CHAPTER 12 124 (a) Under the quasistatic assumption the carriers and hence the device under analysis are assumed to respond to a time-vaying signal as if it were a dc. bias. In the derivation of the generalized two-port model, one specifically equates the total time-varying terminal currents 4p, ic) to the d.c. currents that \iuld exist under equivalent biasing conditions. (b) Two separate definitions are necessary because, contrary 10 the polarities assumed in the development of the generalized small-signal model, the Jp and /¢ currents were previously taken to be positive flowing ous of the base and the collector terminals in a pap BIT, (As noted in Section 10.1, the direction of positive current was so chosen to avoid unnecessary complications, serious sign-related difficulties, in the physical description of current flow inside the BJT when operated in the standard amplifying mode.) {c} The Hybrid-Pi model gets its name from the %-like arrangement of circuit elements with “hybrid” (a combination of conductance and resistance) units. (d) Names (see the first paragraph in Subsection 12.12). &m--transconductance p...output resistance iput Tesistance feedthrough resistance (e) The capacitors model the collector-base arid emitter-base pn junction capacitances which cannot be neglected at higher frequencies, (f) The minority carrier concentration in the base continues to increase as pictured in plot (iii) of Fig. 12.4(¢) until a maximum build-up consistent with the applied biases is auained. The base current varies as Qp/tg and therefore also continues to inciease toward a saturating maximum Value. (In the quantitative analysis, ip increases from [oct at the start of Saturation to a saturating value of /ggtp.) Once saturation biased, ic remains essentially constant at i¢= foc = VeiRL. (g) In words, the base transit time is the average time taken by minority carriers to diffuse across the quasineutral base. Mathematically (see Eq. 12.22), = W2/2Dp. () Bac = Iola = ph G) An ip < 0 aids the widthdrawal of stored charge from the quasineutral base, which in tum reduces both the storage delay time and the fall time. @) A Schouky diode clamp is a circuit arrangement where a Schottky diode is connected between the collector and base of a BST as pictured in Fig. 12.7(a). The Schottky diode conducts at a lower forward bias than a pa junction and therefore minimizes the forward (saturation-mode) bias that is applied to the BJT under turn;on conditions. ‘This reduces the stored charge and speeds up the twm-off transient. (Also see Subsection 12.2.4.) 12-122 The BIT viewed as a qwo-port network and connected in the common-base configuration is pictured below. : ie= [ete ic aaa —_ f——* in > Vents BIT Vent va = out a — invoking the quasistatic assumption we can write ig(Vept+vep, Veptucy) = le(Vep+veb,VeBt teh) = Je(Ves.Vcp) + le icVeptver.Vepttep) = Ic(Veptver,Vcsttep) = IcVep.Vcg) + ic or fe = Inept Pep, Vcp+ te) ~fe(Vep,Vcp) 8 1XVEp+ ep. Von tcp) ~Ic(Ves.Vee) fc Next performing a Taylor series expansion of the first term on the right-hand side of the above equations, and keeping only first order terms, we obiain ah oh Ie(Vept Ven. Vew* ve) = Je(Ven.Vep) +=} vp+ EY nn Weslvce eV cnlven ak ar Ic{Ven+¥ep,Ven+v) = Ie(VenVen) +] vege SE} ay WeBivcs — WVealven which when substituted into the preceding equations gives an an = EY ryt SE} uy OVEBIVcR WVealven ke - Bel wie} ay sslven “Wee Ven 12-2If the direction of positive current flow is as defined in Fig. 10.2 (4/e out and +/¢ in for an npn BIT, 41g in and +c out for a pap BY), then introducing, ale aie ou = = : gn = Wenlicy Veelype Werlyen Vac t t a ft npn Pp npr pnp sein Ic out agin cout yield the eminer and collector a.c. current node equations ig = ByyVet + 812d io = 821 Ved + 822% ‘The low-frequency small-signal equivalent circuit characterizing the a.c. response of the BJT connected in the common base configuration is therefore concluded to be BS tReset Voltages if i=J, VEB=VEBO-0.0001; else; end ie 7 VEB=VEBO+0.0001; else: end }, VEB=VEBO; VEC= VECO-0.01; else: end if + VEC=VECC+0.01; else: end iBS[iB, IB}; iC-[ic, IC}: eng : 12-6Compute Generalized To-Port Model Parameters gil (58(3) -1B(2)) /0,0602 gl2=(1B (5) “4B (4)) 70.02: g2h=(4C (3) 4 (2}} 0.0002; G22=(1C(5} AC (4) ) 70.92 fprintf(*\nflybrid-pi Model Parameters\n') gn-g21-g12 A€ q22+q12—=0 — ro=inf else ro=i/ (g22+g1 25 end xpiel/(glitg12) if gl2—=0, cmu-int else, mm="1/gh2 end fprint{(*\ngu and rpi computed using Eqs. (12.9)\n") gm=3C (2) /0.0259 rpis0.0259/1B (1) 12.6 (@) The high-frequency equivalent circuit of Fig. 12.2(c) with tee = 0 can be manipulased into the form where Y= b+ joc +j0Ce, Th , 12-7Combining node and loop analysis we note ip = Vi0pe' = Yovgs" ay fc = Smdpe + YrV¢H' + Veelro Q ite + Vee’ + (iptidre = 0 @) Dye! — Vee’ + Vey’ = 0 @) Eq. (4) is used to eliminate vey’ in Eqs. (1) and (2). Eqs. (1) and (2) are then combined to eliminate Dye. Next Eqs. (3) is used to eliminate Yee’. Finally, the i/i, ration is formed giving 1 {8m = Yo’ _ ie {io+Bpe+ ¥, +; Je. y iy [¥o— 8m , Pelee +ro¥a-(to+Bfretra -1 La (b) With re = 0 as given in the list of parameters, the io/ip ratio simplifies to (ee feo AW ta) iy (¥2- 8m 4 Lye f=te).y, (ro+ phe <1 Using the MATLAB program to compute lig/ip! versus frequency, one desermines an [fr = 235 MHz]. Data sheets list the ff of the 2N3906 pnp BIT to be approximately 200 MHz. (It should be noted that the Electronics Workbench software program was used to determine the d.c. operating point that produced an Ic= 1 mA. The series resistances listed in the problem statement were those quoted by the EW program. Zero-bias capacitance values employed in computing the Hybrid-Pi parameters were also extracted from the Electronics Workbench program.) A plot of lic/ig! versus frequency, and the MATLAB m-file constructed to generate the plot and determine fy. are reproduced on the next page. 12-8042) MATLAR program script... wProblem 12.6...{7 determination sTnitialization clear; close , Parameters ’lic/ibl vs. frequency gm=3.862-2; egspace (4, 9, 200): xpi=4.65e3; we2.tph tf: r0=2. 0004; Yi-l/rpit3) mmu=3. 59e6; ¥2=1/xmutj.*w.*Cob: Ceb=23.6e-127 ¥2—am) . / (1L4¥2) Ceb=2. 320-12: Den=R.*rc.*¥2 — (¥241/ro).*re ~ 1: rb=10; betas2bs(R./Den); bbeta=lic/ib! ro=2.8; splot re=0; loglog(f,beta): grid mlabel("£ (Hz) "); ylabel(*{ ic / ib 421 ‘The Eqs. (6.68)/(6.69) solution for the /p=r flowing in a narrow base diode is ga De, Bi cosh IEP) (—q¥ ltt _ 1) IDIFR = Lp Np sinh(xe/Lp) For application to a BIT we make the symbol replacements...Dp > Da, Lp > Lp, Np > Np, Xe! -> W, and Va ~> Veg. ‘Then 2 cosh(WiLs) {, oVenuer q = gh Ba Fi cosh(WILB) (p gVeBhT _ Oe Lg Np sinhWikp) ) Since W/Lp << | in a standard transistor, the cosh/sinh factor can be expanded as noted in the problem statement 10 obtain cosh(Wit) al L ail cosh(WIEp) _ Lal; - WiLp << sinh(W/Lg) * 3 ih Me and 2 Dam, I 12°F eavenne u A Bl 4 L/W (eavewer_ 4 DIFF (. W Np 3 lbp it ) Introducing the substitutions cited in Subsection 7.3.2, that is, wp. Ww? 1+ jor, ae ~ Data = pt ote) ” eavente 1) => @veykre EBT yields the corresponding a.c. relationship n2 2 (ca Pezt}hs 4 Wj) Wl 3Dam 3p Finally, ty definition, (22 Je QV egikT iaitt = Gp+jOCp = igi Yen and therefore 2 2 Dp Mm J coven = 3 a [on Dg hh )eovener SAT iq W Np12.8 | The pictured “on” point in Fig. 12.3(b) lies right on the /g = Vs/Rs line. Therefore | fap =Vs/Rs = 30u4, i | Inspecting the plot we find [eo = Vao/Ry = 5.0 mA, We knew fic = /c/In = tp/4, Although base width modulation clearly causes file to vary somewhat depending on the d.c. operating point, itis reasonable to employ a median val in obtaining the desired estimate. Specifically, using the point where the foad line crosses the fa = 15 [UA characteristic, we obeain ~. (0.6245 x 103) te fe - O62 ~ 208 fp 18 15 X 106 ‘Thus, fect — GX103) Tnpt G0 10°)(208) 12-11fab) ‘The required plots and the generating MATLAB m-file are reproduced below. The compute tional relationships used in producing the plots were by TB Gh] ta wl) if E~0 1 tp 2 sere oa) ww ifEet where, 2 sIecw/lppt® of 02 03 af 05 06 O07 08 09 1 {0C taut / 188 tauBOi 02 03 04 a5 06 OF 08 100 taut /1BB ta0B MATLAB program script. @Rise and Storage-Delay Time plots (Prob. 12.9) ‘Initialization clear; close $Rise time computation x=Linspace (0.01,0.99); rise-log(1./(1-x)}; ris plot (%,rise}; grid xlabel (1ICC taut / EBB tauB'); ylabel (‘tr / tauB‘) pause iStorage-Delay Time computation delayO=log (1. /x); ‘$delay0=tsd/tauk, xi-0 delayl=l0g(2./(1+x)); Sdelayl=tsd/tauB, xi=1 plot (x,delay0,x.delayl); grid xlabel (ICC taut / TBR tauB'); ylabel('tsd / tavB") text (0.08,2.8,"xi=0"); text (0.08,0.8, "xi=1") x/taub 12-13 os219 {a) Let ty be the time when ic = 0.9Icc and £2 the ime when i¢ = 0.Uec. Making use of Fg, (12.31b), we can then write iefa) = OSice = Jaw Barden’ — 3) ila) = C.tlcc = Iop® [14H - g i Solving for the ¢'s yields of 1+é re 0.9 ccvingte + §) 14 Q = % ly (0. ecty/ppts + &) and per the measurements-based definition [oStectippt + 4) < mh fo +E f= tty = TlH \otlectlasts + & Ox € where x=Ject//ppt (b) With € = O and &= {m9 wif E=0 eer ea xed , the part (2) relationship simplifies tc ‘The requested 1/73 versus x plot is displayed on the next page along with the script of the MATLAB m-file used to generate the plot. Consistent with the analysis in Subsection 12.2.3, the plotted fall times decrease when E>0. This occurs because an ig < Oaids the withdrawal of charge from the quasincural base. If the x-ratio increases eitht duc to an increase in [oc or a decrease in Ipp, the charge storage is enhanced relative to the charge removal capability of the base 12-14current. Thus, the «4/1 ratio for the £ = 1 curve incicases with increasing, x. When € = the charge removal from the base occurs oaly by secombination and the fall-time collector’ current assumes the simple form, ic = Aexp(i/mp). Since tris always evaluated employing the same relative ic values, io(n Wic(2) = constant = exp(te/tp), and ¢4/zp is seen to be a constant independent of x. 25, us| beeen endo tytaus ot 02 03 04 05 08 67 08 Oo ¢ ICC taut 198 t2uB MATLAB program script... ¥all Time (Problem 12.10) ‘Initialization clear; close a¥all Time computations x0=[0, 1); y0=(1og (9) ,10g(9} }: St€/tauB when xi=0 xdeLinspace (0,1); Yielog((0.9.*x1+1)./(0.1.*x1+41)); $t£/tauB when xi=1 PlOL(KO,yO,x1.yl)2 grid xlabel ("UCC taut / BB taut); ylabel (‘tf / tav8") text (0.47,2.1,'xie0'); text (0.47,0-4,'xie1"y 7 12-15