Week Data Acquisition

ENG3640 Microcomputer Interfacing

Week #8 Data Acquisition Systems Part (B)

2

Topics Sensors Signal Conditioning Analog to Digital Converters Sample and Hold Circuit

ENG3640 Fall 2012

3

Resources Huang, Chapter 12, Sections

12.1 – 12.6 Signal Conditioning Circuits

ENG3640 Fall 2012

4

Sensors and Signal Conditioning

Real

World

Measurand

Transducer

(sensors)

Analog

Mux

Signal

Conditioning

Sample and

Hold Circuit

A/D

ConvMCUD/A

ConvActuator

ENG3640 Fall 2012

Transducers Transducer

a device that converts a primary form of energy into a corresponding signal with a different energy form Primary Energy Forms: mechanical, thermal, electromagnetic, optical,

chemical, etc. take form of a sensor or an actuator

Sensor (e.g., thermometer) a device that detects/measures a signal or stimulus acquires information from the “real world”

Actuator (e.g., heater) a device that generates a signal or stimulus

realworld

sensor

actuator

intelligentfeedbacksystem

5ENG3640 Fall 2012

6

Sensors Sensors connect the

digital world to the analog real world1. Position Based Sensors2. Force Sensors3. Temperature Sensors4. Light Intensity Sensors5. Pressure Sensors6. Humidity Sensors7. ….

ENG3640 Fall 2012

7

Transducers: General Term Transducers convert variable processes such as

pressure, temperature, humidity e.t.c., into electrical signals such as voltage or current. It consists of:

1. Input interface element provides Improved coupling between measurand s(t) and sensor

(matching function) Protection to the sensor from undesirable environmental

effects Conversion of s(t) to another physical variable s1(t) required

by a sensor.2. Sensor3. Output interface element

Input InterfaceElement

SensorOutput

InterfaceElementS(t) X(t)S1(t)

ENG3640 Fall 2012

8

Shaft Angle With Potentiometer Applications:

Accelerator pedal position Steering wheel angle

Voltage Vs changes because of the change in resistance

Simple application of Ohm’s law V = I x R

Shaft angle proportional to voltage Vs Outputs the angular position of the shaft

ENG3640 Fall 2012

9

Linear Variable Displacement Transformer (LVDT)

Moving iron core changes properties of transformer Iron core position changes

primary/secondary voltage ratio Difference in phase is measured

and transformed to a voltage Voltage measured is

proportional to distance moved.

Applications: Fluid level and flow Deflection of Beams

ENG3640 Fall 2012

10

Strain Gauge: Force Measurement

Resistance varies with the amount of stretching (strain) Flexure can be measured with a

strain gauge Force can also be measured

The change in resistance is detected by measuring the voltage change in a Wheatstone bridge.

ENG3640 Fall 2012

Passive Sensor Readout Circuit Photodiode Circuits

Thermistor Half-Bridge voltage divider one element varies

Wheatstone Bridge R3 = resistive sensor R4 is matched to nominal value of R3 If R1 = R2, Vout-nominal = 0 Vout varies as R3 changes

VCCR1+R4

11ENG3640 Fall 2012

12

Temperature: Thermocouples

Outputs a voltage that is related to temp at the tip of the probe Seebeck effect: current will flow through a junction of dissimilar metals if

there is a temperature difference The voltage produced is very minute (milli-volts) The relationship between voltage and temperature is non linear

Varies from 6 uV/C to 90 uV/C

ENG3640 Fall 2012

Temperature Sensor Options Resistance Temperature Detectors (RTDs)

Platinum, Nickel, Copper metals are typically used positive temperature coefficients

Thermistors (“thermally sensitive resistor”) formed from semiconductor materials, not metals

often composite of a ceramic and a metallic oxide (Mn, Co, Cu or Fe) typically have negative temperature coefficients

Thermocouples based on the Seebeck effect: dissimilar metals at diff. temps. signal

13ENG3640 Fall 2012

14

Phototransistor: Light Detection

Current through external resistor varies with light intensity

Can be used to detect light levels or movement

Sensitive to different colors or wavelengths of light

ENG3640 Fall 2012

15

Output Transducers: ActuatorsActuators Some common actuators include

solenoids, relays, (triacs, SCRs switch ac currents).

An increasing analog signal at the gate of the MOSFET increases the amount of current drawn through the dc motor shunt field (field control).

A relay is an electromagnetic switch with a coil and one or more contacts. Applying voltage to the coil will

cause open contacts to close and vise versa.

A solenoid is like a relay but moves a mechanical cylinder instead of electrical contacts.

ENG3640 Fall 2012

16

Example A Temp sensor has a measurement range of -10 to

140C. The output range is -2.5 to +5V. The sensor is connected to an 8-bit A/D. Indicate the offset, span, step size and resolution. Also what is the digital output of the A/D if the temp is +10C?

SOLUTION:1. Offset -2.5V, -10C2. Span 5 – (-2.5V) = 7.5 V, 140C – (-10) = 150C3. Step Size 7.5/28 = 29mv, 150/28 = 0.59C4. Resolution 29mv at 8-bit To find the digital output of A/D we have to solve the

following equation (assuming relationship is linear!) Analog number = m x measurement + K

ENG3640 Fall 2012

17

Cont .. Example1. 5v = 140C x m + K 2. -2.5v = -10C x m + K Subtract 2 from 1 7.5v = 150 C x m m = 7.5V/150C m 0.05V/C Solve for K: 140C x (0.05V/C) + K = 5V K -2V Analog output 10C x 0.05V/C – 2V = -1.5V Digital Number = (Analog Number – Offset)/Step Size Digital output -1.5V – (-2.5V)/29mv = (34.8)10 == (22)16

ENG3640 Fall 2012

18

Signal Conditioning Raw sensor outputs are not always suitable for A/D

conversion. Signal conditioning circuits typically amplify the raw signal

from the sensor (i.e. thermocouple) Signal conditioning also provides:

1. buffering,2. filtering,3. offset shifting

Most signal conditioning circuits employ operational amplifiers

ENG3640 Fall 2012

19

The Inverting/Non-Inverting Amplifiers

ENG3640 Fall 2012

20

Voltage Scaling There are situations in which the transducer output

voltage are in the range of 0 ~ VZ, where VZ < VDD. Because VZ sometimes can be much smaller than VDD, the A/D converter cannot take advantage of the available full dynamic range, and therefore conversion results can be very inaccurate.

A voltage scaling circuit can be used to improve the accuracy because it allows the A/D converter to utilize its full range.

Example: Suppose the transducer output voltage ranges from 0V to

100mV. Design a circuit to scale this range to 0~5V.

ENG3640 Fall 2012

21

Voltage Scaling Circuit

AV = 1 + (R2/R1) = (5V/0.1V) = 50

R2/R1 = 49

Choose R1 = 6.8K, Then R2 = 330K. The R2/R1 ratio is 48.53. Error is smaller within 0.3%.

VOUT

+VIN

R1 R2

OP AMP

Figure 10.4 A voltage scaler

ENG3640 Fall 2012

22

The Unity-Gain Buffer or Voltage Follower

Signal conditioning can also provide buffering so that the sensor signal is not affected by anything else

connected to the circuit (i.e. minimize loading high input impedance and low output impedance)

22ENG3640 Fall 2012

23

Voltage Shifting-Scaling Circuit There are transducers whose outputs are in the range of

V1~V2 instead of 0V~VDD (V1 can be negative and V2 can be smaller than VDD)

The accuracy of A/D conversion can be improved by using a circuit that shifts and scales the transducer output so that it falls in the full range of 0V~VDD.

A Level Shifting/Scaling Circuit would consist of:I. A summing circuitII. Inverting voltage follower

ENG3640 Fall 2012

24

By choosingAppropriate valuesFor V1 and the Resistors, the Desired voltageShifting and scalingCan be achieved.

24

25

Example Choose the appropriate values of resistors and the adjusting voltage so that voltage shifting/scaling circuit can shift the voltage from –1.5V ~ 3.5V to 0V ~ 5V.

Solution:

0 = -1.5 × (Rf/R1) – (Rf/R2) × V1

5 = 3.5 × (Rf/R1) – (Rf/R2) × V1

 By choosing V1 = - 12V and Rf = R1 = R0 = 15K, R2 is solved to be 120K.

Voltage Shifting/Scaling Circuit

ENG3640 Fall 2012

26

A Difference Amplifier. Use superposition to

perform analysis If (R4 = R2), (R3 = R1) then

Vo = R2/R1 (V2 – V1) Can be used to shift offset

ENG3640 Fall 2012

27

Integrator/Differentiator

ENG3640 Fall 2012

28

Analog-to-Digital Converters: Types

A/D converters are classified according to several characteristics

Resolution (number of bits) typically 8 bits to 24 bits Speed (number of samples per second) – several

samples/sec to several billion samples/sec Accuracy – how much error there is in the conversion

Classification Staircase ADC Successive Approximation Converters Tracking ADC Flash A/D Converters Integrating A/D Converters

ENG3640 Fall 2012

29

Flash A/D: Comparators A Flash A/D utilizes comparators and encoders. A comparator compares two voltage values on its two

inputs. If the input on the + input is greater than the voltage

on the – input, the output will be logic high

29ENG3640 Fall 2012

30

Flash A/D: Encoders

What if D3 and D4 both high? Solution?

30ENG3640 Fall 2012

31

Flash A/D: Priority Encoder Example: 4-to-2 line encoder

Chooses the input with highest priority An extra output “V” could be used to validate output

31ENG3640 Fall 2012

32

Priority Encoder with Valid

32ENG3640 Fall 2012

33

Flash (Parallel) A/D Converter

Flash A/D converters can sample at several billion samples/sec

A flash A/D Converter is the simplest to understand.

It compares an input voltage Vin to a large number of reference voltages

An n-bit flash uses 2n – 1 comparators!!!

The output is determined by which of the two reference voltages the input signal is between.

Each succeeding comparator switches from a low output to a high level as the analog input increases by q.

The largest flash A/D converter is 8-bits (255 comparators!)

3-bit A/D Converter

Prio

rity

Enc

oder

ENG3640 Fall 2012

34

Integrating A/D Converters

One of the lowest cost A/D Converters (often used in digital voltmeters)

Slow used if parameter being measured is changing slowly (i.e., temperature)

Has an advantage in noisy environments (noise rejection capability)

Types: Single Slope Integrator Dual Slope Integrator

ENG3640 Fall 2012

35

The Dual-Slope A/D Conversion Method S2 is closed to discharge the capacitor S1 switches between VA and VREF

During the 1st period, converter integrates input signal VA for fixed time (T1) During 2nd period input is connected to VREF of opposite polarity so integration

proceeds to zero during variable time (T2) The counter will count during T2 representing input signal

ENG3640 Fall 2012

36

The Dual-Slope A/D Conversion Method The limited integration period results in normal-mode noise

rejection only when the integration period is equal to one or more periods of the noise signal.

The time integral of this noise over integer multiples of the noise period is ZERO.

At 60Hz minimum signal integrate time is 16.7 ms

ENG3640 Fall 2012

37

Binary Weighted DAC The binary weighted DAC covered earlier is used for small word size

systems. Why? Two major problems

1. The large resistor spread required for large word sizes i.e. R0 = 2R1 = 4R2 = 8R3 ….2. Problem with large resistor spread is difficulty of IC Fabrication3. If value of LSB is inaccurate or drifts slightly due to temp change leads to an

output error that is larger than 1 LSB Solution?

ENG3640 Fall 2012

38

Inverted R-2R Ladder Circuit

Is used to solve the problem of resistor spread and minimize drift problem in DACs with large value of N. The spread of

resistance value for the ladder is now only a 2:1 spread.

ENG3640 Fall 2012

39

Inverted R-2R Ladder: Analysis

MSB LSB

I0 = VREF/R? I1 = VREF/2R I1 will split in half in second node I2 = I1/2 VREF/4R Each succeeding vertical resistor has a value of current flow equal to half

that of previous I1 = 2 I2 = 4 I3 = … = 2N-1 IN Thus as in binary weighted resistive network, the currents controlled by

the switches are binary weighted.

I0

2R//2R = R

ENG3640 Fall 2012

40

The Glitch Problem in DACS A Significant problem occurring in DACs is that of glitching. Certain systems that are driven by a DAC ignore the glitches such as

DC motors (do not respond to these sharp transients) Bits of the code do not change simultaneously Solution? Use a Sample/Hold circuit

ENG3640 Fall 2012

41

Sample and Hold Circuit

Practical Circuit?

ENG3640 Fall 2012

42

Sample and Hold Circuit: Practical Circuit S/H circuit is a high quality capacitor and semiconductor switch. If analog signal changes rapidly during conversion (errors may be

introduce) S/H reduces these errors by quickly sampling the signal and holding it

STEADY while A/D converts Buffer amplifier (voltage follower) has high input impedance to decrease the

discharge of the capacitor. S1, closes during sampling period allows input signal to charge Ch

S1, opens up again leaving capacitor charged to the value of analog input signal.

ENG3640 Fall 2012

43

Sample and Hold Circuit: When is it required? A S/H is required if: The analog input changes more than one resolution

during the conversion time. Let dz/dt be the rate of change (max slope) of the ADC input voltage Let ΔZ be the ADC resolution Let Tc be the ADC conversion time.

A S/H is required if dz/dt . Tc > 0.5 ΔZ

43ENG3640 Fall 2012

44

Sample and Hold Circuit: Example Assume we want to design a system to measure heart sounds, x. The

useful range of y (output of microphone) is from -10 to +10mV. The desired resolution ΔY, is 0.1mV. The maximum dy/dt is 2 V/s. What is the ADC conversion required to eliminate the need for a S/H?

SignalProcessing

ADCMicrophone

x y Z n

I. dy/dt x Tc <= 0.5 ΔY

II. TC = (0.5 ΔY)/(dy/dt)III. TC = (0.5 x 0.0001 V) / 2 V/s = 25 µ seconds.

44ENG3640 Fall 2012

45

Signal Sampling Rate

The rate at which you sample a signal depends on how rapidly the signal is changing.

If you sample a signal too slowly, the information about the signal may be inaccurate.

To get full information about a signal you must sample more than twice the highest frequency in the signal (Nyquist Criteria)

Practical systems typically use a sampling rate of at least four times the highest frequency in the signal.

ENG3640 Fall 2012

46

Sampling A 1,050 Hz signal sampled at 500 Hz looks like a

50 Hz signal.

ENG3640 Fall 2012

47

Summary There are two errors introduced by the sampling

process: Voltage quantization due to finite word size of ADC Time quantization caused by the finite discrete

sampling interval Use a binary weighted DAC when resolution

required is small, else use a ladder type DAC A sample and hold circuit should be used when

signal is changing fast for ADC or if you want to avoid glitch problems in DAC.

Flash ADC are fastest but expensive (use only if speed is of importance).

A successive approximation based ADC is suitable for most applications.

ENG3640 Fall 2012

48ENG3640 Fall 2012

49

Successive Approximation A/D Is based on intelligent trial-and-error method Requires N clock periods for N-bit converter.

ENG3640 Fall 2012

50

I V & V I

Iin

R

Vout = -Iin.R

Vin

RLIout = Vin/R1

R1

ENG3640 Fall 2012