzen fxed chemcal cmpstn Chemcally eactng Flw but there are numerus stuatns n prpulsn systems where chemcal

More information

MAE140 - Linear Circuits - Winter 16 Final, March 16, 2016

MAE140 - Linear Circuits - Winter 16 Final, March 16, 2016 ME140 - Lnear rcuts - Wnter 16 Fnal, March 16, 2016 Instructons () The exam s open book. You may use your class notes and textbook. You may use a hand calculator wth no communcaton capabltes. () You have

More information

Lead/Lag Compensator Frequency Domain Properties and Design Methods

Lead/Lag Compensator Frequency Domain Properties and Design Methods Lectures 6 and 7 Lead/Lag Cmpensatr Frequency Dmain Prperties and Design Methds Definitin Cnsider the cmpensatr (ie cntrller Fr, it is called a lag cmpensatr s K Fr s, it is called a lead cmpensatr Ntatin

More information

A Note on Equivalences in Measuring Returns to Scale

A Note on Equivalences in Measuring Returns to Scale Internatnal Jurnal f Busness and Ecnmcs, 2013, Vl. 12, N. 1, 85-89 A Nte n Equvalences n Measurng Returns t Scale Valentn Zelenuk Schl f Ecnmcs and Centre fr Effcenc and Prductvt Analss, The Unverst f

More information

(2) Even if such a value of k was possible, the neutrons multiply

(2) Even if such a value of k was possible, the neutrons multiply CHANGE OF REACTOR Nuclear Thery - Curse 227 POWER WTH REACTVTY CHANGE n this lessn, we will cnsider hw neutrn density, neutrn flux and reactr pwer change when the multiplicatin factr, k, r the reactivity,

More information

ENG2410 Digital Design Sequential Circuits: Part A

ENG2410 Digital Design Sequential Circuits: Part A ENG2410 Digital Design Sequential Circuits: Part A Fall 2017 S. Areibi Schl f Engineering University f Guelph Week #6 Tpics Sequential Circuit Definitins Latches Flip-Flps Delays in Sequential Circuits

More information

Physics 4B. Question and 3 tie (clockwise), then 2 and 5 tie (zero), then 4 and 6 tie (counterclockwise) B i. ( T / s) = 1.74 V.

Physics 4B. Question and 3 tie (clockwise), then 2 and 5 tie (zero), then 4 and 6 tie (counterclockwise) B i. ( T / s) = 1.74 V. Physcs 4 Solutons to Chapter 3 HW Chapter 3: Questons:, 4, 1 Problems:, 15, 19, 7, 33, 41, 45, 54, 65 Queston 3-1 and 3 te (clockwse), then and 5 te (zero), then 4 and 6 te (counterclockwse) Queston 3-4

More information

SPH3U1 Lesson 06 Kinematics

SPH3U1 Lesson 06 Kinematics PROJECTILE MOTION LEARNING GOALS Students will: Describe the mtin f an bject thrwn at arbitrary angles thrugh the air. Describe the hrizntal and vertical mtins f a prjectile. Slve prjectile mtin prblems.

More information

Week3, Chapter 4. Position and Displacement. Motion in Two Dimensions. Instantaneous Velocity. Average Velocity

Week3, Chapter 4. Position and Displacement. Motion in Two Dimensions. Instantaneous Velocity. Average Velocity Week3, Chapter 4 Moton n Two Dmensons Lecture Quz A partcle confned to moton along the x axs moves wth constant acceleraton from x =.0 m to x = 8.0 m durng a 1-s tme nterval. The velocty of the partcle

More information

2 Analysis of the non-linear aerodynamic loads of hypersonic flow. 1 General Introduction

2 Analysis of the non-linear aerodynamic loads of hypersonic flow. 1 General Introduction 4 TH INTERNATIONAL CONGRESS OF THE AERONAUTICAL SCIENCES PRELIMINARY STUDY OF NON-LINEAR AEROELASTIC PHENOMENA IN HYPERSONIC FLOW Zhang Wewe, Ye Zhengyn, Yang Yngnan Cllege f Aernautcs, Nrthwestern Plytechncal

More information

Lecture 02 CSE 40547/60547 Computing at the Nanoscale

Lecture 02 CSE 40547/60547 Computing at the Nanoscale PN Junctin Ntes: Lecture 02 CSE 40547/60547 Cmputing at the Nanscale Letʼs start with a (very) shrt review f semi-cnducting materials: - N-type material: Obtained by adding impurity with 5 valence elements

More information

OTHER USES OF THE ICRH COUPL ING CO IL. November 1975

OTHER USES OF THE ICRH COUPL ING CO IL. November 1975 OTHER USES OF THE ICRH COUPL ING CO IL J. C. Sprtt Nvember 1975 -I,," PLP 663 Plasma Studies University f Wiscnsin These PLP Reprts are infrmal and preliminary and as such may cntain errrs nt yet eliminated.

More information

GENERAL FORMULAS FOR FLAT-TOPPED WAVEFORMS. J.e. Sprott. Plasma Studies. University of Wisconsin

GENERAL FORMULAS FOR FLAT-TOPPED WAVEFORMS. J.e. Sprott. Plasma Studies. University of Wisconsin GENERAL FORMULAS FOR FLAT-TOPPED WAVEFORMS J.e. Sprtt PLP 924 September 1984 Plasma Studies University f Wiscnsin These PLP Reprts are infrmal and preliminary and as such may cntain errrs nt yet eliminated.

More information

Relationships Between Frequency, Capacitance, Inductance and Reactance.

Relationships Between Frequency, Capacitance, Inductance and Reactance. P Physics Relatinships between f,, and. Relatinships Between Frequency, apacitance, nductance and Reactance. Purpse: T experimentally verify the relatinships between f, and. The data cllected will lead

More information

Department of Civil Engineering & Applied Mechanics McGill University, Montreal, Quebec Canada

Department of Civil Engineering & Applied Mechanics McGill University, Montreal, Quebec Canada Department f Cvl Engneerng & Appled Mechancs McGll Unversty, Mntreal, Quebec Canada CIVE 90 THEMODYNAMICS & HEAT TANSFE Assgnment #6 SOUTIONS. Cnsder a.-m hgh and -m-wde duble-pane wndw cnsstng f tw 3-mmthck

More information

176 5 t h Fl oo r. 337 P o ly me r Ma te ri al s

176 5 t h Fl oo r. 337 P o ly me r Ma te ri al s A g la di ou s F. L. 462 E l ec tr on ic D ev el op me nt A i ng er A.W.S. 371 C. A. M. A l ex an de r 236 A d mi ni st ra ti on R. H. (M rs ) A n dr ew s P. V. 326 O p ti ca l Tr an sm is si on A p ps

More information

Advanced Circuits Topics - Part 1 by Dr. Colton (Fall 2017)

Advanced Circuits Topics - Part 1 by Dr. Colton (Fall 2017) Advanced rcuts Topcs - Part by Dr. olton (Fall 07) Part : Some thngs you should already know from Physcs 0 and 45 These are all thngs that you should have learned n Physcs 0 and/or 45. Ths secton s organzed

More information

Study Group Report: Plate-fin Heat Exchangers: AEA Technology

Study Group Report: Plate-fin Heat Exchangers: AEA Technology Study Grup Reprt: Plate-fin Heat Exchangers: AEA Technlgy The prblem under study cncerned the apparent discrepancy between a series f experiments using a plate fin heat exchanger and the classical thery

More information

Thermodynamics of Materials

Thermodynamics of Materials Thermdynamcs f Materals 14th Lecture 007. 4. 8 (Mnday) FUGACITY dg = Vd SdT dg = Vd at cnstant T Fr an deal gas dg = (RT/)d = RT dln Ths s true fr deal gases nly, but t wuld be nce t have a smlar frm fr

More information

Transient Effects on High Voltage Diode Stack under Reverse Bias

Transient Effects on High Voltage Diode Stack under Reverse Bias Transent Effects n Hgh ltage de Stack under everse Bas Papež zech Techncal Unversty n Prague, Techncká, Prague 6, B Kjecký, J Kžíšek POLOOIČE, as, Nvdvrská 8a, Prague 4, J Hejhal student zech Techncal

More information

BME 5742 Biosystems Modeling and Control

BME 5742 Biosystems Modeling and Control BME 5742 Bsystems Mdeln and Cntrl Cell Electrcal Actvty: In Mvement acrss Cell Membrane and Membrane Ptental Dr. Zv Rth (FAU) 1 References Hppensteadt-Peskn, Ch. 3 Dr. Rbert Farley s lecture ntes Inc Equlbra

More information