ME 7203

Instrumentation and Experimental

Methods
What is this course about?
• This course has two main modules: 1) Instrumentation for mechanical engineering
applications and 2) Design of Experiments (DOE).
• The instrumentation module covers the necessary concepts to analyze static charac-
teristics, uncertainty, and dynamic characteristics of measurements.
• These concepts are applied to design measurement systems for common applications
such as temperature, load, and motion sensing.
• The DOE module covers the necessary concepts to analyze the relationships between
factors (i.e., inputs, causes) and responses (i.e., outputs, effects) of experiments.
• DOE is used to create efficient test plans for experiments, identify significant factors
of an experiment, and optimize factor levels to achieve a desired response.

Illustrated example
ME 7203 2

Chapter 1

Characteristics of sensors
1.1 Generalized Measurement Model
• A sensor is a device which provides a measurement y, of a physical variable x.
• The measured physical variable is also termed as the measurand (denoted as x).
E.g. Temperature, Force, etc.
• A measurement system performs the task of manipulating and indicating measure-
ments in a manner which is useful for a given application.
• Main components of a basic measurement system can be divided as follows:

• The sensing element has a physical characteristic which significantly changes in

response to the measurand x.
• Signal manipulation is the task of converting the output of the sensing element to
a form suitable for indication.
• The indicator provides a scale for reading the measurement.

Example 1.1: Components of a common bulb thermometer

ME 7203 3

• Modern measurement systems mostly operate using electrical signals. This allows to
exploit digital data processing (computing) capabilities in the system.

• The generalized measurement model given below summarizes components of a mod-

ern measurement system.

External Input Modules Data Processing

power Modules
x(t) y(t) yk x^
measurand Sensing Transducer Signal measurement Sampling Estimation estimate
Stage Stage Conditioning
d(t)
Extraneous Indicate
disturbance Store
Communicate
to Control
Output Modules

Sensor stage Uses a natural phenomenon to extract information of a physical

variable (measurand) x, e.g., pressure, temperature, flow.
Input modules Acts as the interface between the sensor and the data pro-
cessing modules. The input modules may include the following:
Transducer stage converts the sensed information to a measurable (elec-
trical) signal.
Signal conditioning Performs amplification and filtering of signal to de-
sired levels and specifications.
Sampling Performs analog to digital conversion of the signal.
Data processing modules Performs different processing operations of the
acquired data which may include the following:
Estimation Performs statistical estimation of the measurand, using a
set of measurements from one or many sensors. The estimate of x
produced by the estimation process is denoted as x̂.
Other establishes standard protocols for communication with devices and
perform indicating, storing, or controlling functions.
Output modules Acts as the interface between the data processing modules
and the actuators. The actuators converts the output electrical signal to a
physical response, e.g., indicator lighting, alarm sounds, conveyer motion,
heating, and ventilation.
ME 7203 4

Example 1.2: Components of a load measurement system

Example 1.3: Components of a thermistor measurement system

+5V

T co
NTC10K
AMPLIFIER. y NONLINEAR EST.

V2
A/D
R1=10K y(t) yk x^
NI6008 x

Sensor stage Resistive element which changes resistance for a given

temperature change
Transducer stage A voltage divider to convert resistance change to a
voltage signal
Signal conditioning non inverting op-amp amplifier scales up the sig-
nal to measurable levels
Sampling An NI USB 6008 DAQ performs the analog to digital con-
version.
Estimation A mathematical relationship corresponding to the thermis-
ME 7203 5

tor maps the reading to an estimate of the measured temperature

After this point the temperature reading can be displayed, the readings
can be stored, or/and the reading can be used to turn on/off a
heater.

1.2 Types of applications

• Measurement systems are used in different application classes. Three most common
applications are embedded system design, experimental Data AcQuisition (DAQ),
and industrial automation.

Embedded Design Data Acquisition Design Industrial Automation

Output modules
Data Processing

Input modules
Data Processing Data Processing

Output modules

Arduino Mini
Input modules

NI cDAQ
Input modules Output modules
Siemens S7-300

Embedded design Embedded designs focus on designing purpose built measurement sys-
tems meeting Size, Weight, Power, and Cost (SWaP-C) specification of a product
design. Typically miniaturization of the measurement system and integrating in an
industrial design is a characteristic of these applications. Examples include wearable
devices, consumer electro-mechanical products, point of care medical devices, pros-
thetics, drones, and satellites. The data processing is often performed on a single
board computer (like Raspberry Pi) or a micro controller (similar to Arduino devices
you used in the mechatronics course).
Experimental data acquisition These applications focus on getting large quantities of
accurate data for experimental purposes. Measurement system designs in labora-
tory or research environments typically employ such designs, which allow easy re-
configuration of the devices depending on the experimental study. Examples include
ME 7203 6

bio-medical instrumentation, wind tunnel instrumentation, instrumentation for struc-

tural strength testing, instrumentation of a renewable energy system optimization,
instrumentation for engine performance tuning. The data processing is typically per-
formed on a PC using data acquisition hardware similar to the National Instruments
(NI) DAQs you use in this course. This video shows a use case of a DAQ

Industrial controls These applications focus on designing measurement systems to be

compliant with industrial control standards (IEC 61131). As a result, sensors, signal
processors, data processors are available as standard plug and play type modules with
standard input output interfaces to meet common industrial applications. Examples
include process plant measurement systems, industrial boilers, marine instrumen-
tation systems, building control systems, HVAC systems. The data processing is
typically performed on a programmable logic controller (PLC) using a standard pro-
gramming syntax (IEC 61131-3). The standards makes these systems reliable, safe,
and easy to maintain. Please follow this video for a 6 minute introduction to PLC
devices.

• In this course we will mainly consider experimental data acquisition case, since we
predominantly use LabView for Labs. A set of case studies in topic 3 will review
measurement system designs for the three application scenarios discussed here.

1.3 Types of tests

• Calibration is the process of establishing the relationship between a measurand x
and a measurement y.

• Static calibration: the input (measurand) is held constant at a known value and
the output is measured after it converges to a steady state value, i.e. The dynamic
or transient behaviour is not considered in this test. Performed by applying known
values of the measurand and recording the output in order to identify a functional
relationship ŷ = f (x). The value ŷ denotes the expected value of the measurement
for a given measurand x.
ME 7203 7

Example 1.4: Static calibration of a load cell

• Dynamics calibration: The dynamic behaviour of the system is established using time
varying input signals. E.g. Finding the transfer function and frequency response of a
sensor using sinusoidal inputs of different frequencies. (The sensor should be a linear
time invariant system to support this type of modelling).

Example 1.5: Dynamic calibration of a sensor

ME 7203 8

• Different types of tests are used to model the static and dynamic behaviour of a
sensor.
• Sequential test applies a sequential set of values of the input variable over the
desired input range. This allows to capture any trends in the measurement such as
hysteresis, temperature drifts etc.
• Random tests applies a random set of values of the input variable over the desired
input range. Random tests allow to break trends in the measurement occurring
due to extraneous variables. E.g., increasing environment temperature during the
experiment. Unless it is required to study effects such as hysterisis, it is always
recommended to perform random tests for experiments.
• Repetitions are repeated measurements made in a single test run. Repetitions
allows to better estimate a quantity which is corrupted by noise.
• Replications are measurements made in different test runs under same operating
conditions. Replication permits to asses how well a set of conditions can be dupli-
cated. (captures more extraneous effects than repetitions) Replications are preferred
over repetitions because replications allows to break trends due to extraneous vari-
ables.
• Dynamic tests are performed to asses the dynamic response of a system using
different signals. I.e, Step input, Sinusoidal input, Impulse input.

Example 1.6: Identify each type of test given below

x y x y
x y x y 10 30
10 30 10 10
20 50 40 40
30 40 40 20
40 10 20 40
50 60 20 10
60 20 30 30
30 20
(a) (b) (c) (d)
ME 7203 9

Example 1.7: Find the calibration equation for the given dataset
A load cell was calibrated using a set of precision standard weights. The voltage
reading was recorded for each applied weight for static conditions. The recored values
are tabulated below. Using a software tool find a linear, and a cubic calibration
equation for the data. How would you decide on a suitable calibration model for the
data.

Load Voltage
(g) (V)
0 0.33
200 1.56
400 2.79
600 4.01
800 5.23
1000 6.449

1.4 Sensor (static) specifications

1.Sensitivity (K) The slope of a static calibration curve. K may or may not
be constant over the range of input values.
dŷ
K= (1)
dx
2.Bias (b) The shift of the static calibration curve from zero. i.e., Output
value when input is held at zero.

ŷ = Kx + b (2)

3.Range The operating range of the sensor. This can be expressed as minimum
and maximum limits of the input (xmin , xmax ) or the output (ymin , ymax )
of the sensor.
4.Span (F SO) The difference between maximum measured value and the min-
imum measured value. Can be expressed in units of the input:

ri = xmax − xmin (3)

Or can be expressed in the units of the output. Also termed the full scale
output (FSO)
ro = ymax − ymin (4)
ME 7203 10

Example 1.8
Find sensitivity, bias, range, and span using the given calibration plot.
ME 7203 11

5.Resolution (δx) smallest increment in the measurand δx that can be mea-

sured. This can also be expressed in the units of the measurement, i.e., δy.
Digitizing in electrical measurement systems are performed using Analog
to Digital Converters (ADC). The resolution δV of an N-bit ADC with a
full scale voltage span of Vspan is,

Vspan
δV = (5)
2N

Example 1.9
Find the resolution of the sensor in grams assuming that a DAQ with a resolution
of 11-bits and a voltage range of -10 to +10 V was used for measurement in the
previous example.

6.Error (e) The difference between the measured value and the true value of
the measurement (True value is the reading if the measurement system
is unaffected by any errors). Since true value is not known the expected
value is used. A deviation plot is used to illustrate errors which records
the errors along y axis and measurand along the x axis.

e = y − ŷ (6)

1.5 Error specifications

1.Accuracy (A) The accuracy is the maximum error expected from a mea-
surement system when calibrated and operated as specified by manufac-
turer. Can be divided as systematic errors (interference) and random
errors (Noise). Accuracy can be expressed in different forms:
ME 7203 12

• shown as absolute error bounds = ±emax

• shown as a percentage of full scale (FS) = ±emax
ro × 100%
• shown as statistical bounds using sample variance Sx2 (preferred)
• Expressing errors as 95% confidence bounds is preferred throughout
this course. It is reasonable to assume that absolute error bounds
represent a conservative estimate of the 95% confidence bounds, in
cases where statistical information is unavailable.

±2Sx
A= × 100% (95% C.I) (7)
ro

2.Precision(eR ) Quantifies the random variation of errors when input is held

at a constant value. This is termed as repeatability, when the sensor is
tested for a relatively short term experiment. This is termed reproducibil-
ity when tested for replications considering many extraneous factors. I.e,
different labs, different batches, different operators.
Precision as statistical bounds
• For each level of x, defined as xlevel find the sample standard deviation
of the errors.i.e, Sx (xlevel )
• Take the maximum of these standard deviations. max(Sx (xlevel ))
• The 95% confidence level of precision expressed in %FSO is
±2max(Sx (xlevel ))
er = × 100% (95% C.I) (8)
ro
Precision as absolute bounds:
• For each level of x, defined as xlevel find the span of error. r(xlevel ) =
max(xlevel )−min(xlevel )
2
• Precision error is the maximum of these spans across all different
xlevel s.
±max(r(xlevel ))
er = × 100% (9)
ro
ME 7203 13

Example 1.10: Accuracy

The deviation plot of a sensor for a random test is given below. Specify the accuracy
of the sensor in absolute bounds, and in statistical bounds. Use suitable units for
your answer (mV , g or %F S).
ME 7203 14

Example 1.11: Precision error

The deviation plot of a sensor for a random test with replications is given below.
Find the precision error of the sensor. (Sensitivity = 6.12 mV /g , FS= 6.12 V )

3.33±0.2
3

2
Error(mV)

1 0.93±0.2

-0.67±0.2 -0.67±0.3
-1
-1.47±0.4 -1.47±0.5
-2
0 200 400 600 800 1000
Load(g)

3.Resolution errors (eδ ) The error due to quantization (round off) during
sampling the measurement.
±δy
eδ = × 100% (10)
2ro
ME 7203 15

Example 1.12: Resolution error

Find the resolution error of the NI6008 DAQ [5] operating in reference single ended
mode. The unit can measure voltages from -10V to +10V with a resolution of 11
bits.
Temperature (C)

26 ±0.2

25

Time(s)

5.Nonlinearity (eL ) The error between the output and the linear approxima-
tion of the sensor. A random test should be performed to find nonlinearity.
There are two methods to measure this, Terminal point method and the
best fit straight line (BFSL) method. In this course we’ll opt for the
terminal point method which is more accurate than the second option.
• For each level of x, defined as xlevel find the average error for each
xlevel . i.e., ē(xlevel )
• Draw the terminal line which connects the first and last points of the
deviation plot. i.e., connect ē(min(xlevel )) and ē(max(xlevel )).
ME 7203 16

• Nonlinearity error is the maximum deviation of ē(xlevel ) from the ter-

minal line, expressed as percentage full scale.

Example 1.13: Nonlinearity error

The deviation plot of a sensor for a random test with replications is given below.
Find the nonlinearity error of the sensor. (Sensitivity = 6.12 mV /g , FS= 6.12 V )
4

3.33±0.2
3

2
Error(mV)

1 0.93±0.2

-0.67±0.2 -0.67±0.3
-1
-1.47±0.4 -1.47±0.5
-2
0 200 400 600 800 1000
Load(g)
ME 7203 17

4.Hysteresis errors (eh ) The difference in values found in measuring up scale

and down scale in a sequential application of the input values. A sequential
test should be performed to find hysteresis.
• For each level of x, defined as xlevel find the average error of all repli-
cations measured up the scale and down the scale. i.e., ēupscale (xlevel ),
ēdownscale (xlevel )
• Define individual hysteresis errors eh for each xlevel .
eh (xlevel ) = ēupscale (xlevel ) − ēdownscale (xlevel ) (11)
• Hysteresis error is the maximum among all eh ’s expressed as a per-
centage of full scale.
±|max(eh (xlevel ))|
eh = × 100% (12)
ro

Example 1.14: Hysteresis error

The deviation plot of a sensor for a sequential test with replications is given below.
Find the hysteresis error of the sensor. (Sensitivity = 6.12 mV /g , FS= 6.12 V )
10 -1
5
0.4667
0.4333
4
0.3733
3
0.2733
2
0.1333
1
Error(mV)

0.0667 0.0667
0

-1
-0.1733 -0.1733
-2
-0.2533
-3

-4
0 200 400 600 800 1000
Load(g)
ME 7203 18

6.Bias, sensitivity errors(eb , eK ) : The bias and sensitivity of the sensor

can change due to temperature, manufacturing tolerances, aging etc. The
expected error bounds of bias (eb ) and sensitivity (eK ) are reported as
part of sensor specifications by manufacturers.

Example 1.15: Bias and sensitivity error

LCM105 10kgf Loadcell has an output of 3 mV /V ±0.0075 mV /V and a zero balance
of ±1%F.S. The sensor is excited using a 10V supply and a 5kgf load is applied on
the sensor. Find the bias and sensitivity error of the sensor in mV .
ME 7203 19

7.Loading errors : The sensing element undergoes an energy transfer with

the physical world whenever we attempt to measure. This causes the
measured physical quantity to change from its initial value. Similarly
each stage of the measurement system undergoes transfer of energy causing
loading errors in readings.
• As an example, a resistive temperature sensor also acts as a heating
element. Therefore if we use a high excitation power we may end up
measuring the effect due to self heating of the sensing element.

Example 1.16: Loading error

PT100 is a temperature sensor that changes its resistance in response to the tem-
perature of the environment. It has a resistance of 109.73Ω at room temperature
and the sensor is excited using a 50mA current source. Assume that the probe has a
self heating rating of 50 mW/◦ C (i.e., dissipation factor: the amount of power that
it takes to the raise the body temperature of the sensing element 1◦ C). Find the
loading error in ◦ C due to self heating.
ME 7203 20

1.6 Dynamic specifications

• All of the characteristics discussed above define the static response of a sensor. This
is the response of a sensor if the input is held at a constant level and all transient
behaviour of the sensor is decayed before reading. i.e, the steady state response of a
sensor for an input with zero frequency.
• The following set of specifications defines the dynamic characteristics of sensors. i.e,
the response of a sensor for an input with different frequency components. (These
concepts are covered in detail in topic “Measurement Dynamics”).
1.Transient response : Transient response characteristics (time constant,
rise time, settling time, overshoot, natural frequency) of a sensor found us-
ing a test input signal. These can be captured using a transfer function for
a sensor if it has a Linear Time Invariant (LTI) input output relationship.

Concept Review 1.1: Transfer function

The dynamic behaviour of a sensor (linear or nonlinear) can be modelled using dif-
ferential equations. For dynamic measurement applications, linear sensors (more
specifically sensors having linear time invariant (LTI) models) are highly recom-
mended. An LTI system can be modelled using linear ordinary differential equations
with time independent coefficients, i.e.,
dn y dy dx d2 x dm x
an + · · · + a1 + a0 y = b0 x + b1 + b2 + · · · + bm (13)
dtn dt dt dt2 dtm
The input output relationship of an LTI model can be described mathematically as
a transfer function G(s), where the order of the system is the highest power of s of
the denominator polynomial.
an sn + · · · + a1 s + a0 Y (s) = b0 + b1 s + b2 s2 + · · · + bm sm X(s)

Y (s) b0 + b1 s + b2 s2 + · · · + bm sm (14)
G(s) = =
X(s) a0 + a1 s + a2 s2 + · · · + an sn
The zeros and poles of the transfer function governs its stability and response to
different signal.
ME 7203 21

Concept Review 1.2: Time domain characteristics

Parametrizes the response of a transfer function in time domain for known inputs.
Typically the response of a system for a units step input is studied.

2.Frequency response : The steady state response of the sensor for sinu-
soidal inputs. This is illustrated using a magnitude and phase plots (sim-
ilar to the plots you used in the control systems course).

Concept Review 1.3: Bode plot

• Is used to visualize the response of a transfer function G(s) for sinusoidal
inputs x = A sin(ωt) with different frequencies ω rad/s (i.e., 2πω Hz).

• The bode magnitude plot has the gain of the transfer function (i.e., the sensi-
tivity of the sensor K ) along y axis for different frequencies f Hz of the input
signal indicated along the x axis. Hence we can use the plot to find the gain
of a sensor for a given input frequency.

• Typically the gain in y axis is shown in decibels (i.e., 20log(K) dB) and the x
axis is shown in logarithm scale.
ME 7203 22

3.Bandwidth : Specifies the minimum and maximum frequency that the sen-
sor produces constant sensitivity (flat magnitude response). Usually a 3dB
specification is used. (corresponds to 70% sensitivity change).
4.Noise density (power spectral density of noise) : The precision error
of a sensor is caused due to electrical noise. This noise can be modelled as a
summation of Gaussian noise signals with different frequencies and similar
power levels at each frequency, i.e, additive white Gaussian noise. This
constant power level is termed the noise density. Noise density allows us
to calculate the standard deviation of noise using the following equation:
√
σnoise = N oise Density Bandwidth (15)
ME 7203 23

Example 1.17: ADXL 335 static and dynamic response

Mark all sensor specification indicated by boxes in the following static response plot
and frequency response plot of the ADXL335 accelerometer. (use the datasheet of
ADXL 335 [6] to answer this question)

20log|K|
V

b 3dB

f
g`s

dB/dec
ME 7203 24

1.7 Standards
• Calibration process uses known values of the measurand, which is assumed to be
highly accurate. These “known values” used for calibration purposes are referred to
as standards.
• Standards are related to fundamental dimensions and units.
• Dimensions define some aspect of a physical variable. Units define a quantitative
measure of a dimension. SI standard defines 7 basic dimensions and their units.
• For each basic unit a primary standard is defined which exactly quantifies the unit
(with zero error).

Dimension SI Units Primary Standard

Time (second) s Time elapsed during
9,192,631,770 periods of
the radiation emitted be-
tween two excitation levels
of the fundamental state
of cesium-133. i.e., specific
atomic transition frequency
(∆vCs )
Length (meter) m a unit meter is c∆vCs , where
c is the speed of light
Mass (kilogram) kg Mass of Platinum Irradium
bar maintained at IBWM -
Severs, France. “Le Grand
K”. As of 2019, a kilogram is
h∆vCs
c2
, where h is the Planck
constant.
Temperature (Kelvin) K Different standards are used
depending on the scale. E.g.
triple point of hydrogen 13.81
K
Current (Ampere) A The current that produces
2e−7 N between two parallel
conductors

• The primary standards (which corresponds to zero error) are difficult to obtain for
calibration. Therefore secondary or derived standards can be used (which has some
error relative to the primary standard). The accuracy of the standard deteriorates
down the hierarchy of the derived standard.
ME 7203 25

• Therefore the error of the standard used for calibration (±es ) deteriorates the ac-
curacy of a sensor. This is taken in to account when a manufacturer specifies the
accuracy of a sensor in the datasheet. If we perform a separate calibration of a sensor
using some standard, then we should take in to account the standard error (±es ) as
a error component which affects the system.

1.8 Root sum squared method

• We talked about many types of component errors of a measurement system. (e.g.
precision, resolution, nonlinearity, standard, hysteresis, bias and sensitivity errors)

• A conservative estimate of the overall measurement system error is the root sum
squared (RSS) of all component errors.
q
etotal = e2L + e2h + e2b + e2K + e2r + e2s (16)

• This calculation assumes that the square of each error term is proportional to its
variance.

• All error components should be established using a statistically large sample (N ≥

30). All component errors should be expressed in same units and all component
errors should be expressed at same confidence level (e.g, 95% ) before applying the
equation.

• If only the absolute bounds of a component error is available, then it is reasonable

to assume it as a conservative estimate of 95% confidence bound of the component
error.
ME 7203 26

Example 1.18:
The following table summarizes specifications of a LM35 Temperature measurement
IC. Find the total accuracy of the measurement system in ◦ C using the provided
data.
 References
[1] Figliola, R.S. and Beasley, D., Theory and Design for Mechanical Measurements, 6th
Edition, Wiley, 2015.

[2] Statease design expert tutorials, Statease, 2018. URL: www.statease.com/docs

[3] e-Handbook of Statistical Methods, NIST/SEMATECH, 2012. URL:

http://www.itl.nist.gov/div898/handbook/

[4] Lye, L., Design and Analysis of Experiments, DOE Course notes, 2015.

[5] NI USB-6008/6009 - User Guide, National Instruments, 2015. URL: 371303n.pdf

[6] Analog Devices 335 3 axis accelerometer datasheet, Analog Devices, 2015. URL:
adxl335.pdf

[7] Temperature probes: How to choose the right temperature sensor type, Omega Inc. 2018.
URL: thermocouple-RTD.html
ME 7203 27

Chapter 2

Signal Conditioning for Mechanical

Measurements
2.1 Types of sensors and signal conditioning
• Sensors can be broadly classified as non-self generating sensors and self generating
sensors.
• Non-self generating sensors are sensing elements that do not produce a voltage or a
current signal in response to a measured physical quantity.
• Self generating sensors are sensing elements that produce voltage or current signals
in response to a measured physical quantity.
• Depending on the type of sensing element used, these can be further categorized as
follows:
External Input Modules Data Processing
power Modules
x(t) y(t) yk x^
measurand Sensing Transducer Signal measurement Sampling Estimation estimate
Stage Stage Conditioning
d(t)
Extraneous Indicate
disturbance Store
Communicate
to Control
Output Modules
ME 7203 28

• Referring back to the generalized measurement model, we should incorporate the

transducer and signal processing stages to convert the signal from the sensor to
measurable levels.

• Industrial standards dictate the output span of the signal conditioning modules to
scale the sensed variable to a 0 − 5V (or 0 − 10V ) Voltage signal, 4 − 20mA current
signal, or a frequency signal (sensed information is communicated using a digital FM
waveform, RS 485, USB etc.).

• Signal conditioning needs vary widely depending on your sensor, so theres no one gen-
eral instrument which can provide all types of conditioning requirements. Following
table summarizes the signal conditioning needs of different types of sensors.

Example 2.1
Briefly explain each of the following signal conditioning stages used for instrumen-
tation.

• Amplification:

• Attenuation:

• Isolation:

• Filtering:

• Excitation:

• Linearization:
ME 7203 29

• Recommended reading: please refer the NI engineers guide to signal conditioning

link to gain an overview of different signal conditioning operations and how they are
applied for commonly used sensors.

2.2 Resistive sensors

• Resistive sensors are frequently employed as the sensing stage of temperature, strain,
load, and pressure transducers.

• Resistance of a conductor is governed by the following equation, where R is the

resistance, ρ is the conductivity of the material, l is the length of the conductor, and
A is the cross sectional area of the conductor.
ρl
R= (17)
A

• A resistive sensors can be thought of as a variable resistor which changes its resistance
in response to some physical variable.

• The resistance of the sensor can be measured using a current source or a voltage
divider.

Concept Review 2.1: Voltage Divider

The supply voltage Vs across two resistors in series is divided between the two
resistors as follows. By measuring the voltage Vo , and knowing Vs and R1 one can
find the unknown resistance Rs .

Vs

R1
Vo Vo= RsensorVs
R1+Rsensor
Rsensor

Concept Review 2.2: Ohm’s Law

The voltage drop across a resistor is equal to the current flow Is times the resistance
Rs of the resistor. By measuring the voltage Vo , and knowing the supply current Is
one can find the unknown resistance Rs .
ME 7203 30

Is
Vo Vo= Rsensor Is

Rsensor

Example 2.2: Voltage divider

A potentiometer is a device used to measure displacement x. A sweep arm is
used to divide the total resistance Rl of the potentiometer. Assume that the
supply voltage is Vs = 10, the total resistance Rl = 10k with l = 10cm.
Find the output voltage when the sweep is displaced 25% of its full range.
1
Vs

Rl l 2
Vo
x Rx
xR
Rx= l
3 l

• In order to measure the output voltage of a voltage divider, we must use a high
impedance measuring device, i.e, a device which draws a very low current from the
voltage divider when measuring. Otherwise the system will produce loading errors
ME 7203 31

Concept Review 2.3: Op amp

Operational amplifier is an integrated device which amplifies the voltage difference
of its inputs V+ , V− with a gain A to produce its output Vo . They have very high
input impedance Rinput → ∞, very low output impedance Routput → 0, and very
high open loop gain A → ∞. The devices can be configured to operate as amplifiers,
followers, filters, etc.
Supply Voltage
Inverting Input +Vs
V+ +
Output

V- - Vo
Non- Inverting Input
-Vs
Vo = A(V+ − V− )

• An op amp voltage follower can be used to minimize the effects of inter stage loading.
• Op-amps have a very high input impedance which draws a very low current from the
connecting circuit, minimizing any loading effects.

Example 2.3: Voltage follower

Find the output Vo for the following Opamp circuit.

+Vs
Vi +

- Vo

-Vs
ME 7203 32

Example 2.4: Interstage loading

Find the output Vo for the following two circuits. What is the interstage loading
error in V when a voltage follower is not used?
Vs=10V Vs=10V

R1=10k Measuring device R1=10k

Measuring device
Vo +
Vo
-
R2=10k Rm=100k R2=10k Rm=100k
ME 7203 33

Concept Review 2.4: Common Op amp Circuits

Find the output voltage Vo for each op amp circuit shown below as a function of
inputs.
Inverting amplifier:
R2

+Vs
Vin R1
- Vo
+
-Vs

non inverting amplifier:

R2

+Vs
R1
-
Vo
Vin +
-Vs
ME 7203 34

differential amplifier:
R1

+Vs
R2
V1 -
Vo
V2 +
R2
-Vs
R1

instrumentation amplifier:
+Vs
V1 + R3 R3

-
R1 +Vs
-Vs
-
Rg Vo
+
-Vs
R1
-
V2 + R2 R2
-Vs
ME 7203 35

2.3 Measurement of temperature

2.3.1 RTDs
• Resistive temperature devices (RTDs) are a class of temperature sensors which change
their resistance in response to change in temperature.

• The change in resistance of metals when subjected to heating, is due to thermal

expansion of the conductor and the change in resistivity of the material.

• The following graph shows resistance vs temperature for common sensing elements.
Platinum devices exhibit good linearity and long term stability.

• The change in resistance of an RTD is best described by the R-T table. e.g., PT100
R-T Table URL: pt100.pdf

• For simplicity the response of an RTD can be approximated by a linear function.

R = R0 + αT (18)

• The change in resistance for a given change in temperature is defined by the temper-
ature coefficient α, the resistance at reference temperature (typically 0 ◦ C) is given
by R0 and the measured temperature in Celsius is denoted as T .

• A higher degree polynomial function is preferred for higher accuracy temperature

measurement. The Callendar-Van Dusen Equation is commonly used to define the
nonlinear response of an RTD.

R = R0 (1 + AT + BT 2 ) f or T > 0◦ C
(19)
R = R0 (1 + AT + BT + C(T − 100)T ) f or T < 0◦ C
2 3

• In the above equation A, B and C are manufacturer provided polynomial coefficients.

ME 7203 36

• The accuracy of RTDs depends on the tolerance class of the sensor and temperature
range that is measured. e.g. PT100 tolerance table URL: pt100.pdf
• Use of a Voltage divider for a RTD makes the output voltage nonlinear. Use of a
current source is recommended for linear response.
• These devices exhibit self heating errors and very low sensitivity. I.e. 0.385Ω/C ◦
sensitivity and 50mW/C o self heating capacity in water for the PT100 RTD.
• The self heating effect of the transducer can be minimized by making use of a very
low excitation current to drive the sensor and an amplification stage is recommended
to improve its sensitivity.

Example 2.5: RTD Excitation

Find a relationship between the input temperature T and output Voltage Vo of the
following two measurement circuits. Also find the power dissipation at T = 0◦ C in
each case.
Vs= 5V

Tco

PT100
+5V
+
Vo1
-
Rs=100

Vs= 5V

LM334

R2=68.3
+5V
+
V02
Is~1mA -
T co
PT100
ME 7203 37

• Lead wire resistance is a major source of error when they run long distances. This is
compensated by 3 wire or 4 wire RTD circuits.

Instrumentation
Amp
+ Vo
-

Example 2.6: Leadwire effects

Find the output Vo for the following two circuits, and show that the 4 wire circuit
cancels off the lead wire effects of an RTD.

Is
Is

Rw Rw
+
+
T co - Vo Vo
PT100 Rw -
Rw T co
PT100
Rw
Instrumentation
Amplifier
Gain =1
Rw
(b)
(a)
ME 7203 38

2.3.2 Thermistors
• Thermistors change their resistance as temperature changes similar to RTDs. The
main difference is that thermistors offer much higher sensitivity than RTDs.

• Thermistors with positive temperature coefficients are categorized as PTC’s, and

ones with negative temperature coefficients are termed NTC’s.

• The temperature response of thermistors are not linear, it can be approximated using
the following equation.
β( 1 − 1 )
R = R0 e T T0 (20)

• The variable R0 is the nominal value of resistance at reference temperature T0

(e.g., 25◦ C), and R is the resistance of the thermistor at the measured temperature
T . The temperature values for this equation are in absolute scale (in Kelvin). By
knowing the beta value β of the thermistor, one can find T by solving the equation.

• Although thermistors offer better response and sensitivity, they have less range, and
are also prone to self heating. Additionally, the nonlinear response (variable sensi-
tivity) of these devices is undesirable when designing precise measurement systems.

• For best linear response about T0 the thermistor can be used in a voltage divider
configuration with a series resistance equal to R0 as illustrated below.

Figure: Response of NTC thermistor with β=3984K ◦ , R25 =10kΩ at reference temperature
25C ◦
ME 7203 39

Example 2.7: Thermistor model

A thermistor is placed in a 100C o environment, and its resistance measured as
20, 000Ω. The material constant, β, for this thermistor is 3650K o . If the thermistor
is then used to measure a particular temperature, and its resistance is measured as
500Ω, determine the thermistor temperature. (Use the Beta equation)

2.3.3 Thermocouples
• Thermocouples(TCs) are self generating devices, i.e., produces a voltage signal in
response to temperature without any excitation.

• Applicable in very high temperatures when compared with RTDs and thermistors
[7].

• Thermocouples operate using the Seebeck effect. I.e., generation of a voltage potential
VT1 −T2 (emf) in an open thermocouple circuit due to a difference in temperature (T1
and T2 ) between dissimilar material junctions in the circuit.

• The voltage due to Seebeck effect is not produced when the junctions are at same
temperature VT1 −T1 = 0.

• Standard thermocouple tables provide output voltage of different thermocouples for

sensed temperature. These assume 0C ◦ as the reference junction temperature. e.g.
refer link for Type K thermocouple table URL: TypeK.pdf
ME 7203 40

• During measurement, if the reference junction is at a different temperature (e.g. 25C ◦

) we should add the voltage corresponding to cold junction compensation (CJC) V0−25
to the measurement, when comparing it with the values from TC tables.

V0−T = V0−25 + V25−T (21)

Material A Material A
T1 T2 T1 T2

V1-2 Material B V1-2 Material B

+
=

Material A Material A
T1 T2 T2 T3

T3 V2-3
Material B Material B

Material C
V1-2

=
=

Material A Material C Material A

T1 T1 T3
T2 V1-2

Material B
V1-3 Material B

Thermocouple Laws

Example 2.8: Cold junction compensation

Figure below illustrates a signal processing module designed for reading a thermo-
couple. An NTC thermistor is placed to record the junction temperature of the
terminal block as shown in figure. The thermistor has a beta value β = 3000 ◦ K
and a reference resistance of R25 = 30 kΩ at reference temperature 25 ◦ C. The
standard thermocouple output voltage V0−T (measured with a reference junction
temperature of 0◦ C) for different temperatures T ◦ C at the measured junction are
given below.

1. Find the junction temperature TJ if VDAQ1 is 2.5V .

2. The thermocouple output voltage VDAQ2 is found to be 47.8mV for a particular

experiment when Vb = 0V and Rg = 49.4kΩ. Find the measured temperature
T ◦ C for this case.
ME 7203 41

T co + +5V + OP AMP

K-Type VTJ-T - A/D VDAQ 1

TC - TJ Co
30kΩ NI6008

NTC
5V
V+ + INSTRUMENTATION
SIG+ AMP
K-Type Thermocouple Voltages
SIG- V- - A/D VDAQ 2
V0-0 =0 mV GND V1 NI6008

V0-25 =1 mV OpAmp
Adder
V0-100 =4.1 mV (Bias Compensation)
V0-600 =24.9 mV Rg=
VT-0 =-V0-T
Vb= Gain=1+(49.4kΩ/Rg)
ME 7203 42

2.4 Measurement of strain

2.4.1 Resistive strain gauges
• Introduced in late 1930’s by Lord Kelvin, the resistive strain gauges are devices which
significantly change its resistance when subjected to strain (when elongated).
• The relation in equation (??) can be transformed to log and differentiated to derive
the following.

dR/R = dρ/ρ + (1 + 2ν)dl/l

• Manufacturers provide a quantity termed the gauge factor (Gf ) to easily represent
the sensitivity of the device. e.g. Omega prewired strain gauge data sheet URL:
KFH.pdf

∆R/R ∆R
Gf = Gf ε =
∆l/l R
• Here ε = ∆l/l is the axial strain experience by the strain gauge and R is the nominal
gauge resistance. Therefore by knowing the change in resistance ∆R, the nominal
gauge resistance R, and the gauge factor Gf , one can find the axial strain ε of the
strain gauge.

Example 2.9: Guage factor

A steel specimen is subjected to 1kN m−2 of stress. This is instrumented with a
strain gauge with R = 100Ω, Gf = 2 along the direction of principal stress. What
is the change in resistance ∆R of the strain gauge? (Esteel = 200GP a)
ME 7203 43

• The sensitivity of a strain gauge on a typical steel specimen will be 10−6 Ω/(kN/m2 ).
I.e., a millionth of the nominal resistance would change for a 1kN/m2 stress.
• Therefore a highly sensitive setup should perform the task of converting the change
in resistance ∆R to a measurable voltage signal.
• The Wheatstone bridge is a circuit which converts minute changes in resistance to
measurable voltage.

Concept Review 2.5: Wheatstone Bridge

The Wheatstone bridge is a combination of two voltage dividers balanced to produce
zero output voltage. The output Vb becomes nonzero whenever theres a resistance
change ∆R/R at any of its bridge arms.
Vs

R1 R4

V2
V1
R2 R3
Vb=V2 - V1
ME 7203 44

R1 R4
= =⇒ Vb = 0 (22)
R2 R3
 
1 ∆R1 ∆R2 ∆R3 ∆R4
Vb = − + − Vs (23)
4 R R R R

• The Wheatstone bridge can be used in many configurations for strain measurement:

• Quarter bridge: The configuration produces an output for both bending and axial
loading. There is no temperature compensation for this configuration.

1
Vb = Gf εVs (24)
4

Vs

R1 R4+ε
V1
V2
R2 R3
Vb=V1 - V2

• Half bridge - Type 1: This configuration also produces an output for both bending
and axial loading. This configuration is temperature compensated because both R3
and R4 experience equal effects due to temperature.
1
Vb = Gf ε(1 + ν)Vs (25)
4
ME 7203 45

Vs

R1 R4+ε
V1
V2
R2 R3-νε
Vb=V1 - V2

• Half bridge - Type 2: This configuration only produces an output for bending.
Strains due to axial loading cancels off. The configuration is temperature compen-
sated because both R3 and R4 experience equal effects due to temperature.
1
Vb = Gf εVs (26)
2

Vs

R1 R4+ε
V1
V2
R2 R3-ε
Vb=V1 - V2
ME 7203 46

• Full bridge: This configuration only produces an output for bending. Strains due to
axial loading cancels off each other. The configuration is temperature compensated
because all resistors experience equal effects due to temperature. The sensitivity of
this configuration is 4 times the quarter bridge.

Vb = Gf εVs (27)

Vs

R1-ε
R4+ε
V1
V2
R2+ε R3-ε
Vb=V1 - V2
ME 7203 47

2.5 Measurement of load, torque, and pressure

2.5.1 Load cells
• Load cells are structural elements which are pre instrumented using calibrated strain
gauges (typically as full bridge).
• Manufacturers provide load cells of different size, shape and measurement ranges.
e.g. S type, beam type, cell type etc. URL: Omega load cell selector
• Manufacturers state the ymax mV /V output of the sensor, i.e. this is the output of
the sensor at its full load xmax kg if an excitation voltage of Vs = 1V is used.
• Therefore you should multiply ymax mV /V by the supply voltage Vs of your setup,
in order to find the corresponding output at full load. Using these quantities the
sensitivity of the load cell can be found as follows:
ymax Vs
K= mV /kg (28)
xmax

Example 2.10: Load Cell

A steel cantilever beam is fixed at one end. A load F is applied at the free end. Two
axially aligned strain gauges (Gf = 2.1) are mounted to the beam as shown. The
bridge deflection output is passed through an amplifier (gain, K = 1) and measured.
(Beam material: Aluminum 6061-O with E = 69GP a and σallow = 60M P a. Beam
dimensions l = 230mm, d = 30mm, w = 26mm and h = 3mm. Guage factor = 2.1)
F Vs

R1
SG1 R+ΔR R4 +Vs
V+ +
h Vo
-
w SG2 R-ΔR R2 R3
V-
-Vs
d AD620
Gain = K
l
Vb=V+-V-

σ My F (l − d)h/2
SG1 = = =
E EI Ewh3 /12

1. Find an equation relating the output voltage (Vo ) and the applied load (in
kg) of the half bridge load cell shown in figure and find the expected mV /V
output rating of the load cell (without any amplification) assuming that the
rated load is 1 kg.
ME 7203 48

2. Find the maximum load (in kg) that can be applied to the load cell as a
function of the parameters.
ME 7203 49

Example 2.11: Load Cell

1. Figure below illustrates a load cell signal processing module used in Engi 7930
Labs. Clearly draw the wiring diagram to show how a 120Ω half bridge load
cell should be connected to this board.
2. The circuit illustrates a bias adjustment circuit used to balance the bridge.
Find the range of bridge imbalance (in millivolts) that can be adjusted using
this circuit.
Half Bridge +10 V From Power
Cell supply
+10V
SG1
120
-10V

+10 V
SG2 120
EX+
+10V
SIG+ + V1 3
SIG- 1 7 Vo
Full Bridge +10V Rg 6
EX- 5
+ 8 4
Cell 3
1 7
6
- V2
2
SG1 SG4 5 -10V
8 4
To DAQ
Gain Amp
+10V - 2
(AD 620)
-10V
SG2 SG3 1K 200 1K Bias Amp
-10V (AD 620)
Bias adjustment
ME 7203 50

2.5.2 Torque measurement

• Strain gauges can also be configured to measure the principal stresses of a shaft.
• The strain gauges should be oriented along the directions of maximum principle
stress, i.e. 45◦
• The bridge should be configured to produce the maximum sensitivity (Voltage out-
put) for applied torque.

Example 2.12: Torque Cell

Find the output voltage as a function of the applied torque for the illustrated mea-
surement system. Assume that the shaft diameter is D and polar moment of inertia
of a circle is J = πD4 /32.
Vs

R1
R4 +Vs
V+ - Vo
V- +
R2 R3 -Vs

T
Source
2 +ε 2 3
-ε 3 Load
T τ
+ε 4
-ε 1 4
1
ME 7203 51

2.5.3 Pressure sensors

• One approach to measure pressure is the use of an instrumented diaphragm.

• Depending on the pressure applied P on the diaphragm strain gauges are elongated
and compressed along tangential and radial directions.

Figure: Strain distribution of clamped diaphragm

• Pressure sensors are available to measure absolute, gauge pressure or vacuums pres-
sure. The output is typically specified as a mV /V rating. e.g. Demo of mV out-
put pressure transducer acquisition in Labview URL: demo-video-pressure-reading-
LabView
ME 7203 52

Example 2.13: Pressure transducer

Match the strain gauges with the bridge arms for the following pressure sensor in
order to produce a positive voltage output for positive gauge pressure.

SG1

SG2
SG1 SG2

SG4 SG2

SG3 SG4
SG3

SG4
SG1 SG4 are radial guages
SG2 SG3 are tangential guages
ME 7203 53

2.6 Voltage, current, frequency output

• The resistive sensors detailed above make use of transducer and signal processing
stages to produce signals of measurable levels. e.g. 0 − 5V

• Many industrial sensors are packaged with the transducer and signal processing stages
to provide standard output levels.

• Typically the output is rated as voltage (0 − 5V, 0 − 10V), current (4 − 20mA), or

frequency (frequency modulation, RS232, USB). E.g. Omega pressure sensor selector
tool URL: Pressure-sensor-selector-tool)

• A 0 − 5V sensor outputs 0V for the minimum rated input xmin and outputs 5V for
maximum rated input xmax .

• Data acquisition systems digitize the voltage signal using an analog to digital con-
verter. E.g NI 6008 DAQ [5]

• When a voltage signal travels long distances at a site, the signal drops due to resis-
tance of the long wires used for the purpose.

• A remedy for this is the use of current signals. Current signals are efficient for
signal transmission over long distances. (e.g. Automation direct 4-20mA temperature
transmitters URL: demo-video-temperature-transmitters)

• A 4 − 20mA sensor outputs 4mA for the minimum rated input xmin and outputs
20mA for maximum rated input xmax of a sensor. Out of range readings are used to
indicate sensor faults or sensor startup routines.

• For basic data acquisition needs a 250Ω current sensing resistor can be used with an
analog to digital converter to perform digitization of 4 − 20mA signals.
ME 7203 54

Example 2.14: 4 − 20mA Sensors

Omega FMA-900-I is a 4 − 20mA output air velocity probe with a range of
0 − 100 SF P M . The NI 6008 DAQ measures the following voltages during an
experiment. What is the measured airspeed in standard feet per minute (SFPM).
a) 0V b) 1V b)5V

Vo

NI6008 -
Ai0 +

250
NI6008
GND
ME 7203 55

2.7 Analog and Digital I/O

• As discussed in the previous subsection, sensors can be selected having a standard
type of output (e.g. 0 to 5V, USB). These standard outputs of the sensors can be
broadly classified into analog and digital output signal types.

• Analog signal: the voltage ( or current) varies between a range of values and the
voltage values directly correspond to the measured quantity following a standard
linear relationship. As an example, a 0-350 mBar pressure sensor with 0-5V output
has a linear relationship with 0V corresponding to 0 mBar, and 5V for 350 mBar.

• If the sensor outputs a standard analog signal (0-5V, 4-20 mA), analog input modules
can be used to interface the signal to a data processing device. Example illustrations
of the analog input ports of embedded, DAQ, and PLC data processing devices are
illustrated below.

Embedded Design Data Acquisition Industrial Control

Digital IN
Digital IN*

Analog IN
Digital OUT
TCP/IP
Analog OUT
Digital IN*
USART
Analog IN
Analog IN
USART

Analog OUT

*can be configured as in or out, PWM

Figure 1: I/O ports - data processing modules

• These analog input modules perform the function of analog to digital (A to D) con-
version, which converts a voltage signal to a digital representation that the data
processing units can work with. Following are typical examples of interfacing stan-
dard analog output of a sensor with an analog input module of a data processing
unit.
ME 7203 56

DAQ
0-5V VS 8-bit
Output SIG 3V AD0
Sensor GND GND USB
VS+ VS-
Power
Supply

DAQ DAQ
4-20mA VS 5mA R=250 8-bit 4-20mA R=250
10mA 8-bit
Output SIG AD0 Output SIG+ AD0
Sensor GND GND Sensor SIG-
USB GND USB
3 wire VS+ VS- 2 wire VS+ VS-
Power Power
Supply Supply

Example 2.15: Analog to digital modules

For each of the examples above find the recorded voltage at the DAQ chan-
nel AD0, the corresponding binary, and decimal representation of the value
recorded by the data processing unit. Assume an ADC voltage range of 0-
10V.
ME 7203 57

• Digital signals have the voltage changing between only two levels. For example high
voltage (2 to 5V) corresponds to digital “1” and low voltage (0 to 0.8V) corresponds
to digital “0” for digital TTL output type sensors.
• The Digital input modules are capable of recording the state (High or Low), keep
track of the frequency, or the pulse width of a digital signal. Example illustrations
of the digital input ports of embedded, DAQ, and PLC data processing devices are
shown in figure ??.

5V
DAQ
TTL VS
Output SIG D0
Sensor GND GND USB
VS+ VS-
Power
Supply

Example 2.16: Digital outputs

a) A 3-bit digital output sensor uses three digital signal wires (V0 , V1 , V2 ) to
communicate with a digital input module of a DAQ, where V0 corresponds to
the least significant bit. The sensor communicates decimal number 5 at time t
= 0 and changes the value to decimal number 3 at time t=5 . Assuming 0-5V
digital logic levels draw the voltage vs time graph of each of the digital lines
connecting to the DAQ.
ME 7203 58

b) An encoder with 150 pulses per revolution is rotating at a speed of 1 rpm.

Assuming 0-5V digital logic levels draw the voltage vs time graph of the en-
coders output signal.

c) A PWM-Output linear Hall effect magnetic sensor has a measurement range

of ±21mT and a sensitivity of 2%Duty/mT . The output at 0mT is 50%duty.
Assuming 0-5V digital logic levels and a 1kHz PWM frequency, draw the
voltage vs time graph of the sensor output signal when the sensor is reading
5mT .

• Digital frequency signals are also commonly used to transmit speed values.

• One option is to modulate the frequency of a digital waveform in response to a sensed

variable. E.g. An optical encoder produces a digital waveform where the frequency
of the waveform is proportional to the speed of rotation of a shaft.

• In this case a data acquisition system captures the frequency signal using a counter
and a clock. E.g NI DAQ digital pulse counting example URL: demo-video-counter-
Labview
ME 7203 59

Example 2.17: Frequency output sensors

A hall effect sensor is attached to a fly wheel of an engine as shown to generate a
frequency signal corresponding to its rotation. A 32-bit 5M Hz counter on NI 6008
records the the number of counts C during each 5 second period. After the 5s period
the counter is reset to restart counting. The flywheel has 150 teeth. Find the speed
of the flywheel if C = 1000 during a certain 5s interval. What is the maximum RPM
that the DAQ can measure using this sampling strategy.

NI6008
CTR0
ME 7203 60

• Digital sensor can also make use of a communication standard to communicate with
a device. Communication standards define how a “packet” of digital data is com-
municated between two devices. These data ”packets” can conveniently carry any
information related to the sensor and its measurements. Three common protocols
are shown below IC to IC (i2c), universal asynchronous receive transmit (USART),
and ethernet (tcp/ip).

Communication standard Description

– Has two signal lines for com-

munication.
VS DAQ
I2C SDA SDA – SDC line communicates a clock
Output SCL SCL
Sensor GND GND
USB
which is used to figure out
VS+ VS-
when to read the SDA data
Power line.
Supply
– SDA line is used to communi-
cate the data.
– I2C supports 100 kbps, 400
kbps, 3.4 Mbps and can have
27 devices connected to the
same line.

– USART has two lines, one for

receive (RX) and the other for
VS DAQ
USART TX RX transmit (TX).
Output RX TX
Sensor GND GND
USB
– Both sender and receiver
VS+ VS-
should be configured to com-
Power municate using a common
Supply speed (baud rate).
– A start bit, stop bit and data
size is also configured which de-
fines the data packet.
ME 7203 61

– Directly interfaces to a ether-

VS Ethernet Router net modem using a standard
Ethernet
Sensor ethernet CAT 5 cable.
GND
ethernet or Wifi
VS+ VS- – The sensor gets assigned an ip
Power
Supply
address.
– The ip address can be used
to communicate with the sen-
sor using a local network or
through an internet network
connection.

• Similar to the input ports discussed above, data processing modules can also have
output ports. Analog output ports perform digital to analog conversation ( D to A)
to output an analog signal. Digital output ports are used to output digital signals to
communicate with another digital device.
 References
[1] Figliola, R.S. and Beasley, D., Theory and Design for Mechanical Measurements, 6th
Edition, Wiley, 2015.

[2] Statease design expert tutorials, Statease, 2018. URL: www.statease.com/docs

[3] e-Handbook of Statistical Methods, NIST/SEMATECH, 2012. URL:

http://www.itl.nist.gov/div898/handbook/

[4] Lye, L., Design and Analysis of Experiments, DOE Course notes, 2015.

[5] NI USB-6008/6009 - User Guide, National Instruments, 2015. URL: 371303n.pdf

[6] Analog Devices 335 3 axis accelerometer datasheet, Analog Devices, 2015. URL:
adxl335.pdf

[7] Temperature probes: How to choose the right temperature sensor type, Omega Inc. 2018.
URL: thermocouple-RTD.html