D ERIVATIONS

OF

S ELECTED

Appendix

E QUATIONS

Equation 23
The average value of a half-wave rectified sine wave is the area under the curve divided by the period (2p). The equation for a sine wave is v = Vpsin u VAVG =
p Vp area 1 = Vpsin u du = ( - cos u)|p 0 2p 2p L 2p 0 Vp Vp Vp = 3 - cos p - ( - cos 0)4 = 3 - ( - 1) - ( - 1)4 = (2) 2p 2p 2p Vp = p

VAVG

Equation 212
Refer to Figure B1.

vC = Vp(rect)e Vp (rect)

t RLC

Vr ( pp)

tdis T

FIGURE B1

When the filter capacitor discharges through RL, the voltage is vC = Vp(rect)e - t>RLC

Since the discharge time of the capacitor is from one peak to approximately the next peak, tdis T when vC reaches its minimum value. Since RC W T, T/RLC becomes much less than 1 (which is usually the case); e - T>RLC approaches 1 and can be expressed as e - T>RLC 1 T RLC vC(min) = Vp(rect)e - T>RLC

B-2

A PPENDIX B

Therefore, vC(min) = Vp(rect) a 1 The peak-to-peak ripple voltage is Vr ( pp) = Vp(rect) - VC(min) = Vp(rect) - Vp(rect) + Vr ( pp) a 1 bV fRLC p(rect) Vp(rect)T RLC = Vp(rect)T RLC T b RLC

Equation 213
To obtain the dc value, one-half of the peak-to-peak ripple is subtracted from the peak value. VDC = Vp(rect) VDC = a 1 Vr (pp) 2 = Vp(rect) - a 1 bV 2fRLC p(rect)

1 bV 2fRLC p(rect)

Equation 61
The Shockley equation for the base-emitter pn junction is IE = IR(eVQ>kT - 1)

where IE IR V Q k T

= = = = = =

At ambient temperature, Q> kT 40, so Differentiating yields

the total forward current across the base-emitter junction the reverse saturation current the voltage across the depletion layer the charge on an electron a number known as Boltzmanns constant the absolute temperature IE = IR(eV40 - 1) dIE = 40IReV40 dV

Since IReV40 = IE + IR, dIE = 40(IE + IR) dV Assuming IR 6 6 IE, dIE 40IE dV

The ac resistance r e of the base-emitter junction can be expressed as dV> dIE. re = dV 1 25 mV dIE 40IE IE

Equation 614
The emitter-follower is represented by the r parameter ac equivalent circuit in Figure B2(a).

D ERIVATIONS

OF

S ELECTED E QUATIONS

B-3

C R1 B Rs ac Ib Rs || R1 || R2 ac Ib

r e E

r e

Vs

R2

RE

Ie

RE

Ve = Vout

(a)

(b)

FIGURE B2

Conventional current direction shown.

By thevenizing from the base back to the source, the circuit is simplified to the form shown in Figure B2(b). Vout = Ve, Iout = Ie, and Iin = Ib. Rout = Ve Ie Ie b acIb

With Vs = 0 and with Ib produced by Vout, and neglecting the base-to-emitter voltage drop (and therefore r e ), Ib Assuming that R1 7 7 Rs and R2 7 7 Rs, Ib Ve Rs Ve R1 || R2 || Rs

Iout = Ie = Vout Iout

Looking into the emitter, RE appears in parallel with Rs> b ac. Therefore, Rout = a Rs b || RE b ac

b acVe Rs Ve Ve Rs = = = Ie b acVe> Rs b ac

Midpoint Bias (Chapter 8)

The following proof is for the equation on page 400 that shows ID 0.5IDSS when VGS = VGS(off )/3.4. Start with Equation 81: ID IDSS a 1 Let ID = 0.5IDSS. 0.5IDSS = IDSS a 1 VGS VGS(off) b
2

VGS VGS(off)

B-4

A PPENDIX B

Cancelling IDSS on each side, 0.5 = a 1 VGS VGS(off) b

2

We want a factor (call it F) by which VGS(off) can be divided to give a value of VGS that will produce a drain current that is 0.5IDSS. 0.5 = J 1 a a VGS(off) F VGS(off) b
2

Solving for F,

VGS(off) F VGS(off)

10.5 = 1 -

= 1 1 F

1 F

10.5 - 1 = -

1 = 1 - 10.5 F 1 F = 3.4 1 - 10.5 Therefore, ID 0.5IDSS when VGS = VGS(off)/3.4.

Equation 92
ID IDSS a 1 = IDSS a 1 IDRS 2 IDRS IDRS b = IDSS a 1 b a1 b VGS(off) VGS(off) VGS(off)
2 2IDRS I2 2IDSSRS IDSSR2 DRS S 2 + 2 b = IDSS ID + 2 ID VGS(off) VGS(off) VGS(off) VGS(off)

Rearranging into a standard quadratic equation form, a IDSSR2 S V2 GS(off) b I2 D - a1 + 2IDSSRS b I + IDSS = 0 VGS(off) D

The coefficients and constant are A = R2 SIDSS V2 GS(off) 2RSIDSS b VGS(off)

B = - a1 + C = IDSS In simplified notation, the equation is

AI2 D + BID + C = 0 The solutions to this quadratic equation are ID = - B ; 2B2 - 4AC 2A

Equation 910
A general model of a switched-capacitor circuit, as shown in Figure B3(a), consists of a capacitor, two voltage sources, V1 and V2, and a two-pole switch. Lets examine this circuit

D ERIVATIONS

OF

S ELECTED E QUATIONS

B-5

0 T /2 T I1 V1 1 2 C V2 0 (a)

Position 1 Position 2 T /2 T

Position 1 Position 2

Position 1

(b)

FIGURE B3

for a specified period of time, T. Assume that V1 and V2 are constant during the time period T and V1 7 V2. Of particular interest is the average current I1 produced by the source V1 during the time period T. During the first half of the time period T, the switch is in position 1, as indicated in Figure B3(b). The capacitor charges very rapidly to the source voltage V1. Therefore, an average current I1 due to V1 is charging the capacitor during the interval from t = 0 to t = T/2. During the second half of the time period, the switch is in position 2, as indicated. Because V1 7 V2, the capacitor rapidly discharges to the voltage V2. The average current produced by the source V1 over the time period T is I1(avg) = Q1(T/2) - Q1(0) T

Q1(0) is the charge at t = 0 and Q1(T/2) is the charge at t = T/2. Therefore, Q1(T/2) - Q1(0) is the net charge transferred while the switch is in position 1. The capacitor voltage at T/2 is equal to V1, and the capacitor voltage at 0 or T is equal to V2. By substituting CV for Q in the previous equation, I1(avg) = CV1(T/2) - CV2(0) T = C1V1(T/2) - V2(0)2 T

Since V1 and V2 are assumed to be constant during T, the average current can be expressed as I1(avg) = C(V1 - V2) T
I1 V1 R V2

Figure B4 shows a conventional resistive circuit with two voltage sources. From Ohms law, the current is I1 = V1 - V2 R

The current I1(avg) in the switched-capacitor circuit is equal to I1 in the resistive circuit. I1(avg) = C(V1 - V2) V1 - V2 = T R
F I G U R E B 4

By solving for R and canceling the V1 - V2 terms, R = T(V1 - V2) C(V1 - V2) T R = C

As you can see, a switched-capacitor circuit can emulate a resistor with a value determined by the time period T and the capacitance C. Remember that the two-pole switch is in each position for one-half of the time period T and that you can vary T by varying the frequency at which the switches are operated.

B-6

A PPENDIX B

Since T = 1/f, the resistance in terms of frequency is R = 1 fC

Equation 101
An inverting amplifier with feedback capacitance is shown in Figure B5. For the input, I1 = Factoring V1 out, I1 = V1 - V2 XC

V1(1 - V2 > V1) XC

The ratio V2> V1 is the voltage gain, - Av. I1 =

V1(1 + Av) V1 = XC XC > (1 + Av)

C

FIGURE B5

I1 V1 Av

I2 V2

The effective reactance as seen from the input terminals is XCin(Miller) = or 1 1 = 2pfCin(Miller) 2pfC(1 + Av) Cancelling and inverting, Cin(Miller) = C(Av + 1) XC 1 + Av

Equation 102
For the output in Figure B6, I2 = Since V1> V2 = - 1> Av, I2 = V2(1 - V1> V2) V2 - V1 = XC XC V2 V2 = XC> (1 + 1> Av) XC> [(Av + 1)> Av] XC (Av + 1)> Av 1 = 2pfC[(Av + 1)> Av]

V2(1 + 1> Av) XC

The effective reactance as seen from the output is XCout(Miller) = 1 2pfCout(Miller)

D ERIVATIONS

OF

S ELECTED E QUATIONS

B-7

Cancelling and inverting yields Cout(Miller) = C a Av + 1 b Av

Equations 1029 and 1030

The total gain, Av(tot), of an individual amplifier stage at the lower critical frequency equals the midrange gain, Av(mid ), times the attenuation of the high-pass RC circuit. Av(tot) = Av(mid ) a fcl = 1 2pRC R 2R +
2

X2 C

b = Av(mid ) a

2 21 + X2 C> R

Dividing both sides by any frequency f, fcl 1 = f (2pfC )R Since XC = 1 > 2pfC, fcl XC = f R Substitution in the gain formula gives Av(tot) = Av(mid) a The gain ratio is Av(tot) Av(mid ) = 21 + ( fcl> f )2 1 21 + ( fcl> f )2 1 b

For a multistage amplifier with n stages, each with the same fcl and gain ratio, the product of the gain ratios is a 21 + ( fcl> f )2 1 b
n

The critical frequency f cl of the multistage amplifier is the frequency at which Av(tot) = 0.707Av(mid), so the gain ratio at f cl is Av(tot) Av(mid ) = 0.707 = 1 1 = 1.414 12

Therefore, for a multistage amplifier,

n 1 1 1 = c d = 12 21 + ( fcl> f c l)2 ( 21 + ( fcl> f c l)2)n

So 21/2 = ( 21 + ( fcl> f cl )2)n Squaring both sides, 2 = (1 + ( fcl> f cl )2)n

B-8

A PPENDIX B

Taking the nth root of both sides, a

fcl 2 b = 21>n - 1 f cl fcl a b = 221>n - 1 f cl fcl f cl = 1>n 22 - 1 f cu = fcu 221>n - 1

21>n = 1 + ( fcl> f cl )2

A similar process will give Equation 1030:

Equations 1031 and 1032

The rise time is defined as the time required for the voltage to increase from 10 percent of its final value to 90 percent of its final value, as indicated in Figure B6. Expressing the curve in its exponential form gives v = Vfinal(1 - e - t>RC)

When v = 0.1Vfinal,

0.1Vfinal = Vfinal(1 - e - t>RC) = Vfinal - Vfinale - t>RC Vfinale - t>RC = 0.9Vfinal ln e - t>RC = ln (0.9) t = - 0.1 RC t = 0.1RC e - t>RC = 0.9

FIGURE B6 Vfinal 0.9 Vfinal

Vfinal 1 e

t RC

0.1 Vfinal 0 tr t

When v = 0.9Vfinal,

0.9Vfinal = Vfinal(1 - e - t>RC) = Vfinal - Vfinale - t>RC Vfinale - t>RC = 0.1Vfinal ln e - t>RC = ln (0.1) t = - 2.3 RC t = 2.3RC

D ERIVATIONS

OF

S ELECTED E QUATIONS

B-9

The difference is the rise time. tr = 2.3RC - 0.1RC = 2.2RC The critical frequency of an RC circuit is 1 2pRC 1 RC = 2pfc fc = Substituting, tr = fcu In a similar way, it can be shown that fcl = 0.35 tf 2.2 0.35 = 2pfcu fcu 0.35 = tr

Equation 1221
The formula for open-loop gain in Equation 1219 can be expressed in complex notation as Aol = 1 + jf> fc(ol ) Aol(mid )

Substituting the above expression into the equation Acl = Aol> (1 + BAol) gives a formula for the total closed-loop gain. Acl = 1 + BAol(mid )> (1 + jf> fc(ol )) Aol(mid ) Aol(mid )> (1 + jf> fc(ol ))

Multiplying the numerator and denominator by 1 + jf> fc(ol ) yields Acl = 1 + BAol(mid ) + jf> fc(ol ) Aol(mid )> (1 + BAol(mid ))

Dividing the numerator and denominator by 1 + BAol(mid ) gives Acl = 1 + j [ f> ( fc(ol )(1 + BAol(mid )))] Acl(mid )

The above expression is of the form of the first equation Acl = 1 + jf> fc(cl )

where fc(cl) is the closed-loop critical frequency. Thus, fc(cl ) = fc(ol )(1 + BAol(mid ))

Equation 141
In Figure B7 the common-mode voltage, Vcm, on the noninverting input is amplified by the small common-mode gain of op-amp A1. (Acm is typically less than 1.) The total output voltage of op-amp A1 is Vout1 = a 1 + R1 R1 b V - a b Vin2 + Vcm RG in1 RG

B-10

A PPENDIX B

A similar analysis can be applied to op-amp A2 and results in the following output expression: Vout2 = a1 + R2 R2 b V - a b Vin1 + Vcm RG in2 RG

Vin1 + Vcm

+ A1 R1 R2 A2

Vout1 R3

R5

A3 + R4 Vout 2 R6 Vout = Acl (Vin2 Vin1)

RG

Vin2 + Vcm

FIGURE B7

Op-amp A3 has Vout1 on one of its inputs and Vout2 on the other. Therefore, the differential input voltage to op-amp A3 is Vout2 - Vout1. Vout2 - Vout1 = a 1 + For R1 = R2 = R, Vout2 - Vout1 = a 1 + 2R 2R bV - a1 + b V + Vcm - Vcm RG in2 RG in1 R2 R1 R2 R1 + b Vin2 - a 1 + + b V + Vcm - Vcm RG RG RG RG in1

Notice that, since the common-mode voltages (Vcm) are equal, they cancel each other. Factoring out the differential gain gives the following expression for the differential input to op-amp A3: Vout2 - Vout1 = a 1 + 2R b (Vin2 - Vin1) RG

Op-amp A3 has unity gain because R3 = R5 = R4 = R6 and Av = R5/R3. = R6/R4. Therefore, the final output of the instrumentation amplifier (the output of op-amp A3) is Vout = 1(Vout2 - Vout1) = a 1 + The closed-loop gain is Vout Vin2 - Vin1 2R Acl = 1 + RG Acl = 2R b (Vin2 - Vin1) RG

Equation 161
R( - jX )> (R - jX ) Vout R( - jX ) = = Vin (R - jX ) + R( - jX )> (R - jX ) (R - jX )2 - jRX

D ERIVATIONS

OF

S ELECTED E QUATIONS

B-11

Multiplying the numerator and denominator by j, Vout RX RX = = Vin j(R - jX)2 + RX RX + j(R2 - j2RX - X2) RX RX = = 2 2 RX + jR + 2RX - jX 3RX + j(R2 - X2) For a 0 phase angle there can be no j term. Recall from complex numbers in ac theory that a nonzero angle is associated with a complex number having a j term. Therefore, at fr the j term is 0. R2 - X2 = 0 Thus, Vout RX = Vin 3RX Cancelling yields Vout 1 = Vin 3

Equation 162
From the derivation of Equation 161, R2 - X2 = 0 R2 = X2 R = X Since X = 1 , 2pfrC R = 1 2pfrC 1 fr = 2pRC

Equations 163 and 164

The feedback circuit in the phase-shift oscillator consists of three RC stages, as shown in Figure B8. An expression for the attenuation is derived using the mesh analysis method for the loop assignment shown. All Rs are equal in value, and all Cs are equal in value. (R - j1> 2pfC )I1 - RI2 + 0I3 = Vin - RI1 + (2R - j1> 2pfC )I2 - RI3 = 0 0I1 - RI2 + (2R - j1> 2pfC )I3 = 0
C C C

Vin

I1

I2

I3

Vout

FIGURE B8

B-12

A PPENDIX B

In order to get Vout, we must solve for I3 using determinants: 3 I3 = 3 (R - j1> 2pfC ) -R 0 -R (2R - j1> 2pfC ) -R -R (2R - j1> 2pfC ) -R Vin 0 0 3

(R - j1> 2pfC ) -R 0

0 3 -R (2R - j1> 2pfC )

I3 =

(R - j1> 2pfC )(2R - j1> 2pfC )2 - R2(2R - j1> 2pfC ) - R2(R - 1> 2pfC )

R2Vin

Vout RI3 = Vin Vin = R3 (R - j1> 2pfC )(2R - j1> 2pfC )2 - R3(2 - j1> 2pfRC ) - R3(1 - 1> 2pfRC )

R3 R3(1 - j1> 2pfRC )(2 - j1> 2pfRC )2 - R3[(2 - j1> 2pfRC ) - (1 - j1> 2pfRC )] R3 R3(1 - j1> 2pfRC ) (2 - j1> 2pfRC )2 - R3(3 - j1> 2pfRC )

Vout 1 = Vin (1 - j1> 2pfRC ) (2 - j1> 2pfRC )2 - (3 - j1> 2pfRC ) Expanding and combining the real terms and the j terms separately. Vout = Vin 1 5 6 1 a1 b - ja b 2 2 2 2 2pfRC 4p f R C (2pf )3R3C 3

For oscillation in the phase-shift amplifier, the phase shift through the RC circuit must equal 180. For this condition to exist, the j term must be 0 at the frequency of oscillation fr. 6 1 = 0 2pfrRC (2pfr)3R3C3 6(2p)2fr2R2C2 - 1 (2p)3fr3R3C3 = 0

6(2p)2fr2R2C2 - 1 = 0 fr2 = 1 6(2p)2R2C2 1 fr = 2p 16RC

Since the j term is 0, Vout = Vin 1 5 1 2 2 2 2 4p fr R C = 1 a 1 5

2 1 b R2C2 16RC

1 1 = 1 - 30 29

The negative sign results from the 180 inversion. Thus, the value of attenuation for the feedback circuit is B = 1 29