Ch 1.

Amplifiers
Ideal OP Amps Basic OP Amp Circuit Blocks Analog Computation Nonlinear OP Amp Applications OP Amp Considerations Guarding Passive Filters Active Filters VCO(Voltage Controlled Oscillator)

Function of Amplifiers


Amplifiers provides
 GAIN  Filtering,

Signal processing, Correction for Nonlinearities

Temperature Pressure Flow Motion .

Sensor

Signal Conditioning Circuitry

Digital Computer

Chap 0

Ideal OP Amps


Transfer Function = Output / Input

 Voltage

Amp TF (Gain):

Av !

vo vi

Av u 1  OP Amp is preferred
 Usually

Easy to use in circuit designed compared to discrete Transistor circuits

Chap 0

Ideal OP Amps (Cont.)



Assumptions
 Open

loop Gain = Infinity  Input Impedance Rd = Infinity  Output Impedance Ro = 0  Bandwidth = Infinity
Infinite Frequency Response
 vo=0

when v1 = v2

No Offset Voltage

Chap 0

Ideal OP Amps (Cont.)



Note


v0 = A(v2 v1)
If v0 = g, A = g (Typically 100,000)
Then v2 v1 = 0 v2 = v1

Since v2 = v1 and Rd = g
We can neglect the current in Rd


Rule 1


When the OP Amp is in linear range the two inputs are at the same voltage No Current flows into either terminal of the OP Amp

Rule 2


Chap 0

Basic OP Amp Circuit Blocks

     

Inverting Amplifier Noninverting Amplifier Unity-Gain Amplifier Differential Amplifier Instrumental Amplifier The Electrocardiogram Amplifier

Chap 0

Inverting Amplifier


Inverting Amp with Gain = - Rf / Ri

From Rule 1
 v-

= v+ = 0

From Rule 2 & KCL

+ if = 0 ii = -if  From Ohms law
 ii

ii = vi / Ri , , if = vo / Rf
 vi


/ Ri = - vo / Rf
vo / vi = -Rf / Ri

Virtual Ground

Inverting Amp Gain



-Rf / Ri
7

Chap 0

Inverting Amplifier (Cont.)



Linear Range


Input Impedance
Low (Ri)  Increasing Ri Decreasing Gain


By Power Supply Voltage

Increasing Gain by increasing Rf

But there is practical limit

Saturation
Chap 0 8

Why High Input Impedance ?



Concept of Loading


: Sensor

Sensor amplitude . Frequency

Open Loop Output



Vx Vy = Vx Vx v Rx / (RL + Rx)

Voltage Drop by Load



Digital


Let RL >> Rx
Vy = Vx  Amp


Rx x Vx Vy RL

Chap 0

Noninverting Amplifiers


Noninverting Amp


By Rule 2
Vo = If v (Rf + Ri)  Vi = If v Ri  Vo = Vi v (Rf + Ri)/Ri


Gain = (Rf + Ri) / Rf

  

Gain: Vo/Vi = 1 + Rf / Ri Gain u 1, Always Input Impedance



Very Large (Infinite)

By Rule 1 Vi

Chap 0

10

UnityUnity-Gain Amplifier


Homework #2-1


 

Verify that the Gain of Unity-Gain Amp is 1

Vo = Vi Applications


Buffer amplifier
Isolate one circuit from the loading effects of a following stage

Impedance converter
Data conversion System (ADC or DAC) where constant impedance or high impedance is required

Chap 0

11

Differential Amplifiers
  

Combination of Inverting and Noninverting Amp Can reject 60Hz interference Electrocardiogram amplifier

Chap 0

Differential Instrumentation

Noninverting
12

Differential Amplifiers (Cont.)



Gain of Differential Amp



By Rule 2
V5 = I2 * R2 V2 = I2 * R1 + V5 = V5 * R1 /R2 + V5 V5 = R2 * V2 / (R1 + R2)

By Rule 1
V1 = R1 * I1 + V5 V5 = R2 * I1 + V6 V6 = (V2 V1) * R2 / R1

Chap 0

13

Differential Amplifiers (Cont.)



CMV (Common Mode Voltage)



If V1 = V2, then V6 = 0

 

CMG (Common Mode Gain) = 0 DG(Differential voltage Gain)



If V1 { V2, then V6 = (V2-V1)*(R2/R1)

 

In practice, CMG { 0 CMRR (Common Mode Rejection Ratio)

Measure of the ability to reject CMV  CMRR = DG / CMG


The Higher CMRR, the better quality Typically, 100 ~ 10,000 60Hz noise common to V1 and V2 can be rejected
Chap 0 14

Instrumentation Amplifiers


One OP Amp Differential Amplifier

 Input

Impedance is not so High

Good for Low impedance source

Strain gage Bridge

Bad for High impedance source



Instrumentation Amplifier
 Differential

Amp with High Input Impedance and Low Output Impedance  Two Noninvering Amp + One Differential Amp

Chap 0

15

Instrumentation Amplifiers (Cont.)



Instrumentation Amp = Noninverting Amp + Differential Amp

Homework #2-2
 Show

that

DG = (V1-V2) / (V3-V4) = (2*R4 + R3) / R3 V6 = (V3-V4)*DG*R2 / R1



First Stage CMRR

 CMRR

= DG / CMG = DG

Overall CMG = 0
 High

CMRR

 

High Input Impedance Gain is adjustable by changing R3

16

Chap 0

The Electrocardiogram Amplifier

Low Pass Filter < 100Hz < 0.2 V Gain = 40

Gain = 32 Maximize CMRR

Chap 0

High Pass Filter >0.05Hz

17

Analog Computation


Digital Signal Processing is preferred

 Flexibility  Easy

to Change  Elimination of hardware



Analog Signal Processing

 Is

preferred when DSP consumes too much time

Chap 0

18

Inverter and Scale Changer



Inverting Amp with Gain = - Rf / Ri

Inverter


Rf / Ri = 1

Inverter and Scale Changer



Proper choice of Rf / Ri Use of inverter to scale the output of DAC

Application


Chap 0

19

Adders (Summing Amplifiers)



Adder


Inverter with Several inputs

Vo = -Rf(V1/R1 + V2/R2 + + Vn/Rn)

If = I1 + I2 + In  I1 = V1/R1,  Vo = -If * Rf


Rf determines overall Gain Ri determines weighting factor and input impedance

Chap 0

20

Integrator


Homework #2-3


Drawbacks


Show that 1 t1 v0 ! 0 vi dt  vic RC

Vo will reach saturation voltage, if Vi is left connected indefinitely

Integrator operates as an open-loop amplifier for DC inputs

Chap 0

21

Practical Integrator

Reset
 S1

Closed, S0 Open
Inverter C is initialized to Vr

Controlled By Relay or Solid State Switch or Analog Switch

Integrate
 S1

Open, S0 Closed

Hold
 S1

Open, S0 Open  Keeps Vo constant

Read and Process
22

Chap 0

Differentiators


Homework #2-4


Drawbacks


Show that dvi v0 !  RC dt



Instability at High frequencies To Stable R A0[ 0 C

Practical Differentiator


Ri !

Chap 0

23

Comparators


Compare Two Inputs



Drawbacks


Vi > Vr
Vo = -Vs

If Vi = Vr + small noise
Rapid fluctuation between s Vs

Vi < Vr
Vo = Vs

Chap 0

24

Comparators with Hysteresis



Positive Feedback
Hysteresis loop  Can remove the effect of Small Noise


Homework #2-5


Show that

Reduce Fluctuation

(VS  Vr ) R1 R1  R2 (VS  Vr ) R1 Vr  (Vr ! Vr  R1  R2 Vr  (Vr ! Vr 

Chap 0

25

Rectifiers


Precision Half Wave Rectifier

Precision Full Wave Rectifier Limiters

Chap 0

26

OP Amp Considerations


Effects of Nonlinear characteristics



Compensation
Undesirable Oscillation at High frequency
Add external Capacitance according to Spec sheet

GBW (Gain Bandwidth Product)

Gain v Bandwidth = Constant (Typically 1MHz)
For Noninverting Amp: Bandwidth = GBW / Gain

Input Offset Voltage

Practical OP Amp
Zero input Does NOT give Zero output

Input Offset Voltage

Applied input voltage to obtain Zero output

Nulling the offset Voltage

Adding External Resister according to Spec sheet
Chap 0 27

OP Amp Considerations (Cont.)



Input Bias Current

Practical OP amp
Current flowing into the terminal is NOT Zero To keep the input Tr of OP amp turned on Causes errors proportional to feedback network R

To minimize errors
feedback R should be low (<10K;)


Slew Rate
Maximal rate of change of amplifier output voltage
Ex: Slew rate of 741 = 0.5 V / Qs Time to output change from 5V to 5V = 20 Qs

To Minimize slew rate problem

Use OP amp with smaller external compensating C

Chap 0

28

OP Amp Considerations (Cont.)



Power Supply
Usually s15V
Linear Range s13V

Reducing power supply voltage

Results reduced linear range Device does not work < 4V


Different OP Amps
Bipolar Op Amps
Good input offset stability Moderate input bias current and Input resistances

FET
Very Low input bias current and Very High Input resistances Poor Input offset voltage stability

Chap 0

29

OP Amp Considerations (Cont.)



Common OP amps, Typical Specifications

 1986

Prices

Chap 0

30

Guarding
 

Elimination of Surface Leakage Currents Elimination of Common Mode Signals Very important in practice
 But

skip in this course

Chap 0

31

Passive Filters


Passive Circuits
 Contains

only passive elements

Registers, Capacitors and Inductors

 Examples

Bridge Circuit Voltage Divider Filters



Filters
 Eliminate

unwanted signal from the loop  Low Pass, High Pass, Band Pass, Notch,
Chap 0 32

Passive first-order Low pass Filter first

Pass desired Audio signal and reject undesired RF Order of Filter



Homework #2-6


Show that
Vo 1 ! , X ! RC Vi 1  j[X

Number of C and L

Plot Magnitude and Phase plot (Bode plot)  Meaning of [C



Chap 0

33

Passive first-order High pass Filter first

Pass desired High frequency signal and reject undesired low frequency signal

Homework #2-7


Show that
Vo j[X ! , X ! RC Vi 1  j[X

Plot Magnitude and Phase plot (Bode plot)  Meaning of [C



Chap 0

34

Passive second-order Low pass Filter second

To increase the attenuation of transfer function Order of Filter



Homework #2-8


Show that
Vo 1 ! Vi ( j[ / [ c ) 2  (2^ j[ / [ c )  1 [c ! 1 R C ,^ ! 2 L LC

Number of C and L


Meaning of Quality factor

[c 1 Q! ! , ([ ! 3dB BW 2^ ([ ^ ^ !1 ^ "1 1

Chap 0

35

Passive second-order High pass Filter second

To increase the attenuation of transfer function Order of Filter



Homework #2-9


Show that
Vo [2 ! Vi ( j[ / [ c ) 2  (2^ j[ / [ c )  1 [c ! 1 R C ,^ ! LC 2 L

Number of C and L

^ !1 ^ "1

Chap 0

36

Active First-order Low Pass Filter First

Inverting Amp + Feedback Capacitor

Identical frequency response with Passive filter Very Low Output impedance


Negligible Loading Effect

Chap 0

37

Active First-order High Pass Filter First

Inverting Amp + Input Capacitor

Identical frequency response with Passive filter Very Low Output impedance


Negligible Loading Effect

Chap 0

38

Active High-order Filters High

Low Pass Filters

High Pass Filters

Chap 0

39

Bandpass and Band-reject Filters Band

Butterworth Filters
Maximally Flat Magnitude response in pass band  High Attenuation Rate


Chebyshev Filters
Maximum Attenuation Rate  Ripple in pass band


Bessel Filters
Maximally flat time delay in response to step input  Attenuation Rate is very gradual


Chap 0

40

Filter Design Table



C when [0 = R0 = 1

Chap 0

41

Filter Design Example



Low pass five-pole Butterworth filter with a corner frequency of 200Hz and input resistance of 50K;
Economic Solution = 3rd order + 2nd order  Desired R and C ?


C1A = ([0 R0 C0 ) / ([ R) = 1x1x1.753 / 2Tx200x50K = 27.9 nF C2A = 21.6 nF, C3A = 6.7 nF, C1B = 51.5 nF, C2B = 4.9 nF

Chap 0

42

VCO(Voltage Controlled Oscillator)



VCO = Voltage to Frequency(V/F) Converter

VCO converts an input voltage to a series of output digital pulses whose frequency is proportional to the input voltage Applications
ADC  Digital Transmission  Telemetry  Digital Voltmeter


Chap 0

43

VCO (Cont.)


Module form


Better linearity, Lower Gain drift, Higher full-scale frequencies than IC

Monolithic IC form
Less expensive, Small size  Lower drift, Better flexibility of frequency range


Examples


LM331
Low cost VCO from National Semiconductor Maximum nonlinearity 0.01% over 1 ~ 100KHz

CD4046B
PLL contains VCO Maximum nonlinearity 1.0% over 1 ~ 400MHz

Chap 0

44

PLL(Phase Locked Loop)



VCO is commonly used in PLL Applications

Communications  Radar  Time and frequency control  Instrumentation system


Control loop


Goal
Minimize z(t) s(t) = r(t)

Change r(t) until z(t)=0

s(t) can be obtained By reading r(t)

Homework #2-10


PLL

Chap 0

45

VCO Interfacing


Output of VCO


# of pulse / Duration


Digital pulses whose frequency is proportional to input voltage

Duration
Controlled by Sampling Gate

# of Pulse
Counted in Counter

Chap 0

46

More on Passive Circuits

Divider Bridge Filter : See Chapter 1 Sensor Time Response

Divider Circuit


Convert Register Variations to Voltage Variations Vs R1 Vo R2

Output Voltage


Vo = {R2 / (R1 + R2)} Vs

R1, R2

Chap 0

48

Divider Circuit: Drawbacks



Vo is not linearly changed



Ex: Vs = 5V, R1 = 1K;, R2 = 0 ~ 1K;(Sensor)

Vo

Vs/2 Vs/3 R2 500 1K

 

Output Impedance(R1 || R2) is not so High Large Power Consumption



Both R1 and R2 dissipate power

 

Sensor


0
Insulation
49

R1, R2

Chap 0

Divider Circuit: Example



R1 = 10K;, R2 = (4K ~ 12K;), Vs = 5V

 Maximum

Vo = 5 {12 / (10+12)} = 2.73V  Minimum Vo = 5 { 4 / (10 + 4)} = 1.43V  Maximum Z = (10K || 12K) = 120/22 K;  Minimum Z = (10K || 4K) = 40/14 K;  Maximum Power = (Vo)2/R2 = (2.73)2/12K = 0.62mW  Minimum Power = (1.43)2/4K = 0.51mW

Chap 0

50

Bridge Circuit


Convert Impedance variance to Voltage variation Basic DC Wheatstone Bridge R1 E a R3 Eo R2 b R4

Eo = Ea Eb  Ea = {R3/(R1 + R3)} E  Eb = {R4/(R2 + R4)} E  Eo = {(R2R3 R1R4) / (R1+R3)(R2+R4) } E Null Condition (Eo = 0)  R1R4 = R2R3  E Example (R1 is sensor)  Eo : q


Null Condition R2

Chap 0

51

Using Galvanometer as a Null Detector



Thevenins Equivalent R1 E a R3 a IG b RG G R2

b R4


RTH = R1 || R3 + R2 || R4 = R1R3/(R1+R3) + R2R4/(R2+R4) VTH = {(R2R3-R1R4) / (R1+R3)(R2+R4)} E = Eo IG = VTH / (RTH +RG) Example
 R1=R2=R3=2K;,

RTH VTH
Chap 0

R4 = 2.05K;, RG = 50;, V=5V IG = -15.0QA

52

Bridge Sensitivity
R E R G R+(R R E R G R+(R R+(R R E R+(R R+(R G R+(R R+(R

Output (Approximation)
Eo = ((R/4R)E Eo = ((R/2R)E Eo/(R = E/2R Eo = ((R/R)E Eo/(R = E/R

Sensitivity
Eo/(R = E/4R

More Sensitive


Accuracy of Approximation
 

(R < 0.05R, then 98% Accurate (R < 0.1R, then 95% Accurate
53

Chap 0

Bridge Resolution
 

Detector Example

Resolution

 R1=R2=R3=R4=120;,

E=10V  If resolution of Detector is 10mV, Then resolution of R4 ?

10mV = {R3 / (R1+R3)}E - {R4 / (R2+R4)}E R4 = 119.52; (R = 120 119.52 = 0.48; R4 120 s 0.48 (119.52 ~ 120.48) Null Detect

Chap 0

54

Lead Compensation


Bridge at Control room and Sensor in Remote Plant Eo = {R/(R+R)} E {(R+RL)/(R+R+RL)} E {0

Compensation using 3 Lead Lines Eo = {R/(R+R+RL)} E {(R)/(R+R+RL)} E =0

R E R G

R RL R RL E

R G R

R RL R RL
55

Chap 0

Potential Measurement using Bridge

 

Galvanometer Null Then Ex ??

R1 E a R3 Ex

R2


G b R4

Ec = Ea + Ex (E = Ec Eb = - (Eb Ea Ex) = Ex + {R3/(R1+R3)}E - {R4/(R2+R4)}E At Null ((E = 0)



Ex = {R4/(R2+R4)}E - {R3/(R1+R3)}E

Chap 0

56

AC Bridges


Eo = {Z3/(Z1+Z3)}E - {Z4/(Z2+Z4)}E Null Condition



Example
Z1 = 1K;  Z2 = 2K;  Z3 = R(1K;) + C(1QF)  Z4 = R4 (?) + C4(?)


Z1Z3 = Z2Z4

Z1 E a Z3 G

Z2 b Z4

Z3 = R + 1/(j[C) = 1K j/(1Q x [) Z1Z3=Z2Z4

Re1 + j*Im1 = Re2 + j*Im2 R4 = 2K; C4 = 0.5QF
57

Chap 0

Summary of Bridge


Convert variation of resistance to variation of voltage Nonlinear response to a linear variation of resistance
 Assumed

Linear in a small range

Chap 0

58

Sensor Time Response



Desired Response


Step Response

No Sensor Gives Ideal Response



Model
Zero order First order Second order

Real Application
Assumed Zero Order or First Order  Model or Table is used


Chap 0

59

Zero Order Response



b(t) = Kvc(t-td) + bd


b(t) b(t) Kcf Kci

td = bd = 0
b(t) = K v c(t) Ideal with Gain K

td = 0
b(t) = K v c(t) + bd Biased

b(t) Kcf+bd Kci+bd b(t) Kcf Kci td t

bd = 0
b(t) = K v c(t-td) Time Delayed

t
60

Chap 0

First Order Response



b(t ) ! bi  (b f  bi )[1  e t / X ]

bi : Initial Sensor Output  bf : Final Sensor Output  X : Sensor Time Constant



To Reduce Transient Error

X << 1  Modeling the response


Mathematic Table Look Up



b(t )  bi ! (b f  bi )[1  e  t / X ]


t=X
b(t )  bi ! [1  e1 ](b f  bi ) ! 0.6321(b f  bi )

63%

t = 5X
b(t )  bi ! 0.993(b f  bi )
61

X
Transient
Chap 0

Steady State

Example: 1st Order Response



Temperature sensor
 Linear

TF: 33mV/rC, with X=1.5sec  Find output at 0.75sec after input 20 41rC
From

b(t ) ! bi  (b f  bi )[1  e t / X ] b(0) ! bi = 20rC x 33mv/rC = 660mV b(g) ! b f = 20rC x 33mv/rC = 1353mV b(t ) ! 660  (1353  660)[1  e t /1.5 ]

mV

b(0.74) = 932.7 mV
932.7 mV / 33 mV / rC = 28.3rC

Error = 41rC 28.3rC = 12.7rC

Chap 0 62

Second Order Response



R (t ) w R0 e  at sin(2T f n t )

R(t) : Transient Output  a : Output Damping Constant  fn : Natural Frequency  Ro : Amplitude



Chap 0

63