ENR 206-Sensors, Instruments and Experimentation.pptx (1).pdf

ENR 206 - Sensors, Instruments and Experimentation Professors: Baibhav Kumar Gupta (Section 1), Ashok Ranade (Section 2), Bhawnath Tiwari (Section 3) 1 Examination Scheme 2 ...

ENR 206 - Sensors, Instruments and Experimentation Professors: Baibhav Kumar Gupta (Section 1), Ashok Ranade (Section 2), Bhawnath Tiwari (Section 3) 1 Examination Scheme 2 Grade Rubric Grades A A- B+ B B- C+ C D NP Marks 90-100 80-89 70-79 60-69 50-59 40-49 35-39 30-34 0-29 Range 3  Introduction Sensors, Instruments and Experimentation 29-06-2021 | 4 A re-look around ourselves: Sensors: A tool to bring smartness→ outcomes dependent on the way it is used 5 Can you list some of the sensors available in this room ? 6 What to expect from this course: Fundamentals of Sensors Use Sensors measurement parameters Factors to consider in the selection of sensor(s) Hands-on practice on sensing systems We will cover: Light, Temperature, Pressure, Flow, Sound, Power, Position sensing Also basic use of some actuators will be performed 7 Measurement and control systems Engineering applications require measurement and control of physical quantities ❖ Satellite communication ❖ Antenna position 8 Robot position 9  Other Examples Measurement and control of temperature of oven Measurement and control of Flow in flood control 10 Transducers Capable of converting one physical quantity into another type of physical quantity The device which converts a non-electrical quantity into an electric quantity (generally) is called a sensor Examples: Thermocouple, Microphone The device which converts an electrical quantity into a non electrical quantity (mechanical) is called an actuator Examples: Piezoelectric, Shape Memory Alloy (SMA) 11  Differences Transducer Sensor Actuators Function Converts one form of Detects physical input from Converts electrical signals energy into another, the environment and into mechanical motion or including sensors and converts it to an electrical action actuators. signal. Direction of Conversion Any form of energy → Physical phenomenon → Electrical signal → Physical Another form of energy Electrical signal. action Role in Systems Used to convert energy Used to gather data from the Used to perform actions forms, acting as a bridge in environment based on control signals various systems. 12 Importance of Transducers Measurement Energy Conversion Signal Processing Systems For example, a photovoltaic cell Microphones and speakers are This allows for accurate monitoring (a transducer) converts solar examples that convert sound to and control in various applications, energy into electrical energy. electrical signals and vice versa. such as industrial automation and scientific research. 13 Importance of Sensors Data Collection Healthcare For example, temperature sensors in For instance, heart rate monitors and weather stations collect climate data blood glucose sensors provide critical that help in forecasting. health data. Safety and Security Automation and Control Smoke detectors, gas leak detectors, For instance, in a smart home, sensors and motion sensors are examples of detect occupancy and adjust lighting sensors that help protect lives and and temperature accordingly. property. 14 Importance of Actuators Mechanical Automation and Control Systems Movement Robotics They are used in robotics, industrial This is crucial in manufacturing For example, in a machines, and automotive systems processes, where precision and thermostat-controlled heating to perform tasks such as moving repeatability are required for system, the actuator adjusts the robotic arms, controlling valves, or tasks such as assembling parts or heater based on the temperature steering vehicles packaging products. reading from a sensor 15  Examples Sensors Actuators ❑ Temperature Sensor: ❖ Electric motor ❑ Pressure Sensor ❖ Hydraulic Cylinder ❑ Proximity Sensor ❖ Pneumatic Cylinder ❑ Light Sensor ❖ Solenoid ❑ Humidity Sensor ❖ Stepper Motor ❑ Motion Sensor ❖ Servo Motor ❑ Gas Sensor ❖ Piezoelectric Actuator ❑ Magnetic Sensor ❖ Thermal Actuator ❑ Sound Sensor ❖ Voice Coil Actuator ❑ Image Sensor etc. ❖ Shape Memory Alloy (SMA) Actuator 16 Applications Consumer Electronics Automotive Enhancing the functionality Improving vehicle safety of smartphones, gaming and performance through consoles, and wearable applications like anti-lock devices braking systems (ABS) and Healthcare airbag deployment Environmental Monitoring Measuring vital signs, Tracking weather such as heart rate and conditions, pollution levels, blood pressure and natural phenomena Industrial Automation Home Automation Monitoring and Enabling smart home controlling machinery, Applications devices like processes, and safety thermostats, security systems systems, and lighting control. 17 Real-time Applications Smart Home Smart Irrigation System In-vehicle sensors 18  An application Temperature alarm Temperature Comparator Sensor Vr 19  One more application Speed and acceleration measurement Speed 38 RPM A to D Speed sensor Processor converter Acceleration 4 m/sec2 20 Input output characteristics of a sensor Manufacturers specify the sensor characteristics x is the input y is the output y = f(x) 21 Processes in Measurement System Variable Input Quantity (physical Signal Processing Unit Quantity) conditioning Output X Y Sensor/transducer Standard certified meter/ Calibrated meter for input quantity Compare and identify measurement errors. The manufacturer typically provides a datasheet detailing a sensor's characteristics. 22 Curve fitting This is the process of finding a suitable equation between y and x when table of values of x and y is known. Possible equations Best fit line A polynomial A power formula (y = kxm) Piecewise linear approximation Excel has the facility for curve fitting 23 24 25 26 27 28 Class Exercise 1 Let x vary from 0 to 1 with increment of 0.1 Let y = x3 In Excel calculate y for each value of x and find the polynomial for y in terms of x. It will be close to x3. Calculate y using the polynomial and compare with original values 29 Solution to Exercise 1 30 Class Exercise 2 Let x vary from 0 to 1 with increment of 0.1 Let In Excel calculate y for each value of x and find the polynomial for y in terms of x. Calculate y using the polynomial and compare with original values. Experiment with two or more trend-lines for more accurate results 31 Solution to Exercise 2 32 More error with single trendline 33 Some aspects of instrument design and errors in measurements 34 Analog Instrument Suppose we want to build an analog temperature measuring instrument with range of 0 to 100 degree C, using a given sensor and a voltmeter Sensor characteristics are 10mV/degree centigrade Voltmeter is 0 to 5 volts. Output of sensor is 1 volt at 100 degrees so for full scale deflection the amplifier with gain 5 is required V= 0.01 T Vo = 0.05T V Vo T Amplifier Sensor Voltmeter Gain 5 35 Calibrating the meter Vo = 0.05T T= Vo/0.05 Vo T 0 0 1 20 2 40 3 60 4 80 5 100 36 Linear and non linear x Temp y1(mV) y2(mV) 0 0 0 10 100 10 20 200 40 30 300 90 40 400 160 50 500 250 Output in 60 600 360 mV 70 700 490 80 800 640 90 900 810 100 1000 1000 Temperature in degrees Centigrade Scale is crowded at low temperatures for non linear sensor for analog display 37 Digital Instrument A to D Sensor Processor Display converter y = k 1x y1 = k 2 y y 1 = k 2k 1x In general Formula or Piecewise linear approximation can be used Easy calculation for Difficult calculation for linear sensor non-linear sensor 38 Algorithm for the Digital instrument (1) Read y1 (2) calculate x = f-1(y1) (3) Send to display (4) Delay (5) Go to (1) 39  Class Exercise 2.1 A force sensor has input output characteristics given by Where V is the output voltage and F is the force in Newtons We wish to build an analog Force meter which uses a voltmeter of the range 0-10 volts. The calibration should be valid for the force range from 0 to 10 Newtons. (Full scale). So in your block diagram you have Force sensor and 0 – 10 dc voltmeter. What else you need? Think and design the complete block diagram for the meter. Show the calibration so that you get one decimal digit accuracy. 40 Solution to Exercise 2.1 We have So when F = 10 Newtons V = 6.324555 volts Clearly we can not get full scale deflection if we connect the output of sensor to the voltmeter. So we need an amplifier with gain = 10/6.324555 = 1.581139 41  Block diagram Vo Force F Amplifier sensor 42 Calibration F V Vo 0 0 0 0.1 0.632 1 0.5 1.414 2.2 1 2 3.2 2 2.828 4.5 3 3.464 5.5 4 4 6.3 5 4.472 7.1 6 4.899 7.7 7 5.292 8.4 8 5.657 8.9 9 6 9.5 10 6.325 10 43 Error in measurements What is Error ? Difference between measured and true value usually in % of full scale reading Accuracy Conformity of the measurement to the true value 44 What is a true value ? As calculated assuming perfect theory or As measured by an instrument without error This does not exist However highly accurate instruments called Primary standards are available These are available in International and National laboratories Less accurate instruments are calibrated using Primary standards These are secondary standards more often used Working standards These are calibrated using secondary standards 45 Calibration procedure Source with regulator to Instrument control the quantity under test under test Heat source Secondary or Compressor working standard Voltage source 46  Types of Measurement errors Two types of errors 1. Gross Errors Mainly covers the human mistakes in reading instruments and recording and calculating measurement results 2. Systematic Errors 1. Instrumental errors: Due to loading effect, misuse, inherent shortcomings 2. Environmental errors 3. Observational errors 4. Random Errors 47 Error distributions (Magnitude) Systematic errors True value (Magnitude Measured measurement) values Random Errors True value Measured values 48 Error distribution (Magnitude and phase) Systematic errors Measured values True value Magnitude Phase 49  Error distribution (Magnitude and phase) Random errors Measured values True value Magnitude Phase 50  Example of voltmeter V 51 Calibration table Voltage(Volts) Current 0 10 Assume meter resistance 20 30 to be zero 40 Express current in 50 microamperes 60 70 80 90 100 52 Calibration table Voltage(Volts) Current 0 0 μA Assume meter resistance 10 10 μA to be zero 20 20 μA 30 30 μA 40 40 μA μA 50 50 μA 60 60 μA 70 70 μA 80 80 μA 90 90 μA 100 100 μA 53 Voltage measurement 54 Error calculation V R1 R2 Vtrue Rm R2eff Vmeas Error E % Let V = 100 volts : Rm = 1000 kohms : R1 = 100 kohms: Calculate error and error % for values of R2 ranging from 100 kohms to 1000 kohms Discuss the results 55 Results V R1(Kohms) R2(KO) Vtrue Rm(KO) R2eff(KO) Vmeas Error Error% 100 100 100 50 1000 90.9090909 47.61905 -0.04762 -4.7619 100 100 200 66.66667 1000 166.666667 62.5 -0.0625 -6.25 100 100 300 75 1000 230.769231 69.76744 -0.06977 -6.97674 100 100 400 80 1000 285.714286 74.07407 -0.07407 -7.40741 100 100 500 83.33333 1000 333.333333 76.92308 -0.07692 -7.69231 100 100 600 85.71429 1000 375 78.94737 -0.07895 -7.89474 100 100 700 87.5 1000 411.764706 80.45977 -0.08046 -8.04598 100 100 800 88.88889 1000 444.444444 81.63265 -0.08163 -8.16327 100 100 900 90 1000 473.684211 82.56881 -0.08257 -8.25688 100 100 1000 90.90909 1000 500 83.33333 -0.08333 -8.33333 56 Overcoming the effect of errors Systematic errors Calibration can help Gain could be changed to correct for the drift If loading is known the measurement can be corrected Random errors Reduced by taking averages 57 Average of many readings We will take two examples (1) An object is to be weighed. Then to get better accuracy , it will be weighed many times. Let us say the readings are W1, W2, --- Wn. And we will take average of these readings 58 Standard Deviation We want to estimate the error as measured in a given balance Every time the weight is measured the measured weight is slightly different because of random errors To reduce the measurement error we take many readings and take the average. Standard deviation of this sample is an indication of the error. 59 Standard deviation of means Standard error We measure several such samples Each sample will have slightly different average We take the average of averages This will reduce the measurement error further Standard deviation of these averages is called standard error 60 Class Assignment Generate a sample of 7 numbers Each number is generated by adding 10 to a random number generated to have plus minus value of 0.5. Use function RAND() in excel Repeat above to generate six more samples Calculate the average and standard deviation of each sample also the average of SDs Find standard deviation of 7 averages. This is standard error 61  Standard Deviation and standard error calculation !0 samples each with 10 values Random values generated lying between 9.5 to 10.5 S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 9.62 10.48 9.56 9.63 10.36 10.36 10.05 9.76 10.42 9.57 9.83 9.99 9.91 10.32 10.12 9.98 9.55 10.11 10.18 9.61 10.07 9.60 9.52 10.15 10.23 10.28 9.50 10.28 9.60 10.19 10.30 9.82 9.81 9.58 10.03 10.14 10.05 10.35 10.13 10.37 9.57 9.80 10.26 9.67 9.89 9.75 9.72 10.16 10.47 9.62 9.85 10.37 10.24 10.33 10.25 10.02 9.99 9.61 9.52 10.38 9.86 10.03 9.92 9.53 10.04 9.70 10.16 10.29 9.59 10.47 10.01 9.79 9.53 9.66 9.77 10.02 9.56 9.79 9.76 9.79 10.47 10.10 9.66 10.07 10.32 10.08 10.11 9.91 10.23 9.89 9.96 10.19 9.56 9.60 9.67 10.05 10.48 10.27 10.47 10.05 Standard Deviation 0.28 0.28 0.28 0.32 0.23 0.20 0.32 0.26 0.38 0.35 Average 9.95 10.02 9.80 9.85 10.07 10.04 9.92 10.05 10.04 9.99 Standard Error 0.091 Average Of Averages 9.972 62 Standard error equation 63 Correction in SD Standard deviation measures dispersion of the data set relative to its mean. 64 Standard error formula It is expected to depend on SD and n. Larger the n closer will be the mean to the true value. 65 Drill Exercise 2.2 An object is weighed 7 times. The measured values are shown below. Clearly there are errors. Find the mean value of the weight and the standard error using Excel Object weight(in KG) 9.9 10.1 10.4 9.8 9.7 10 9.6 66 Answers Object weight(in KG) 9.9 10.1 Mean 9.929 10.4 Std Dev 0.269 9.8 Std Dev (n-1) 0.291 9.7 Std Error 0.11 10 9.6 67 Accuracy and precision Precision degree of reproducibility of a measurement. Target location in GPS More accurate, Less Less accurate, more precise precise 68 Resolution Smallest measurable increment A voltmeter capable of measuring minimum value of 1mV has more resolution than the one capable measuring minimum value of 10 mV considering same full scale value for both. 69 Span and range Span : Linear operating range A temperature transducer may have a linear relation between temperature and output voltage over a temperature range of 0 to 150 degrees centigrade Range : The range of measurable values Linearity : Conformity to an ideal linear calibration 70  Non-linearity Deviation from the linearity over a specified range 71 Measuring input-Output Characteristics (2) We wish to find input – Output characteristics of a device (i.e. Sensor). Let us say input is denoted by x and output is denoted by y. There is a regulator to set x and we adjust the regulator to set x from x1 to xn and measure corresponding y1 to yn values. However for better accuracy , each regulator setting is repeated for , say, nx times and average of each x and y value is taken. 72 Average of many measurements Many values of the variable, say x , are measured Similar equations apply for y values 73 Best fit line A practical sensor will have some non-linearity Manufacturers specify the best fit line Linearization of sensor Data pairs (x1,y1), (x2,y2),(x3,y3),---(xn,yn) Least square method y = mx + b 74 An Example: x y 0 0 1 2.5 2 3.2 3 5.4 4 8.5 5 9.8 6 12.5 7 13.4 8 16.3 9 17.5 10 20.4 75 The best fit line: x y z 0 0 -0.105 1 2.5 1.907 2 3.2 3.919 3 5.4 5.931 4 8.5 7.943 5 9.8 9.955 6 12.5 11.97 7 13.4 13.98 8 16.3 15.99 9 17.5 18 10 20.4 20.01 76 Worst case Fitted y Output x Input Max Deviation = Max | ymi – yfi| : i = 1 to n ym is y measured and yf is y fitted 77 Most likely deviation n-2 is used due to two reasons Measured values are mean of samples and not true values Two variables are involved in measurement , x and y 78 Hardware interface When a computer board has an Ato D converter built in , it will accept analog input. Its range will be fixed For example , for Arduino ,the range is 0 to 5 volts Between sensor and the analog input you need an interface so that the interface output matches as closely as possible to the analog input range 79 Nature of interface Many sensors have their resistance changing as the input variable changes ( Thermistor, Force sensor, Light dependent resistor) This change in resistor has to be converted into change in voltage Voltage divider can be used Some sensors produce a output voltage. But it might be small. So an amplifier might be necessary 80 Frequency output Some sensors might provide a periodic signal with the frequency dependent on the quantity being measured. You might need an amplifier and/or a level shifter as the interface 81  Design of the voltage divider interface Two possibilities If Rs decreases with variable The output voltage will increase If Rs increases with the variable output voltage will decrease If Rs decreases with variable The output voltage will decrease If Rs increases with the variable output voltage will increase 82 The choice of R Let us say max value of Rs is Rmax and minimum value of Rs is Rmin. We refer to the first circuit so that Rs is in the series arm The R >> Rmin so that output is high Also R> 1 kohms and 8kohms