Fundamentals of measurement and automation technology 1 (WS 2019/20)

1st Exercise: 44. and 45. CW

Signal flow diagrams. Static characterization of the transmission elements.
Linearization of characteristic curves.

Signal flow diagrams

Problem 1B: spring balance

A spring balance is loaded with a mass m. The spring balance itself is weightless. In the stationary
case, the deflection s is proportional to the weight of the mass. Develop the signal flow diagram
for the description of the spring balance measuring system!

input signal: xe = ∆m
output signal: xa = ∆s

m - weight of mass
g - acceleration of gravity
s - deflection
c - spring constant

Problem 2: level meter

The change of the level h causes a change of the level of the float 1 and thus an adjustment s1 at
the potentiometer 2. With the adjustment of the potentiometer 2 the voltage U changes. The
higher the voltage, the greater the displacement s2 of the core of the plunger coil. This core is
connected to a display 4 via a lever. We are looking for the signal flow diagram for the relation
∆α = f(∆h) .

input signal: xe = ∆h
output signal: xa = ∆α

Problem 3L: centrifugal tachometer

The masses m ("centrifugal weights") rotate with the speed ω around the vertical axis. As a result
of the centrifugal force Fω, the masses m move radially outwards until the centrifugal force Fω is
compensated by the spring force FF. The movement of the centrifugal weights m is transmitted by
the linkage 2 to the sleeve 3 and manifests itself in an upward movement of the pointer 5 around

1
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
1st Exercise: 44. and 45. CW
Signal flow diagrams. Static characterization of the transmission elements.
Linearization of characteristic curves.

the distance yr. For the creation of the signal flow diagram, the change in radius can be neglected
when calculating the centrifugal forces. Develop the signal flow diagram for this system.

input signal: xe = ∆ω
output signal: xa = ∆yr
4
Fc Fω
1 - centrifugal weights
2 - linkage m m
3 - sleeve
1
4 - spring 2
5 - pointer yr
3 r

5
n
Problem 4: self-compensating voltage compensation
With self-compensating voltage compensation, the voltage difference ∆U = Ux − UK is not dis-
played, but is transmitted to the motor of a motor potentiometer via an amplifier. This adjusts
the slider until the adjustment is reached, i.e. ∆U = 0 .The motor no longer receives current and
stops. The motor is coupled with a pointer in front of a scale, where the measured value can be
read. The input signal is U x , the output signal is ∆s. Develop the signal flow diagram for this sys-
tem.

Static characterization of the transmission elements. Linearization of characteristic

curves.

Problem 5B: glow lamp

A glow lamp is a gas discharge tube for generating a weak, so-called glow light. The glass bulb of a
glow lamp is usually filled with the noble gas neon, possibly with admixtures of helium, at low gas
pressure. An orange-red light colour is typical for glow lamps, which are mainly used as signal

2
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
1st Exercise: 44. and 45. CW
Signal flow diagrams. Static characterization of the transmission elements.
Linearization of characteristic curves.

lamps in mains-operated electrical appliances to indicate the operating state.

To start the glow discharge (ignition), an electrical voltage of up to 80 to 200 V is required. If this
ignition voltage is reached, the glow discharge begins and a current flows through the glow dis-
charge. As a result, the voltage drops to the combustion voltage. If the voltage falls below the hold-
ing voltage, the lamp goes out. To prevent thermal destruction of the glow lamp by an arc dis-
charge, a series resistor ensures that the operating voltage lies in the linear part of the character-
istic curve. The U-I characteristic can be determined by the following relationship

U(I) = U0 + U1 ⋅ Ib ⋅ e −cI + U2edI

V
be approximated. The values for the constant parameters are U0 = 180 V , U1 = 2 and
mA1,1
1
U2 = 1 V , the values for the constants in the exponents are b = 1,1 , c = 0,07 and
mA
1
d = 0,035 .
mA

a) Draw the static characteristic curve of the glow lamp U = f(I) for I = 0  100 mA .
b) Determine for the three work points I1 = 20 mA , I2 = 35 mA , and I3 = 70 mA the
respective sensitivities via a tangent linearization.
c) For the three work points from subtask b), determine the respective sensitivities via a secant
linearization with ∆I = ±10 mA .
d) Analyse the results at the three work points.

Problem 6L: resistance change

According to DIN EN 60751, the change in resistance of a Pt-100 measuring resistor is:

ϑ /°C 100 200 300 400 500 600 700 800

R ϑ /Ω 138,51 175,86 212,05 247,09 280,98 313,71 345,28 375,70
1 1
and R ϑ = R 0 ⋅ (1 + 3,9083 ⋅ 10 −3 ⋅ ϑ − 0,5775 ⋅ 10 − 6 ⋅ ϑ2 ) with R 0 = 100 Ω bei ϑ = 0°C ,
°C (°C)2

respectively.
a) Draw the static characteristic curve. Is it linear, progressive or degressive?
b) Calculate the temperature for R ϑ = 244 Ω by linear interpolation using the table and using
the specified equation (to two decimal places).
c) How high is the sensitivity (accurate to four decimal places) due to the formation of the
∆R ϑ
difference quotient ESek = (secant linearization) around the operating point
∆ϑ
R ϑ = 244 Ω ? R ϑ1 = f(ϑ1) and R ϑ2 = f(ϑ2 ) with ∆R ϑ = R ϑ2 − R ϑ1 = f(∆ϑ) are resistors
from the table that includes the operating point.

3
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
1st Exercise: 44. and 45. CW
Signal flow diagrams. Static characterization of the transmission elements.
Linearization of characteristic curves.

d) How high is the sensitivity (to four decimal places exactly) due to differentiation
dR ϑ
of the static characteristic curve with respect to ϑ as E Tan = at the operation point
dϑ AP
R ϑ = 244 Ω (tangent linearization)?

Problem 7: water boiler

In a water boiler, the water flowing in at 18 °C is heated depending on the valve opening. With a
working range of 0 to 15 cm2 for the cross-sectional area A of the valve, the following results are
obtained in accordance with the characteristic curve
1
ϑ = ϑ0 + a b ⋅ A a = 6 °C b= cm − 2
2
for the outlet temperatures between ϑ 0 = 18 °C and ϑmax (A max = 15 cm2 ) ≅ 34,43 °C .
a) Determine the transfer factors (sensitivities) for tangent linearization at the operating point
A 0 = 2 cm2 and with secant linearization in the range of A = 1  3 cm2 .
b) Determine the transmission factors for both types of linearization at
A 0 = 12 cm2 and in the range A = 11  13 cm2 , respectively.
c) In both cases, determine the percentage differences ∆E related to the transfer factor for
tangent linearization and discuss the result.
Note: For tasks a) and b), an accuracy of two decimal places is sufficient.

Problem 8H: SI units

The volume flow rate at the discharge from a large vessel with the filling level h can be calculated
from the following approximation

V = α ⋅ π ⋅ d2 ⋅ 2 ⋅ g ⋅ h .
4 V
With the given values
- fill level h = 1,2 m
- flow coefficient α = 0,9
- inner diameter of the valve dV = 20 mm
- acceleration of gravity g = 9,81 m/s²
determine the flow rate
a) in SI units
b) in l/s
c) in m³/h

Problem 9H: differentiation and fractional arithmetics

x1
a) Calculate all partial derivatives of the expression y = .
x 14 − x 24
b) Calculate the partial derivative ∂y / ∂x 1 and ∂y / ∂t of the expression y = x12 x2 (t) .
c) Calculate the derivative ∂y / dα of the expression y = cos2 (2α) .

4
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
1st Exercise: 44. and 45. CW
Signal flow diagrams. Static characterization of the transmission elements.
Linearization of characteristic curves.

1  Ua Ue − Ua 
d) Simplify the expression y =  −  by main denominator formation.
C  R 2 R1 
1 Ua U
+ a
C R 2 CR 1
e) Simplify the expression y = .
R1
+1
R2

____________________________________________________________________________________________________
Note: Problems with B are example tasks, problems with H are to be solved as homework
and problems with L are tasks for the learning group letter.

5
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
2nd Exercise: 46. und 47. KW
Signal flow diagrams. Current and voltage measurement.
Wheatstone bridge.

Signal flow diagrams

Problem 1L: spring pendulum

A spring pendulum or spring oscillator is a harmonic oscillator consisting of a coil spring and a
mass m attached to it. While the mass causes an expansion of the spring, the tension of the spring
(characterized by the spring constant k) causes its contraction. When the mass deflected from its
rest position is released, a harmonic oscillation begins, the amplitude of which decreases with time
due to the presence of damping (with the damping constant γ). The input variable is the resulting
force acting on the mass m, the output variable is the position x of the mass which changes with
time. Draw the signal flow diagram for this system and characterize each transmission element
with a formula or mathematical operation. Note: The effect of gravity can be neglected as it only
affects the position of the rest position.

Current and voltage measurement

Problem 2: reference resistance for current supply

An unknown resistance Rx is to be measured with the aid of a reference resistor Rr. For this pur-
pose, a current-feeding circuit is to be used, whereby only voltmeters are available as measuring
instruments.
a) Sketch the necessary circuit.
b) Which two conditions must be fulfilled by the internal resistors RMV of the voltmeters?
c) How is the unknown resistance Rx calculated?

Problem 3H: four resistors

In order to simplify the manufacture of an electronic assembly, the resistor network shown in the
figure is to be replaced by a substitute resistor RE.

1
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
2nd Exercise: 46. und 47. KW
Signal flow diagrams. Current and voltage measurement.
Wheatstone bridge.
a) Calculate the substitute resistance RE in case Ri=5 Ω for i=1...4.
b) For two modifications, Ri=5 Ω (i=1, 3, 4) and R2=10 Ω or Ri=5 Ω (i=1, 2, 3) and R4=10 Ω, the net-
work value is to be replaced by a substitute resistor RE=4 Ω. For which of the two resistor as-
signments is this possible?

Problem 4B: two-terminal device

A circuit is given with two connections A and B (active two-terminal device), see figure a).

figure a) figure b) figure c)

a) How large is the change in voltage across terminals A and B by connecting a voltmeter with
resistor RM?
b) How large must the resistance RM be so that the voltage change via terminals A and B is < 5 %
due to the connection of the voltmeter?

Note: The active two-terminal device can be substituted by a series connection of a voltage source
UL and a resistor Ri (the internal resistance), [equivalent circuit diagram, see figure b)]. To calculate
the internal resistance Ri of the two-terminal device, the voltage source E is short-circuited [figure
c)].

Problem 5: two-terminal device with ammeter

The current I in the circuit (see figure) is to be measured by means of a current measuring device
with an internal resistance RM, which is connected between terminals A and B. What percentage
error is caused by the connection of the ammeter? (R1 = 5 Ω, R2 = 7,5 Ω, R3 = 7 Ω, RM = 0,5 Ω). Take
account the note from problem 4B.

2
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
2nd Exercise: 46. und 47. KW
Signal flow diagrams. Current and voltage measurement.
Wheatstone bridge.
Wheatstone bridge

Problem 6B: Wheatstone bridge with constant current

The bridge shown in the picture is fed with a constant current I0.

a) Determine Ud = f(I0) .
b) What are the differences to a bridge fed with constant voltage?

Aufgabe 7L: Wheatstone bridge with variable R3

A variable resistor R3 = R0⋅α is connected in a Wheatstone bridge as follows:

R1= 200 Ω, R2 = 1 kΩ, R4 = 600 Ω, U0 = 10V, α - change factor

a) Calculate the nominal resistance R0 of R3 if the bridge is calibrated for case α= 1.

b) Calculate α and determine the change in % if the differential voltage Ud = 3 mV is set for the
supply voltage U0.

3
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
2nd Exercise: 46. und 47. KW
Signal flow diagrams. Current and voltage measurement.
Wheatstone bridge.
Aufgabe 8: Wheatstone-Brücke with variable R1, R4
In a Wheatstone bridge, whose 4 resistors are originally all equal to a basic resistance R 0 , the
resistance R 1 is increased by a small amount, R 1 → R 1 + ∆R , and the resistance R 4 is decreased
by the same amount, R 4 → R 4 − ∆R . It applies: ∆R << (R 1 , R 4 ) . Determine the diagonal voltage
Ud as a function of ∆R . Note: Use the linearization of expressions of shape (1 ± x )
−1
≈ 1  x when
x << 1 is.

____________________________________________________________________________________
Note: Problems with B are example tasks, problems with H are to be solved as homework
and problems with L are tasks for the learning group letter.

4
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
3rd Exercise: 48. und 49. KW
Bridge circuits. Error calculation.

Bridge circuits

Problem 1L: Bridge circuit in light meter

A light meter in a camera contains two similar LSDs (light sensitive device) and two similar fixed
resistors. The camera is powered by a miniature cell with a voltage of 1.5 V, the exposure time is
fixed to tBelich = 1 / 250 s , a film with ISO light sensitivity 100 is inserted. The LSDs have a linear
characteristic and a resistance of R LSD = 100 kΩ at bright sunlight (illuminance Ehell = 75600 Lux
, aperture 11). The following relationship exists between the aperture value k, the exposure time
tBelich , the illuminance E and the ISO light sensitivity Si

k = {tBelich } {E} ⋅ Si
C

with a typical value of C = 250 for this constant. {E} denotes the numerical value of the illuminance
when E is given in lux; {tBelich } the numerical value of the exposure time in seconds.
a) How is the bridge in the light meter to be connected if a maximum signal is to be generated
with the help of the LSDs?
b) What are the fixed resistances when the bridge is balanced in bright sunlight, i.e. when Ud = 0
and the light meter pointer is horizontal, respectively?
c) When the sky is cloudy, the pointer of the light meter falls to the lower stop point, which cor-
responds to a diagonal voltage of Ud = 0,15 V . When opening about two steps with respect to
the aperture, i.e. opening to an aperture of 5.6, the bridge is balanced again. Determine the
sensitivity of the LSDs in Ω / Lux .

Problem 2: Four-capacity bridge

The four capacitance bridge is used to measure small capacitances and their loss angle. The loss
angle is defined as the arc tangent of the ratio between active power and reactive power. C x is
the capacitance to be measured and tan δ x the loss angle to be determined. The unknown capac-
itance can be understood as the parallel connection of a loss-free capacitor C 3 = C x and a resistor
R 3 = 1 /(ωC 3 tan δ x ) . The resistor R 1 and the capacitor C 2 are adjustable.

1
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
3rd Exercise: 48. und 49. KW
Bridge circuits. Error calculation.

a) What expressions are obtained for C x and tan δ x when starting with the general adjustment
condition Z1 ⋅ Z 4 = Z 2 ⋅ Z 3 for AC bridges?
b) As an alternative, choose the matching conditions from the lecture!

Problem 3H: frequency measuring bridge

The following circuit allows to determine the frequency of the supply voltage.

a) Consider which adjustment condition you start with.

b) Determine three meaningful relations for the components of the brigde from the matching
condition in order to determine the frequency f with capacitances and resistances that are as
similar as possible.

Note: To solve the following tasks, also use the "Error Calculation Guide" available in the
network.

Error calculation

Problem 4B: error calculation for resistance calculation

The value of a resistance is to be determined from a current and voltage measurement. A class 1
measuring instrument with a final value of 5 A is available for current measurement. The initial
value is zero ampere. A current I=4 A is measured. The voltage is measured with a measuring in-
strument of class 0.5 with a measuring range of 100 V. The measured voltage is U=45 V.
Determine the
- relative
- maximum relative
- probable relative
- maximum absolute and the
- probable absolute
deviation for the resistance.

2
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
3rd Exercise: 48. und 49. KW
Bridge circuits. Error calculation.

Problem 5: pressure transmitters

According to the manufacturer's specifications, a pressure transmitter should map the pressure
range p=(1 ... 10) bar linearly to a standard current signal I=(4 ... 20) mA. When testing with a piston
pressure gauge, the following pairs of values were determined (note the units of measurement!):

p kp/ cm² 1 2 3 4 5 6 7 8 9 10
I mA 4,15 5,9 7,35 9,50 11,25 12,70 14,75 16,20 18,20 19,80

a) Calculate the transmission factor a resulting from the manufacturer's specifications.

b) To which error class shall the pressure transmitter be assigned as a result of the test?
c) Determine the error limits of the pressure transmitter and display the error curve ∆I=f(pist)
graphically.

Note: permitted error classes are: 0,01; … 1,5; 1,6; 2,5; 4; … 10;

Problem 6: Pt100 resistance thermometer

The measuring resistance of a Pt100 resistance thermometer was measured 8 times with a Wheat-
stone bridge: Rϑ/Ω=125,8; 124,3; 126,9; 123,5; 125,7; 127,1; 125,3; 125,1. Calculate
a) the probably correct value of the resistance R .
b) the mean error mR (confidence interval of the mean value).
c) the measurement uncertainty of the resistance with P=95%? (t=2,36).
d) the temperature ϑ of the measurement sensor. (The sensor is characterized by the mean tem-
R − R O°C
perature coefficient α = ϑ = 38,5 ⋅ 10 − 4 / °C .)
R O°C ⋅ ∆ϑ
~ .
e) the uncertainty of temperature measurement m ϑ

Problem 7L: output of a appliance

During a measurement on a appliance whose resistance is known with an accuracy of 3%, the
current and the voltage drop were measured in order to determine the power. The following are
given:

devices measurement value measuring range error class

voltmeter 72 V 100 V 1,5
ammeter 470 mA 1A 2

Determine the maximum relative measurement deviation for the calculation of the power, ignor-
ing circuit-related errors
a) from voltage and resistance.
b) from voltage and current.
c) Now the measurement deviations by amount and sign are known. The systematic measure-
ment deviations are: ΔU/U=-0,013, ΔI/I=0,034 and ΔR/R=-0,024. What is the maximum relative
measurement error of the power when determining from voltage and resistance?

3
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
3rd Exercise: 48. und 49. KW
Bridge circuits. Error calculation.

d) What is the maximum relative measurement error of the power for these known measurement
deviations for the determination from current and resistance? Also now the measurement de-
viations according to their sign are to be considered again.

Problem 8: capillary viscometer

To determine the viscosity η of a liquid using a capillary viscometer, the pressure drop ∆p along a
capillary through which the measuring fluid flows is used. The Hagen-Poiseuille law applies to a
formed laminar pipe flow, which leads to the following correlation:

πR 4 ∆p .
η =
8q V L

R is the radius of the capillary, L its length, qV the volume flow and ∆p the difference in pressure
between the inlet and the outlet of the capillary. Given are ∆p = (15,2 ± 0,3) kPa ,
−6
R = (3 ± 0,05) mm , L = (15 ± 0,5) mm und q V = (40 ± 3) ⋅ 10 m3 / s . Calculate the probable
relative error ∆ηrel, wahr for the measured viscosity.

____________________________________________________________________________________
Note: Tasks with B are example tasks, tasks with H are to be solved as homework
and tasks with L are tasks for the learning group letter.

4
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
4th Exercise: 50. und 51. KW
Error calculation. Dynamic characterization of zero order transmission elements.

Error calculation

Note: To solve the following tasks, also use the "Error Calculation Guide" available in the
network.

Problem 1L: water bath for chocolate coating

With the help of a water bath, chocolate coating (for decorating Christmas cookies) is to be kept at
a temperature of about 40°C. The temperature of the water bath is measured with a type K ther-
mocouple, which has an error of 2.5 K. The thermocouple is directly connected to a reference junc-
tion with a temperature of 0°C. This operating mode causes an error of 0.4% of the measuring
range, which is 100 K. The following measured values have been recorded.

number of measurement 1 2 3 4 5 6 7 8
ϑTE, i in °C 43,4 41,9 45,2 42,7 43,7 44,6 42,3 43,6

a) Determine the mean temperature ϑTE for the measurement with the thermocouple and the
associated absolute maximum error. Specify the result in the form ϑ = ϑTE + ∆ϑ TE,abs,max .
b) To reduce the uncertainty of measurement for ϑ an additional measurement is performed
with a mercury thermometer (measuring range from 20°C to 70°C) of error class 1. The
measured value is ϑHg = 43,1 °C . Which weighted mean value ϑres and which associated er-
ror ∆ϑres result for the temperature ϑ ? Specify the result in the form ϑ = ϑres + ∆ϑres .

Problem 2: relative probable error

Calculate the relative probable error of the following quantities
x 12
a) y 1 =
2x 2

b) y 2 = x 1 + x 2
3
x3
x + 2x 2
c) y 3 = 1
4x 2
n 2
 ∂y ∆xi 
first using the general expression ∆yrel,prob = ∑ 
i =1 
⋅ 
∂xi y 
from the Gaussian law of error

propagation and then using the rules for special cases, if possible.

Problem 3B: best fit line

During the calibration of a measuring system, the following measured values of the output variable
y were determined as a function of the input variable x:

x 1 2 3 4 5 6 7 8 9 10
y 0,67 2,23 3,46 5,37 4,35 5,93 6,00 9,22 8,89 9,32

1
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
4th Exercise: 50. und 51. KW
Error calculation. Dynamic characterization of zero order transmission elements.

The functional relationship is to be described with a compensation straight line y = ax + b. The

coefficients a and b, their errors and the correlation coefficient r are to be determined!

Problem 4L: radiation power measurement

When measuring the radiant power of a test specimen, the following measured values for the ra-
diant power P were determined as a function of the temperature T:

T in K 275 300 325 350 375 400 425 450 475 500

P in W 0,1028 0,0865 0,1373 0,31 0,3316 0,289 0,5644 0,5701 0,7676 0,8975

The functional connection between T and P is given by P(t) = p1T 4 + p0 .

a) Transform the measured values appropriately so that you can subsequently work with a linear
regression.
b) For the linear regression line, determine the slope, the axis intersection, and their errors.
c) Use the results from b) to determine the coefficients p0 and p1 of the functional relationship.

Dynamic characterization of zero order transmission elements

Note: To solve the following tasks, also use the "Static and dynamic characterization of
measuring elements" script available in the network.

Problem 5B: rotor turning mass

A drive torque MA and a braking torque MB act on a rotor turning mass. The output signal is the
change in the speed of the shaft. The input signal represents the difference between the two tor-
ques. Determine the type and characteristic values of the transmission element.

2
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
4th Exercise: 50. und 51. KW
Error calculation. Dynamic characterization of zero order transmission elements.

Problem 6: piezoelectric force transducer

A piece of quartz cut out of a single crystal is used as a piezoelectric force transducer. The upper
and lower cut surface is metallized and serves as an electrode. If a force acts on the quartz disc,
charges are induced, whereby the amount of charge increases linearly with the applied force with
the proportionality factor k. The quartz can be understood as a capacitor that stores charges,
which makes it possible to tap a voltage, see picture. The quartz block has the base area A, the
thickness d and the relative dielectric constant εr . Determine the type of transmission element
when the input signal is the change of the applied force ∆F and the output signal is the resulting
voltage change ∆U .

Problem 7: running rope

The speed of a pulley is determined by the speed of a running rope. The input variable is the dis-
placement of the rope, the output variable is the resulting change in the angular speed of the
pulley ∆ω. The search is for ∆ω = f(∆s). Which transmission element with which characteristic values
does this correspond to?

3
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
4th Exercise: 50. und 51. KW
Error calculation. Dynamic characterization of zero order transmission elements.

Problem 8: spring-damper-lever-system
Given is the shown mechanical system of a spring-damper-rod-system. The displacement change
of the suspension point ∆s is the input signal xe, which via a lever (with lever arms a and b) should
lead to the change of the force F, so that the lever remains in equilibrium. This is the output signal:
x a = ∆F . Assume a force balance to determine the type and characteristic quantities of the trans-
mission element. Consider the relationships for the spring force Fc and the damping force FD. The
spring constant c and the damping constant k are given.

____________________________________________________________________________________
Note: Tasks with B are example tasks, tasks with H are to be solved as homework
and tasks with L are tasks for the learning group letter.

4
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
5th Exercise: 2. und 3. KW
Dynamic characterization of zero and first order transmission elements.

Note: For the solution of the following tasks, also use the script "Static and dynamic char-
acterization of measuring elements" available in the network

Problem 1L: Direct gravimetric particle mass sensor

The measuring principle of a direct gravimetric particle mass sensor follows the principle of a
spring pendulum clamped on one side and excited by an electric motor. The spring element is
formed by a hollow metal tube with the outer diameter R and the inner diameter r. This spring
element is magnetically excited by an actuator coil to natural frequency oscillations with fSystem . A
sensor coil registers the oscillations of the spring pendulum. At the free end of the pendulum there
is a replaceable filter cartridge in which the particles flowing in through the sampling hose (with
upstream filter) are deposited. As a result, the effective total particle mass mp is deposited in the
filter cartridge over time, causing the natural frequency of the spring pendulum to change to fw .

a) Consider the tube, clamped at one end, as a bending beam of the length L , at the end of which
a force F is applied across the beam. Determine the natural frequency of the spring pendulum
fSystem , i.e. the frequency of the oscillating system of the mass msystem without deposition of
particles. Notes: The modulus of elasticity E and the moment of inertia of area I are known.
The deflection of the beam can be interpreted as its deflection within the framework of the
model of a spring pendulum.
b) How is the accumulated particulate mass mp determined when the associated frequency fw
is measured. Note: For the determination of the moment of inertia per unit area, the cut sur-
face through the tube is considered to be an annular surface.
c) How does the frequency of deposition ( x a = ∆fw ) change in linear approximation depending
on the mass of particles ( x e = ∆mp )? The working point is the natural frequency of the spring
pendulum. Which type of transmission element results with which characteristics?

1
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
5th Exercise: 2. und 3. KW
Dynamic characterization of zero and first order transmission elements.

Problem 2: Step and rise response PIT0 element

From the differential equation of a PIT0-element determine the step and rise response and sketch
their progressions.

Problem 3B: room heating

A room heating system is considered, which should heat the air in a closed room, see picture. The
 of the heating system is used as the input variable and the output variable forms
heat flow Q zu
the temperature difference ∆ϑ= ϑ - ϑa. Set up the difference equation for the system using a bal-
ance equation. Which type of transfer element results with which characteristics? Note that the
room loses heat to the outside (constant outside temperature ϑa) via the walls.


Q zu - heat flow
ϑ - room temperature
ϑ a - outside temperatuere (constant)
m - room air mass
cp - specific heat
k - heat transfer coefficient
A - entire wall surface

Problem 4L: flow with pressure jump

A liquid flow with a flow velocity c0 shall be accelerated in a pipeline of diameter D and length L
(rusty cast iron pipe with λ = 0,06) by a pressure jump ∆p = 100 Pa between inlet and outlet.
a) Set up the differential equation for determining the change in velocity ∆c = f(∆p) using the
equilibrium of forces. Assuming a laminar flow, i.e. for the Reynolds numbers smaller than a
64 ⋅ ν
critical value of 2320, holds λ = . For the frictional force holds FR = U ⋅ L ⋅ τ and for the
c ⋅D
λ ⋅ ρ ⋅ c2
acceleration force FB = m ⋅ ∆c . The wall shear stress is τ = . The following parame-
8
ters are given: D = 3 cm, L = 100 m, ρ = 998 kg/m3, ν = 1,01⋅10-6 m²/s and c0 = 5 cm/s.
b) Determine the type of the transmission element. Calculate the characteristic values of the
differential equation. What speed is achieved in the steady state? Note: Use the DGL solu-
tions from the "Characterization of Measuring Elements".

2
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
5th Exercise: 2. und 3. KW
Dynamic characterization of zero and first order transmission elements.

Problem 5: gas pressure gauge

The picture sketched the function of a gas pressure gauge. The gas pressure p(t) acts on the piston
surface A, causing the piston-rod system to be pressed down [s(t)] until the friction (friction con-
stant r) caused by the liquid-filled cylinder-piston system brings the movement to a standstill. The
movable linkage system has the total mass m, whereby the weight force m-g is neglected for the
analysis. The input variable is the pressure p, the output variable is the change of movement over
time, v(t)= s (t). Set up the corresponding differential equation for the system. Specify the type of
transmission element and the corresponding characteristic values.

Aufgabe 6: pressure vessels

In the case of the gas-filled pressure vessel shown in the figure, the internal pressure p shall be
represented as a function of the position y of the inlet valve when isothermal changes of state and
an ideal gas are present. It is given: m
 zu = a ⋅ y , m
 ab = b ⋅ p − p a , pe=const., pa=const.; b is a
valve constant. What kind of transmission element is present and what are its characteristic val-
ues? Note: The necessary linearization is performed at the operating point (p 0 , p 0 ) .

3
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
5th Exercise: 2. und 3. KW
Dynamic characterization of zero and first order transmission elements.

Problem 7H: CR high pass

Given is the CR high pass shown in the picture. The transmission behavior of this network with the
input voltage Ue and the output voltage Ua is sought.

Ue R Ua

∆Ue ∆Ua

a) Describe the differential equation of this network using the mesh sets (balance equations) and
the relationships between the quantities Ue, Ua, R and C.
b) Determine the characteristic values of the differential equation (time constant and transfer
factor) if R=100 Ω und C= 10 µF. Note: 1 F=1 (A·s)/V.

Problem 8: piezoelectric pressure transducer

For a piezoelectric pressure transducer the following equivalent circuit diagram can be used.

dQL
IL =
dt

dUC
Ic = C
dt

UR
IR =
R

a) Determine the differential equation of the transmission element under the assumption that
the voltage measurement takes place without current consumption.
b) Determine the transmission behaviour of the pressure sensor and the characteristic values of
the transmission element.
c) Specify the equation for the transition function and sketch its progression. How are the time
constant and the transfer factor determined graphically?

4
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
5th Exercise: 2. und 3. KW
Dynamic characterization of zero and first order transmission elements.

Problem 9H: arithmetic operations using complex numbers

Convert the following complex numbers so that the denominator becomes real. Present the result
as z = Re(z) + i ⋅ Im(z) . Also calculate the absolute value of these complex numbers and summa-
rize the result as far as possible.

i⋅c D E
a) z1 = b) z 2 = a + i ⋅ c c) z 3 = 2
d) z 4 =
1+i⋅b 1+i⋅b 1− a +i⋅b id(1 − a2 + i ⋅ b)

____________________________________________________________________________________
Note: Tasks with B are example tasks, tasks with H are to be solved as homework
and tasks with L are tasks for the learning group letter.

5
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
6th Exercise: 4. und 5. KW
Dynamic characterization of first order transmission elements. Laplace transfor-
mation. Digital image processing.

Problem 1L: Pneumatic displacement sensor

The nozzle flapper system is a path-pressure converter s=f(p). Depending on the distance s of the
baffle plate from the nozzle d2, the pressure p is set in the working volume (space between the
nozzles d1 and d2). Even small changes of the distance s between nozzle and baffle plate lead to a
considerable change of the pressure p. Derive the differential equations for the description of the
dynamic behaviour of this system under isothermal state changes in the working point (p0, s0)!
What is the transmission behaviour and how are the characteristic values calculated?

V – working volume
m – mass in the working volume
V 1
=
m ρ0
pV – constant prepressure
pamb - constant ambient pressure
m
 out = a ⋅ s ⋅ p − pam b

m
 in = b ⋅ pv − p

Problem 2B: IT0 element with shock or jump as input

Write down the differential equation for an IT0 element in the time domain. Perform a Laplace
transformation and use the result to determine the transfer function G(s)= Xa(s)/Xe(s). For the input
signal Xe(s), set a shock and then a jump. Which expressions result for Xa(s)? Into which results xa(t)
are these results transformed back into the time domain?

Problem 3: R1-C-R2 network

Examine the behavior of the network shown in the picture. The voltage Ue is the input signal and
voltage Ua is the output signal.
a) Set up the differential equation of the network. What type of transmission element is present?
b) Determine the characteristic values of the transmission element. Assume R 1 = 1 MΩ ,
R 2 = 3 MΩ and C = 1 µF .
c) Specify the equation for the transition function and sketch its graph. How to determine the
time constant and the transfer factor graphically.

1
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
6th Exercise: 4. und 5. KW
Dynamic characterization of first order transmission elements. Laplace transfor-
mation. Digital image processing.

Problem 4: R1-C-R2 network in frequency space

Given is the normal form of the differential equation of the transfer element from problem 3, see
also summary of the results.
a) What is the transfer function G(s) in the frequency domain?
b) Determine G(s) for the case of harmonic excitation, that means for x e = x̂ e ei ω t !
c) Display G(s) via amplitude and phase response (also called Bode diagram) with the values
from problem 3!
d) What is the limit value of the amount of G for ω→∞? What is the limit value of the phase for
ω→∞ and ω→0?
e) Represent G(s) from subtask b) using the Nyquist plot (also called frequency response locus
or locus curve).

Problem 5L: DIT1 element in frequency space

Given is the normal form of the differential equation of a DIT1 transfer element

∫
T1 x a + x a = K I x e dt + K D x e .
a) What is the transfer function G(s) in the frequency domain?
b) Determine G(s) for the case of harmonic excitation, that means for x e = x̂ e ei ωt !
c) Display G(s) via the locus curve! Which values do the real and imaginary parts of G take at
ω → ∞, at ω→0 and at ω = 1 / T1 ? Note: It is KD=2 s, T1=1 s and KI=4 s-1.
d) Create the amplitude diagram for G(s) from subtask c). With which power α approaches
| G(ω) |~ ωα the limit value for small ω ?

Problem 6: complex circuit

The transmission behavior of a complex circuit with a current input signal I and a voltage output
signal U shall be analyzed.

a) Determine the type of the individual transmission elements from the circuit diagram.
b) Formulate the corresponding differential equation in the time domain and the corresponding
algebraic equation in the frequency domain
c) Determine the differential equation in the time domain, the type of transmission element and
the transfer function for the overall transfer behaviour.

2
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
6th Exercise: 4. und 5. KW
Dynamic characterization of first order transmission elements. Laplace transfor-
mation. Digital image processing.
Problem 7H: Decomposition of complex transfer function
The following transfer function

A(B + T1s)
G(s) =
s + 7s2 + 10s
3

is given which results from the interconnection of individual transmission links. Break down this
transfer function into individual simple transmission elements (IT0, DT0, PT1) and draw the signal
flow diagram with the simple individual elements. Note: Divide the denominator of G(s) into linear
factors.

Digital image processing

Problem 8B: characterisation homogeneous image

A digital black and white image is described with the following mathematical model:

G = {0, Gmax } grey value set consisting of the grey values from 0 to Gmax ,

S = [s(x, y)] image matrix of the digitized image with L image lines and R image columns:
0 ≤ x ≤ L − 1, 0 ≤ y ≤ R − 1 ,

(x, y) spatial coordinates of the pixel at the position (row, column)=(x,y),

s(x, y) = g ∈ G Gray value of the pixel at the position (x,y).

The average gray value of the image S, also called average value of S, is calculated according to

L −1 R −1
1
mS = ∑∑
M x =0 y =0
s(x, y) ,

where the number of pixels of S is M = L ⋅ R . A parameter that allows a statement about the con-
trast in the image is the mean square deviation

L −1 R −1
1
qS = ∑∑ [s(x, y) − mS ]2 .
M x =0 y =0

a) Reshape qS in such a way that you arrive at an easier calculation.

b) Determine mS and qS for a homogeneous image with the grey value 127.

3
 Fundamentals of measurement and automation technology 1 (WS 2019/20)
6th Exercise: 4. und 5. KW
Dynamic characterization of first order transmission elements. Laplace transfor-
mation. Digital image processing.

Problem 9: characterisation checkerboard pattern image

Characterize a checkerboard pattern image (alternating black and white, with both colours occu-
pying only one pixel each) that contains only the gray values 0 and 254. For characterization, use
the same method as in problem 8 the mean value (a) and the mean square deviation (b).

____________________________________________________________________________________
Note: Tasks with B are example tasks, tasks with H are to be solved as homework
and tasks with L are tasks for the learning group letter.

4
Sample solution

Problem 1-1B: spring balance

The aim is to create a signal flow plan for a spring balance.

The input signal is the mass change, xe = ∆m , the output
signal is the pointer position change xa = ∆s . There are two
possible forms of interpretation for the signal flow dia-
gram.
On the one hand, it can be represented as a simple chain,
whereby the mass placed on the shell leads to a mass
change ∆m . This produces a change in force ∆Fm on the
spring. This results in a change in length of the spring ∆s1
via Hook's law, which in turn moves the display around
∆s . So we get the measurement chain below as a signal
flow diagram.

In the second possible representation, the force change ∆Fm causes a change in the position of
the display, which leads to an elongation of the spring. This spring elongation leads to a back-
driving force, which determines the final deflection ∆s in a feedback via a subtractive mixing
point in a force equilibrium. In this case, the signal flow diagram contains the feedback men-
tioned above, which describes the equilibrium of forces.

Problem 1-5B: glimm lamp

Using the example of the nonlinear U-I characteristic curve of a glow lamp, the linearization of
characteristic curves is illustrated.

a)

In this subtask, the characteristic curve of the glow lamp U = f(I) is to be drawn for a value range
of the current from 0 to 100 mA . Note that due to the unit of the given constant, values for I
from 0 to 100 mA are used and not values from 0 to 0,1 A . The result is shown in the graph,
whereby the three characteristic points of the characteristic curve mentioned in the text are drawn.
 A: ignition of the glow lamp at
ignition voltage

B: holding voltage

C: upper end point of the interval for the

operating voltage

b)

In this subtask, a tangent linearization is to be performed at three different points of the charac-
teristic curve. Tangent linearization means the derivation of the characteristic curve, here after the
current I, and calculation of this derivation at the operating points I1 = 20 mA , I2 = 35 mA , and
I3 = 70 mA . The derivative from U to I provides

Etan =
dU
dI
( )
= U1e − cI bIb −1 − cIb + U2dedI .

For safety reasons the physical unit of the obtained expression can be checked, because the phys-
ical unit of Etan must be V/mA. Inserting the units results in

V  1,1 
[Etan(I)] =  mA 0,1 − mA  + V = V
mA1,1  mA  mA mA


and thus the correct physical unit of sensitivity. Inserting the numerical value for I1,

( )
Etan(I1 ) = U1e −cI1 bI1b −1 − cI1b + U2dedI1 ,

I2 and I3 yield

V V V
Etan(I1 ) ≅ −0,129 , Etan(I2 ) ≅ −0,213 und Etan(I3 ) ≅ +0,319 .
mA mA mA

c)

In this subtask, a secant linearization is to be performed at three different points of the character-
istic curve. Secant linearization means the formation of a difference quotient,

∆x a ∆U
Esek = = ,
∆x e ∆I

where in the task setting the value of the interval width for the difference formation is given by
∆I = ±10 mA . The working points are the same as in subtask b).
The graph shows the position of the intervals ∆U and ∆I for the first operating point.

The corresponding calculation results in

∆Ua1 U(I1 + ∆I) − U(I1 − ∆I) U(I1 + ∆I) − U(I1 + ∆I)

Esek (I1 ) = = =
∆Ie1 (I1 + ∆I) − (I1 − ∆I) 2∆I

U(I = 10 mA) − U(I = 30 mA)

=
20 mA

193,181579 − 193,922345 V V
Esek (I1 ) = ≅ −0,037 .
20 mA mA

It follows analog graphics and calculations for the second and third operating point.

190,474081 − 194,386929 V V
Esek (I2 ) = ≅ −0,196
20 mA mA
 197,361669 − 190,876109 V V
Esek (I3 ) = ≅ +0,324
20 mA mA

d)

For the first operating point, two very different results for the sensitivity result from the two line-
arizations:

V V
Etan(I1 ) ≅ −0,129 und Esek (I1 ) ≅ −0,037 .
mA mA

The reason is that the choice of the interval ∆I1 for the secant linearization is not meaningful, since
within this interval the characteristic curve U(I) undergoes a change of monotony.

For the second operating point, two results for the sensitivity with relatively good agreement result
from the two linearizations:

V V
Etan(I2 ) ≅ −0,213 und Esek (I2 ) ≅ −0,196 .
mA mA

Since the selected interval covers in a first approximation a linear part of the characteristic curve,
both values for the sensitivity correspond well. The selected interval is suitable for a secand linear-
ization.

For the third operating point, two results are obtained for the sensitivity with very good agreement
from the two linearizations:

V V
Etan(I3 ) ≅ +0,319 und Esek (I3 ) ≅ +0,324 .
mA mA

The selected interval ∆I3 covers another part of the characteristic curve, which can be regarded as
linear in a first approximation. Therefore, the two values for sensitivity match rather well. The se-
lected interval is suitable for a secand linearization. It can be selected relatively freely in this range
of the characteristic curve as long as the interval limits lie within this approximately linear range of
the characteristic curve.
Sample solution

Problem 2-4B: two-terminal device

The two-pole circuit below is considered, i.e. a circuit characterized by two outgoing terminals,
which must be regarded as a unit depending on the situation.

In order to determine the voltage change for task part (a) by connecting the measuring device, the
open circuit voltage at these terminals must first be calculated. This corresponds to the voltage UL
across the resistor R 2 (the resistors R 1 and R 2 are connected in series)

E = U1 + U2 = I(R1 + R 2 )

UL = U2

UL
I=
R2
R2
UL = E .
R1 + R 2

For further consideration, the total resistance of the two-terminal circuit - its internal resistance Ri
- is now important. To determine this, we look at the two-terminal device from terminals A
and B. The internal resistance can thus be calculated from the parallel connection of R1 and R2.

R 1R 2
Ri = .
R1 + R 2

The active two-terminal device can thus be replaced as a series connection of a voltage source UL
and a resistor Ri.

The previously calculated open circuit voltage generates a current

UL
I=
Ri + RM
in the circuit consisting of Ri and RM.

This must clearly (series connection) be equal to the current through the measuring instrument.

UM
I=
RM

One gets
RM
UM = UL .
Ri + RM

This means for the voltage difference

 RM  Ri
∆U = UL − UM = UL 1 −  = UL .
 Ri + RM  Ri + RM

The relative deviation of the voltage measurement is therefore

∆U Ri R
= ≈ i
UL Ri + RM RM

for R M >> R i . The demand for a deviation of a maximum of 5% thus leads to

Ri 1
≤ 0,05 = → R M ≥ 19 R i .
Ri + RM 20

Problem 2-6B: Wheatstone bridge

(a) constant current supply

Ud = f (I0 = const.) (1)

First the voltage U0 is calculated, which depends on the resistances at a constant current. The total
resistance results from the parallel connection of the voltage dividers given by R1 and R2 as well as
R3 and R4.

−1

U0 = I0 ⋅ 
1
+
1 
= I0 ⋅
(R1 + R 2 )(R 3 + R 4 ) (2)

 R1 + R 2 R 3 + R 4  R1 + R 2 + R 3 + R 4

For the diagonal voltage (left stitch) applies:

 R1 R3   R1R 4 − R 2R 3 
Ud = U1 − U3 = U0  −  = U0   . (3)
 R1 + R 2 R 3 + R 4   (R1 + R 2 )(R 3 + R 4 ) 

If inserting Eq. (2) in Eg. (3), one gets

 R1R 4 − R 2R 3 
Ud(I0 ) = I0   . (4)
 R1 + R 2 + R 3 + R 4 

(b) supply with constant voltage

Ud = f (U0 = const.) (5)

Here the expression derived in the lecture is valid for the diagonal voltage (the right stitch was
used):

 R4 R2   R1R 4 − R 2R 3 
Ud(U0 ) = U4 − U2 = U0  −  = U0   . (6)
R
 3 + R 4 R 1 + R 2 (
 1R + R 2 )(R 3 + R )
4 

Comparing equations (4) and (6), it becomes clear that the counters are the same and the denom-
inators are different.
Sample solution

Problem 3-4B: error calculation for resistance calculation

Given are:
ammeter: class 1, measuring range 5 A
measured current: 4,0 A
voltmeter: class 0,5, measuring range 100 V
measured voltage 45 V

In general, on the basis of the law of error propagation, the maximum absolute deviation for a
quantity y = f (xi , i = 1  n) is calculated by

n
∂f
∆y abs, max = ∑ ∂x
i =1 i
⋅ ∆xi .

For the probable absolute deviation, the following results then apply

n 2
 ∂f 
∆y abs, prob = ∑ i =1

∂
 ix
⋅ ∆xi 

.

If the corresponding relative deviations are to be determined, the corresponding absolute devia-
tions must be divided by the value of the variable, i.e.

n
∆y abs, max ∂f ∆xi
∆yrel, max =
y
= ∑ ∂x
i =1 i
⋅
y

and

n 2
∆y abs, prob  ∂f ∆xi 
∆yrel, prob =
y
= ∑
i =1
 ⋅ 
 ∂xi y 
.

Applied to the example task, the following calculations result.

Measurement error of the current measurement

class 1 means 1% of the measuring range, i.e. ∆I = ± 0,05 A.

With a measured current of I = 4.0 A, the relative error of the current measurement is therefore
0,05 A
∆Irel = ± = ± 0,0125 . (1)
4,0 A

Measurement error of the voltage measurement

class 0,5 means 0,5% of the measuring range, i.e. ∆U = ± 0,5 V.

With a measured voltage of U = 45 V, the relative error of the voltage measurement is therefore
0,5 V
∆Urel = ± = ± 0,0111 . (2)
45 V

1
Resistance from current and voltage

U 45 V
R = = = 11,25 Ω (3)
I 4,0 A

Maximum relative deviation of the resistance determination

∆R abs, max ∂R 1 ∂R 1 1 I U I ∆U ∆I
∆R rel, max = = ∆U ⋅ + ∆I ⋅ = ∆U + − 2 ∆I = + −
R ∂U R ∂I R IU I U U I
(4)
∆R abs, max
∆R rel, max = = ∆Urel + ∆Irel = 0,0111 + 0,0125 = 0,0236
R

Probable relative deviation of the resistance determination

according to Gaussian error propagation

2 2 2 2
∆R abs, prob  ∂R 1  ∂R 1 1 I   U I 
∆R rel, prob = =  ∆U  +  ∆I  =  ∆U  +  − 2 ∆I 
R  ∂U R  ∂I R I U   I U 
(5)
2 2
 ∆U   ∆I 
∆R rel, prob =   + −  = (∆Urel )2 + (∆Irel )2 = 0,01112 + 0,01252 = 0,0167
 U   I 

Maximum absolute deviation of the resistance determination

with (3) and (4)

∆R abs, max = R ⋅ ∆R rel, max = 11,25 ⋅ 0,0236 Ω ≈ 0,27 Ω (6)

Probable absolute deviation of the resistance determination

with (3) and (5)

∆R abs, prob = R ⋅ ∆R rel, prob = 11,25 ⋅ 0,0167 Ω ≈ 0,19 Ω (7)

2
Sample solution

Problem 4-3B: best fit line

If you calculate the best fit line "on foot", it makes sense to determine the required sums next:

xi yi xi2 yi ² xi⋅yi
1 0,67 1 0,0449 0,67
2 2,23 4 4,9729 4,46
3 3,46 9 11,9716 10,38
4 5,37 16 28,8369 21,48
5 4,35 25 18,9225 21,75
6 5,93 36 35,1649 35,58
7 6,00 49 36,0000 42,00
8 9,22 64 85,0084 73,76
9 8,89 81 79,0321 80,01
10 9,32 100 86,8624 93,20
[x]=55 [y]=55,44 [x2]=385 [y2]=387,2206 [x∙y]=383,29

Then you can determine the parameters a and b for the best fit line y = a ⋅ x + b using the formulas
from the lecture (or from the error calculation script) with

the slope a=
[x ⋅ y] ⋅ n − [x] ⋅ [y] = 1 (n ⋅ [x ⋅ y] − [x] ⋅ [y]) = 0,9499 ,
n ⋅ [x2 ] − [x]
2 ∆

where [ ]
∆ = n ⋅ x 2 − [x] ⋅ [x] = 825 , and

the axis intersection b=

[x ] ⋅ [y] − [x] ⋅ [x ⋅ y] = 1 ([x ] ⋅ [y] − [x] ⋅ [x ⋅ y]) = 0,3193
2
2 .
n ⋅ [x ] − [x]
2 2 ∆

These values can be used to calculate the y-values of the best fit line (yB) using xi. The difference
between the calculated (yB) and the measured (yi) values results in ∆yi and thus the sum of all
(∆yi)². One gets:

xi yi yB = a⋅xi + b ∆yi=yi-yB (∆yi)2

1 0,67 1,2693 -0,5992 3,59128e-01
2 2,23 2,2191 0,0100 1,16378e-04
3 3,46 3,1692 0,2908 8,45928e-02
4 5,37 4,1191 1,2509 1,56477e+00
5 4,35 5,0690 -0,7190 5,17005e-01
6 5,93 6,0190 -0,0890 7,91561e-03
7 6,00 6,9689 -0,9689 9,38785e-01
8 9,22 7,9188 1,3012 1,69300e+00
9 8,89 8,8688 0,0212 4,49954e-04
10 9,32 9,8187 -0,4987 2,48729e-01
∑ (∆y )i
2
= 5,4145

1
This data can then be used to determine the errors for the parameters (see script for error calcu-
lation).

1
(∆y )2 =
n−2
⋅ ∑ (∆y ) i
2
= 0,6768

error of the slope: (∆a)2 = (∆y )2 ⋅ n = 0,0082 ⇒ ∆a = 0,0906

∆

2
[ ]
error of the axis intersection: (∆b)2 = (∆y )2 ⋅ x = 0,3162 ⇒ ∆b = 0,5622
∆

(∆b)2 = (∆a)2 ⋅
[x ] 2

n
In addition, the correlation coefficient r

n ⋅ [x ⋅ y] − [x] ⋅ [y]
r = = 0,9655 mit [x]2 = [x]·[x]
(n[x ] − [x] ) ⋅ (n[y ] − [y] )
2 2 2 2

can be determined for the data set. As an illustration, the following graph shows the measured
values and the approximated functional relationship.

Problem 4-5B: rotor turning mass

balance equation:

For the balance equation, we first determine the difference between the effective torques that
accelerate the rotor mass:
 .
MA − MB = ∆M = J ⋅ ϕ (1)

coupling relationship:

Next we need a coupling between the output signal ∆n and the change of the angular velocity ϕ :
ω = ϕ = 2π ⋅ n (2)
ω
 =ϕ = 2π ⋅ n
(3)

2
With equation (3) inserting in equation (1) one gets:

MA − MB = ∆M = J ⋅ 2π ⋅ n . (4)

This relation is solved with respect to n and one integrats with respect to time

1
∆=
2π ⋅ J ∫
n ⋅ ∆Mdt . (5)

By comparing with the differential equation

x=
a KI ⋅ ∫ xe dt (6)

one recognizes that it is an IT0 transmission element, where the proportionality factor KI has the
form
1
KI = . (7)
2π ⋅ J

3
Sample solution

Problem 5-3B: room heating

balance equation:

For the balance equation one first determines the balance of the heat flows: the difference be-
tween the incoming and outgoing heat flow forms the temporal change of the stored heat in the
room.

 −Q
Q  
in out = Qstored heat . (1)

coupling relations:

Next you need the coupling relations, so that, except for constants, only the output and input var-
iables appear in the equations. The outgoing heat flow is proportional to the total wall area and
the temperature difference; the proportionality factor is the heat transfer coefficient. This results
in

out = k ⋅ A ⋅ (ϑ − ϑa ) = k ⋅ A ⋅ ∆ϑ .

Q (2)

The temporal change of the stored heat depends on the temporal change of the room tempera-
ture, because the specific heat and the mass in the room are constant. Using that the outdoor
temperature is constant, the resulting expression contains the output variable ∆ϑ .

 dQ d d
Q stored heat = = cp ⋅ m ⋅ ϑ = cp ⋅ m ⋅ (ϑ − ϑa ) = cp ⋅ m ⋅ ∆ϑ . (3)
dt dt dt

insert in balance equation and normalize:

With the insertion into the balance equation, i.e. Eq. (2) and Eq. (3) in Eq. (1), the following results

 − k ⋅ A ⋅ ∆ϑ = c ⋅ m ⋅ ∆ϑ
Q  . (4)
in p

The normalization is characterized by the fact that the output variable is shown on the left side of
 , is shown on the right side. Furthermore, the coeffi-
the equation, ∆ϑ , and the input variable, Q zu
cient in front of the output variable itself is equal to one. The equation must therefore be resorted
and divided by k ⋅ A .

cp ⋅ m 1 
zu . (5)
∆ϑ +  =
⋅ ∆ϑ ⋅Q
k⋅A k⋅A

By comparison with the differential equation

x a + T1 ⋅ x a = K ⋅ x e (6)

we can see that it is a PT1 transmission element, where the proportionality factor K has the form
 1
K = (7)
k⋅A

and the characteristic time T1 has the form

cp ⋅ m
T1 = . (8)
k⋅A
Sample solution

Problem 6-2B:

The differential equation for an IT0 transmission element is

x a(t) = K I ∫ x (t) dt .
e (1)

According to the properties of the Laplace transform, the integral is transformed over a function
of time t as follows:

Laplace {∫ f(t)dt} = F(ss) . (2)

This results in the following relation for equation (1):

X e (s)
X a(s) = KI . (3)
s

With the result from equation (3) the transfer function G(s)=Xa(s)/Xe(s) can now be determined. The
result is

X a(s) K I
G(s) = = . (4)
X e (s) s

The Laplace transform of an impact is Xe(s)=1. So equation (3) gives

KI K K
X a(s) = X e (s) = 1⋅ I = I , (5)
s s s

which, transformed back into the time domain, provides a jump function,

0 at t < 0
x a (t ) =  .
KI at t > 0

This result is already known from the relationships between the standardized test functions, be-
cause the integration of an impact results in a jump.
The Laplace transform of a jump is Xe(s)=1/s. So equation (3) gives

KI 1 KI K
X a(s) = X e (s) = = 2I , (5)
s s s s

which, transformed back into the time domain, provides a ramp function,

0 at t < 0
x a (t ) =  .
KIt at t > 0

This result is already known from the relationships between the normalized test functions, be-
cause the integration of a jump results in a ramp.
Problem 6-8B:

a)

To transform the expression for the mean square deviation, the binomial formula is explicitly writ-
ten out and constant quantities are written before the sums.

L −1 R −1 L −1 R −1 L −1 R −1 L −1 R −1
2mS m2
qS =
1
∑∑ [s(x, y) − mS ]2 = 1 ∑∑s(x, y)2 − ∑∑ s(x, y) + S ∑∑ 1
M x =0 y =0 M x =0 y =0 M x =0 y =0 M x =0 y =0

The first term cannot be further simplified, the second term contains the mean value of s(x, y) and
the third term is summed using the constant "1".

L −1 R −1 L −1 R −1 L −1 R −1
1 1 m2
qS = ∑∑
M x =0 y =0
s(x, y)2 − 2mS ⋅ ∑∑
M x =0 y =0
s(x, y) + S
M ∑∑ 1
x =0 y =0

The result is:

L −1 R −1
1 m2
qS = ∑∑
M x =0 y =0
s(x, y)2 − 2mS ⋅ mS + S ⋅ M .
M

The simplification of the last expression results now:

L −1 R −1
1
qS = ∑∑
M x =0 y =0
s(x, y)2 − m2S .

b)

The homogeneous image is characterized by the fact that the grey value s(x,y) is identical for all
pixels. In the given case, this is assigned the value 127: s(x,y)=127. This constant value can be placed
before the sums for the determination of the mean value as well as for the mean square deviation.

L −1 R −1 L −1 R −1
1 127
mS = ∑∑
M x =0 y =0
127 = ∑∑
M x =0 y =0
1

L −1 R −1
1272
qS =
M ∑ ∑ 1 − 127
x =0 y =0
2

The execution of the summation returns the value M = L ⋅ R in all cases, which results in

127
mS = ⋅ M = 127 and
M

1272
qS = ⋅ M − 1272 = 0 .
M
Solutions of the 1st exercise

Problem 2:

variant 1:

variant 2:

Problem 3L:

Problem 4:
Problem 6L:

a) The static characteristic curve is degressive.

b) linear interpolation

(400 − 300) °C
ϑ(R ϑ = 244 Ω) = (244 − 212,05) Ω + 300 °C = 391,18 °C
(247,09 − 212,05) Ω

1 −6 1
0 = −R ϑ + R 0 + R 0 aϑ + R 0bϑ2 mit a = 3,9083 ⋅ 10 − 3 und b = −0,5775 ⋅ 10
°C (°C)2

a R 20a2 + 4R 0b(R ϑ − R 0 ) a R 20 a2 + 4R 0b(R ϑ − R 0 )

ϑ1 = − + ϑ2 = − −
2b 2R 0b 2b 2R 0b

ϑ1 (R ϑ = 244 Ω) = 391,04 °C ϑ2 (R ϑ = 244 Ω) = 6376.578 °C entfällt

c) secant linearisation

∆R ϑ (247,09 − 212,05) Ω Ω
ESek = = = 0,3504
∆ϑ R =244 Ω (400 − 300)°C °C

d) tangent linearisation

dR ϑ 1 1
ETan = = R 0 ⋅ (3,9083 ⋅ 10−3 − 0,5775 ⋅ 10−6 ⋅ 2 ⋅ ϑ(R ϑ = 244 Ω)
dϑ R ϑ =244 Ω °C (°C)2

dR ϑ 1 1 Ω
ETan = = R 0 ⋅ (3,9083 ⋅ 10−3 − 0,5775 ⋅ 10−6 ⋅ 2 ⋅ 391,04°C) = 0,3457
dϑ R ϑ =244 Ω °C (°C)2
°C
Problem 7:

dϑ ab
a) ϑ = ϑ 0 + a b ⋅ A =
dA 2 bA

dϑ °C
Etan(A = 2 cm2 ) = = 1,500
dA A =2 cm2 cm2

ϑ(3 cm2 ) − ϑ(1 cm2 ) (25,35 − 22,24) °C °C

E sek (A = 2 cm2 ) = 2
≅ 2
≅ 1,555
(3 − 1) cm 2 cm cm2

dϑ °C
b) Etan(A = 12 cm2 ) = ≅ 0,612
dA 2
A =12 cm cm2

ϑ(13 cm2 ) − ϑ(11 cm2 ) (33,30 − 32,07) °C °C

E sek (A = 12 cm2 ) = ≅ ≅ 0,615
(13 − 11) cm2 2 cm2 cm2

c) for A 0 = 2 cm2 Esek=103,67 % also ∆E=3,7%

for A 0 = 12 cm2 Esek=100,49 % also ∆E=0,5%

cause of the larger deviation at A 0 = 2 cm2 : characteristic curve is there

more nonlinear

Problem 8H:

3
 = α ⋅ π ⋅ d2 ⋅ 2 ⋅ g ⋅ h = 0,9 ⋅ π ⋅ (0,02 m)2 ⋅ 2 ⋅ 9,81 m ⋅ 1,2 m = 1,372 ⋅ 10 − 3 m
a) V
4 V 4 s2 s

3
 = 1,372 ⋅ 10 − 3 m = 1,372 l m3 3 l
b) V weil 1 = 10
s s s s
3 3 3
 = 1,372 ⋅ 10 − 3 m = 1,372 ⋅ 10 − 3 m ⋅ 3600 s = 4,939 m
c) V
s s h h

Problem 9H:

a)
∂y
= 1
( )
x 4 − x24 ⋅ 1 − x1 4x13
=
( )
− 3x14 − x24
∂x1 (
x14 − x24
2
) x14 − x24
2
( )
∂y
= x1
∂y 4
(
x1 − x24 )
−1
(
= −x1 x14 − x24 ) (− 4x ) =
−2 3 4x1x32
∂x2 ∂x2
2
(x
4
1 − x24 )
2
 ∂y
b) = 2x1 x2 (t)
∂x1

∂y ∂y ∂x2 1 ∂x2 x12x 2

= = x12 =
∂t ∂x2 ∂t 2 x2(t) ∂t 2 x2(t)

∂y
c) = 2 cos(2α)[− sin(2α)]2 = −4 sin(2α) cos(2α)
∂α

1  UaR 1 UeR 2 − UaR 2  Ua (R 1 + R 2 ) − UeR 2

d) y =  −  =
C  R 2R 1 R 1R 2  CR 1R 2

Ua  1 1  Ua  R 1 + R 2 
 +   
C  R 2 R 1  = C  R 1R 2  = Ua  R 1 + R 2  R2 U
e) y =  = a
R1 R 2
+
 R1 + R 2 
 
C  R 1R 2  (R 1 + R 2 ) CR 1
R2 R2  R2 
Solutions of 2nd exercise

Problem 1L:

1. variant

equation of forces: F = mx
 + γx + kx

 = 1 ⋅ F − γ ⋅ x − k ⋅ x
 : x
solve with respect to x
m m m
 to x: double integration
from x x = ∫ x dt und x= ∫ x dt

2. variant: create a dependency tree starting at the output signal

xa = x = ∫ x dt

x = ∫ x dt

 = 1 Fres
x
m

(
Fres = F − Fγ + Fk )

Fγ = γx

Fk = kx
Problem 2:

a)

b) – both voltmeters must have identical internal resistances RMV

– it must be valid

RMV >> (R x , Rr )

Ux Ur
c) it holds: I0 = I0 =
Rx Rr

Ux U Ux
also = r → Rx = ⋅ Rr
R x Rr Ur

Problem 3H:

a)

R123 = R1 + R2 + R3

R 4 ⋅ R 123 15
R 4123 = = Ω = 3,75 Ω
R 4 + R 123 4

b) for Ri=5 Ω (i=1, 3, 4) and R2=10 Ω R 4123 = 4 Ω = R E

for Ri=10 Ω (i=1, 2, 3) and R4=10 Ω R 4123 = 6 Ω > 4 Ω = R E

for the first resistor assignment it is possible

Problem 5:

To determine the equivalent internal resistance R i , the circuit must be viewed from the open
terminals A and B.
 R1R 2
R 3 in series connection with parallel connection of R1 and R 2 : Ri = R3 +
R1 + R 2

UK
equivalent circuit: IK =
Ri

measuring instrument with R M in series connection with R i

Uk Ri
measuring current: IM = = IK
Ri + RM Ri + RM

 Ri  RM
difference of currents: ∆I = IK − IM = IK 1 −  = Ik
 Ri + RM  Ri + RM
∆I RM R
relative error: = ≈ M if R M << R i
IK Ri + RM Ri

∆I RM 0,5
with Ri = 10 Ω and RM = 0,5 Ω : = = ≅ 4,76 %
IK Ri + RM 10,5

When the ammeter is connected to an active two-terminal device, it must be Ri >>RM .

 currents must always be measured with an ammeter with low internal resistance!

Problem 7L:

R1 ⋅ R 4 200 Ω ⋅ 600Ω
a) R1⋅R4 = R2⋅R3 = R2⋅R0⋅α  R0 = = = 120 Ω
R2 1000 Ω
b) Ud = U1 - U3

U1 R1 U0 ⋅ R 1
= → U1 =
U0 R1 + R 2 R1 + R 2

U3 R0 ⋅ α R0 ⋅ α
= → U3 = U0 ⋅
U0 R4 + R0 ⋅ α R4 + R0 ⋅ α

 R1 R0 ⋅ α 
Ud = U0  − 
 → solve with respect to α
 R1 + R 2 R 4 + R 0 ⋅ α 

 R1 Ud  R 4  200Ω 0,003 V  600Ω

 
R + R − U  ⋅ R  − ⋅
10 V  120Ω
 1 2 0  0  1200Ω
α = = = 0,9978
  R1 Ud    200Ω 0,003 V 
1 −  −  1 −  − 
  R 1 + R 2 U0 

  1200Ω 10 V 

The resistance R3 has been changed by 0.22%.

Problem 8:

for R 4

 R4 R2   R − ∆R R0 
Ud = U0  −  = U0  0 − 
 R 3 + R 4 R1 + R 2   2R 0 − ∆R 2R 0 + ∆R 
 R − ∆R  ∆R 
−1
1 ∆R 
−1 
U d = U0  0 1 −


 − 1 + 


 2R 0  2R 0  2 2R 0  
 
 R − ∆R  ∆R  1 ∆R 
U d ≅ U0  0 1 +

 − 1 −
 2


 2R 0  2R0   2R0 

 1 ∆R ∆R ∆R 2 1 ∆R  U0 (∆R)2
Ud ≅ U0  − + − 2
− + =−
 2 2R 0 4R 0 4R 0 2 4R 0  4 R 20

or for R 1

 R1 R3   R + ∆R R0 
Ud = U0  −  = U0  0 
  2R + ∆R − 2R − ∆R 
 R1 + R 2 R 3 + R 4   0 0 
 R + ∆R  ∆R 
−1
1 ∆R 
−1 
U d = U0  0 1 +


 − 1 −




 2R 0  2R 0  2 2R 0  
 
 R + ∆R  ∆R  1 ∆R 
U d ≅ U0  0 1 −

 − 1 +
 2


 2R 0  2R 0   2R 0 

1 ∆R ∆R ∆R 2 1 ∆R  U0 (∆R)2
Ud ≅ U0  + − − 2
− − =−
 2 2R 0 4R 0 4R 0 2 4R 0  4 R 20
Solutions of the 3rd Exercise

Problem 1L:

a)

Since the changes (due to a change in luminous intensity) in diagonally arranged resistors add up
(for the largest possible signal), the LSDs must be arranged diagonally, i.e. half bridge.

R1 = R 4 = R LSD and thus R 2 = R 3 = R fix

b) in bright sunlight:

2
Ud = 0 ↔ R1R 4 − R 2R 3 = 0 ↔ R LSD = R 2fix

→ R fix = R LSD, bright = 100 kΩ

c) for cloudy sky:

determination of illuminance

k = {tillu } {E} ⋅ Si → E[Lux ] =

k2 ⋅ C
= 19600 Lux
C {tillu } ⋅ Si

determination of the change in resistance (half-bridge): factor 2 for the

resistance change ∆R LSD

U0 ∆R LSD 2Ud
Ud = 2⋅ → ∆R LSD = R LSD, bright = 20 kΩ
4 R LSD, bright U0

- LSD has a higher resistance at lower illuminance, i.e.

R LSD, bright < R LSD, cloudy

- due to linear characteristic curve:

~ R LSD, bright − R LSD, cloudy R LSD, bright − (R LSD, bright + ∆R LSD ) ∆R LSD
ELSD = = = −
Ebright − Ecloudy Ebright − Ecloudy ∆E
 ~ 20000 Ω Ω
ELSD = − ≅ −0,36
(75600 − 19600) Lux Lux

Problem 2:

1 1 1 1 1 1 1 + iωC1R1
a) R1 || C1 : = + =
Z1 ZR1 ZC1 R1
+
−i
=
R1
+ iωC1 =
R1
⇒ Z1 =
R1
1 + iωC1R 1
ωC1
−i −i
C2 : Z2 = C4 : Z4 =
ωC2 ωC 4
1
R3 ωC x tan δ x 1
Cx : Zx = = =
1 + iωC xR 3 1 iωC x (1 − i tan δ x )
1 + iωC x
ωC x tan δ x

R1  i  − ωC1R12 − iR 1
Z1 ⋅ Z4 = ⋅  −  =
2 2 2
1 + iωC1R1  ωC 4  ωC 4 (1 + ω C1R1 )

1  i  − 1 − i tan δ x
Z2 ⋅ Z3 = ⋅  −  =
iωC x (1 − i tan δ x )  ωC2  ω C2C x (1 + tan2 δ x )
2

C 4 (1 + ω2C12R12 )
Re(Z1 ⋅ Z4 ) = Re(Z2 ⋅ Z3 ) yields Cx = (1)
C1R12 ω2C2 (1 + tan2 δ x )

Im(Z1 ⋅ Z4 ) = Im(Z2 ⋅ Z3 ) yields R1ω2C2C x (1 + tan2 δ x ) = tan δ x ωC 4 (1 + ω2C12R12 ) (2)

insert equation (1) in equation (2) einsetzen and solve for tan δ x

1
tan δ x = (3)
ωC1R 1

insert expression (3) in equation (1) gives

C 4C1
Cx = (4)
C2

b)
real part: R1R 4 − X1X 4 = R 2R 3 − X2 X3

imaginary part: R 1X 4 + X1R 4 = R 2 X 3 + X 2R 3

1 1 1 1
X1 = − X2 = − X3 = − X4 = −
ωC1 ωC2 ωC x ωC 4

1
R1 ≠ 0 R2 = 0 R3 = R4 = 0
ωC x tan δ x
  1  1   1  1  C 4C1
−  −  −  = − −  −  gives Cx =
 ωC1  ωC 4   ωC2  ωC x  C2

1 1  1  1
− R1 =  −  gives tan δ x =
ωC 4 ωC3 tan δ x  ωC 2  ωC1R 1

Problem 3H:

a) It must Re(Z1 ⋅ Z4 ) = Re(Z2 ⋅ Z3 ) and Im(Z1 ⋅ Z4 ) = Im(Z2 ⋅ Z3 ) be used. The capacitive resistor
Z4 is a series connection of an Ohmic resistor and a capacitor. The capacitive resistor Z3 is a
parallel connection of an Ohmic resistor and a capacitor. Thus, each diagonal has a different
complex resistance.

b) Z1 = R1 Z2 = R 2

1 1 1 1 + iωC3R 3 R3
= + = → Z3 =
Z3 ZR3 ZC3 R3 1 + iωC3R 3
i
Z4 = R 4 −
ωC 4

−R 1
Z1 ⋅ Z4 = R 1R 4 + i
ωC 4

R 2R 3 (1 − iωC3R 3 )
Z2 ⋅ Z3 =
1 + ω2C23R 23

equal real parts: R1R 4 (1 + ω2C23R 23 ) = R 2R 3 (1)

equal imaginary parts: R1(1 + ω2C23R 23 ) = ω2C3C 4R 2R 23 (2)

R1R 4 R R
divide equation (1) by (2): = 2 2 3 2
R1 ω C3C 4R 2R 3

1
solve with respect to f 2 : f2 = 2
(2π) C3C 4R 3R 4

two meaningful conditions: C3 = C 4 = C und R3 = R 4 = R

1
that gives for f: f=
2πCR

inserting in equation (1) yields: R 2 = 2R 1

Problem 5:

a) I=I0+a⋅(p-p0) I0=4 mA p0=1 bar

I − I0 16 mA mA
from p=10 bar and I=20 mA : a= = = 1 ,778
p − p0 9 bar bar

conversion pressure: p[bar]=p[kp/cm2]·0,980665

mA
nominal values from : Inom = 4 mA + 1,778 ⋅ (p − 1) bar
bar

pact kp/cm2 1,00 2,00 3,00 4,00 5,00 6,00 7,00 8,00 9,00 10,00

pact bar 0,98 1,96 2,94 3,92 4,90 5,89 6,87 7,85 8,83 9,81

Iact mA 4,15 5,9 7,35 9,50 11,25 12,70 14,75 16,20 18,20 19,80

Inom mA 3.97 5,71 7,45 9,2 10,94 12,69 14,43 16,18 17,92 19,66

∆I=
mA 0,18 0,19 - 0,10 0,30 0,31 0,01 0,32 0,02 0,28 0,14
Iact-Inom

∆Imax 0,32 mA
b) error class 2,5 from: = = 0,020 = 2%
∆Imeasuring range 16 mA

c) graph of errors

------ guide for the eye, no functional relation

Problem 6:

a) R = 125,46 Ω

∑ (R )
2
i −R
i =1 σ
b) mR = = = 0,428 Ω
n ⋅ (n − 1) n

~
c) mR = t ⋅ mR = 2,36 ⋅ 0,428 Ω = 1,01 Ω result: R = (125 ± 1) Ω

d) temperature of the measurement sensor with αR = 38,5 ⋅ 10−4 / °C

(R ϑ − 100) Ω 25,46 Ω
∆ϑ = ϑ(°C) − 0°C → ϑ= = = 66,1°C
100 Ω ⋅ αR 38,5 ⋅ 10 − 4 / °C ⋅ 100 Ω

~ ∆R ϑ 1 ,01 Ω
e) uncertainty of temperature measurement: m ϑ = = = 2,6 °C
αR ⋅ 100 Ω 38,5 ⋅ 10− 4 / °C ⋅ 100Ω

result: ϑ = (66 ± 3) °C

Problem 7L:

U 72 V
a) R = = = 153,19 Ω
I 0,47 A

U2
P = U ⋅I =
R

maximum relative measurement uncertainty:

∂P 1 ∂P 1 2U R U2 R ∆U ∆R
∆Prel,max = ∆U + ∆R = ∆U + − ∆R = 2 + − = 2 ∆Urel + ∆Rrel
∂U P ∂R P R U 2 2
R U 2 U R
1,5 V 4,6 Ω
∆Prel,max = 2 ⋅ + = 0,072 bzw. 7,2%
72 V 153,19 Ω

b) P = U ⋅ I

maximum relative measurement uncertainty

∂P 1 ∂P 1 1 1 ∆U ∆I
∆Prel,max = ∆U + ∆I = I ∆U + U ∆I = + = ∆Urel + ∆Irel
∂U P ∂I P U⋅I U⋅I U I

1,5 V 0,02 A
∆Prel,max = + = 0,063 bzw. 6,3%
72 V 0,47 A
 U2
c) P =
R

maximum relative uncertainty of measurement under consideration of the signs:

∂P 1 ∂P 1 ∆U ∆R
∆Prel,max = ∆U + ∆R = 2 − = 2∆Urel − ∆Rrel
∂U P ∂R P U R

∆Prel,max = 2 ⋅ (−0,013) − (−0,024) = −0,002 bzw. − 0,2%

d) P = I2 ⋅ R

maximum relative uncertainty of measurement under consideration of the signs:

∂P 1 ∂P 1 1 1 ∆I ∆R
∆Prel,max = ∆I + ∆R = 2I ⋅ R 2 ∆I + I2 ⋅ 2 ∆R = 2 + = 2∆Irel + ∆Rrel
∂I P ∂R P I ⋅R I ⋅R I R

∆Prel,max = 2 ⋅ 0,034 + (−0,024) = 0,044 bzw. 4,4%

Problem 8:

4 2 4
 ∂η ∆xi 
∆ηrel,prob = ∑  
 ∂x η  =
i =1  i 
∑f
i =1
i
2

i=1
∂η ∆(∆p) ∆(∆p) ∆(∆p) 0,3
η = K1 ⋅ ∆p f1 = = K1 = = ≅ 1,974%
∂(∆p) η K1 ⋅ ∆p ∆p 15,2

i=2
∂η ∆R ∆R ∆R 0,05
η = K2 ⋅ R 4 f2 = = K 2 4R 3 4
=4 =4 ≅ 6,667%
∂R η K2 ⋅ R R 3

i=3
1 ∂η ∆qV (−1) ∆qV ∆q 3
η = K3 ⋅ f3 = = K3 2 =− V =− = −7,500 %
qV ∂qV η qV 1 qV 40
K3 ⋅
qV
i=4
1 ∂η ∆L (−1) ∆L ∆L 0,5
η = K4 ⋅ f4 = = K4 2 =− =− ≅ −3,333 %
L ∂L η L K ⋅1 L 15
4
L
probable relative error

∆ηrel,prob = f12 + f22 + f32 + f42 ≅ 115,704 % ≅ 10,76%

Solutions of the 4th Exercise

Problem 1L:

a) calculation mean value:

N= 8

∑ϑ
i=1
TE, i
ϑTE = = 43,425°C
N

calculation absolute maximum error:

∆ϑTE, abs, max = ∆ϑTE + ∆ϑdirect connection ∆ϑTE = 2,5 K

0,4
∆ϑdirect connection = ⋅ 100 K = 0,4 K
100
∆ϑTE, abs, max = 2,9 K

Why calculation of maximum error by this way and not by calculation the mean error a la 3.
exercise, problem 6L (assumption: probability P=95% for measurement uncertainty results in a t-
factor of t=2,306)?

∑ (ϑ
i =1
TE − ϑ TE, i )2
mϑTE = ⋅ 2,306 ≅ 0,79 K
8(8 − 1)

This error is smaller than the error from the measurement uncertainty of the thermocouple and
the direct connection.

result: ϑTE = (43 ± 3)°C

b) ϑTE = 43,425°C ∆ϑTE, abs, max = 2,9 K

1 range  of measurement

 
ϑHg = 43,1°C ∆ϑHg =
100
(
⋅ ϑ end of range of measurement − ϑbegin of range of measurement )

error class

50 K
∆ϑHg = = 0,5 K
100

1 1 1 1
(
g ∆ϑTE, abs, max = ) = K −2 (
g ∆ϑHg =) = K −2 = 4 K −2
(∆ϑTE, abs, max ) 2
2,9 2
(∆ϑHg ) 2
0,52

1
⋅ 43,425 + 4 ⋅ 43,1
ϑres =
( ) ( )
g ∆ϑTE, abs, max ⋅ ϑTE + g ∆ϑHg ⋅ ϑHg
=
2,92
°C ≅ 43,11°C
g(∆ϑTE, abs, max ) + g(∆ϑHg ) 1
+ 4
2,92
 1 1
∆ϑres = = ≅ 0,49 K
∑ gi ( ) (
g ∆ϑTE, abs, max + g ∆ϑHg )
i

Problem 2:

x12
a) y1 =
2x2

∂y1 x1 ∂y1 x2
with Gaussian law of error propagation = = − 12
∂x1 x2 ∂x2 2x2

2 2 2 2
x  2  2x   x2  2  2x 
∆y1, rel,prob =  1  (∆x1 )  22  +  − 12  (∆x2 )  22 
   2x   x 
 x2   x1   2  1 
2 2
 ∆x1   ∆x2 
∆y1, rel,prob = 2  
 x  + x  =
 (2rx )2 + (rx )2
1 2
 1   2 

with rule for special cases

~
y1 = x12 ⋅ x2−1 ( ) with ~
x2 = 2x2

∆y1, rel,prob = (rx )2 + (r~x )2

1 2

1st factor: power 2; 2nd factor: power -1

~
∆x1 ~ ∆x 2∆x2 ∆x
rx1 = 2∆x1rel = 2 r~x = (−1) ⋅ ∆ x2rel = − ~ 2 = − =− 2
x1 2
x2 2x2 x2

2 2
∆y1, rel,prob = (rx ) + (r~x )
2 2  ∆x   ∆x2 
=  2 1  +  − 

 x1   x2 
1 2

x1 + x2
b) y2 =
x33
∂y2 1 ∂y2 1 ∂y2 3(x1 + x2 )
with Gaussian law of error propagation: = 3 = 3 =−
∂x1 x3 ∂x2 x3 ∂x3 x34

2 2 2 2 2
 1  2 x33   1  2 x33  3(x1 + x2 ) 2 2 x33 
∆y2, rel,prob =  3  (∆x1 )  +  3  (∆x2 )  + (− ) (∆x )
3 

x  x + x  x  x + x  x34 
 3  1 2  3  1 2  x1 + x2 
2
∆x12 + ∆x22 ∆x12 + ∆x22
∆y2, rel,prob =
 ∆x 
+  3 3  = (
+ 3rx 3 )2
(x1 + x2 )2  x3  (x1 + x2 )2
 with rule for special cases
~ ( )
y2 = x ⋅ x3−3 mit ~
x = x1 + x2

∆y 2, rel,prob = (r~x )2 + (rx )2 3

∆x3
für rx3 : rule for powers rx3 = (−3) ⋅ ∆xrel
3 = −3
x3

~
für r~x : rule for sums ∆ xabs, prob = ∆x12 + ∆x22

 ∆~
2
xabs, prob  ∆x12 + ∆x22
(~
∆ xrel, prob )
2
= 
 ~
x


=
(x1 + x2 )2
 

2
∆x12 + ∆x22
∆y2, rel,prob = (r~x )2 + (rx )2 = 2
 ∆x 
+  − 3 3 
3
(x1 + x2 )  x3 

x1 + 2x2 x 1
c) y3 = = 1 +
4x2 4x2 2

∂y3 1 ∂y3 x
with Gaussian law of error propagation = = − 12
∂x1 4x2 ∂x2 4x2

2 2 2 2
 1   4x2   x   4x2 
∆y3, rel,prob =   (∆x1 )2   +  − 1  (∆x2 )2  
 2
 4x2   x1 + 2x2   4x2   x1 + 2x2 

∆x12 ∆x22 x12 x12  ∆x12 ∆x22 

∆y3, rel,prob = + =  + 2 
(x1 + 2x2 ) 2
x22 (x1 + 2x2 ) 2
(x1 + 2x2 ) 2  x2
 1 x2 

x1 + 2x2 x 1
if rules for special cases (quotient) y3 = = 1 +
4x2 4x2 2
 ∆x2 ∆x2 
∆y3, rel,prob =  21 + 22 
 x x2 
 1

enumerator(x1, x2 )
no agreement, since ; in the enumerator as well as in the denominator
deno min ator(x2 )
appear the same variable, namely x2

Problem 4L:

a) P = p1T 4 + p0 is a straight line of the kind y = m ⋅ x + t

with y = P , x = T 4 , the slope m = p1 and the axis intersection t = p0

X = {X} [X] {X} digit of quantity X,

[X] physical unit of quantity X
x = {T}4 2754 300 4 3254 350 4 3754 400 4 4254 450 4 4754 500 4

y = {P} 0,1028 0,0865 0,1373 0,31 0,3316 0,289 0,5644 0,5701 0,7676 0,8975

b)

xi yi xi2 yi2 xi ⋅ yi

2754 0,1028 2758 0,01056784 2754 ⋅ 0,1028

300 4 0,0865 3008 0,00748225 300 4 ⋅ 0,0865
3254 0,1373 3258 0,01885129 3254 ⋅ 0,1373
350 4 0,31 3508 0,0961 350 4 ⋅ 0,31
3754 0,3316 3758 0,10995856 3754 ⋅ 0,3316
400 4 0,289 4008 0,083521 400 4 ⋅ 0,289
4254 0,5644 4258 0,31854736 4254 ⋅ 0,5644
450 4 0,5701 4508 0,32501401 450 4 ⋅ 0,5701
4754 0,7676 4758 0,58920976 4754 ⋅ 0,7676
500 4 0,8975 5008 0,80550625 500 4 ⋅ 0,8975

[x] ≅ 2,7240 ⋅ 1011 [y] = 4,0568 [x ] ≅ 1,0738 ⋅ 10

2 22

[y ] ≅ 2,3648
2
[x ⋅ y] ≅ 1,5839 ⋅ 1011

best fit line y = m ⋅ x + t

[x ⋅ y] ⋅ n − [x] ⋅ [y] = 1 (n ⋅ [x ⋅ y] − [x] ⋅ [y]) ≅ 0,1443 ⋅ 10−10

n ⋅ [x2 ] − [x ] ⋅ [x]
slope m=
∆

with [ ]
∆ = n ⋅ x2 − [x ] ⋅ [x ] ≅ 3,3181 ⋅ 1022

axis intersection
[x ]⋅ [y] − [x] ⋅ [x.y] = 1 ([x ]⋅ [y] − [x] ⋅ [x ⋅ y]) ≅ 0,0126
2
2
n ⋅ [x ] − [x ] ⋅ [x]
t = 2 ∆
 xi yi yFit
i ∆yi = yi − yFit (∆yi )2
i

2754 0,1028 0,0951 0,0077 5,9003 ⋅ 10-5

300 4 0,0865 0,1295 -0,0430 1,8470 ⋅ 10-3
3254 0,1373 0,1736 -0,0363 1,3168 ⋅ 10-3
350 4 0,31 0,2291 0,0809 6,5382 ⋅ 10-3
3754 0,3316 0,2980 0,0336 1,1313 ⋅ 10-3
400 4 0,289 0,3820 -0,0930 8,6526 ⋅ 10-3
4254 0,5644 0,4834 0,0810 6,5606 ⋅ 10-3
450 4 0,5701 0,6043 -0,0342 1,1729 ⋅ 10-3
4754 0,7676 0,7472 0,0204 4,1536 ⋅ 10-4
500 4 0,8975 0,9145 -0,0170 2,8979 ⋅ 10-4

∑ (∆y )
2
i ≅ 2,7894 ⋅ 10-2
1
(∆y )2 =
n−2
⋅ ∑ (∆y ) i
2
≅ 3,4979 ⋅ 10-3

(∆m)2 = (∆y )2 ⋅ n ≅ 1,0542 ⋅ 10-24 ⇒ ∆m ≅ 1,0267 ⋅ 10-12 error of the slppe

∆

(∆t )2 = (∆y )2 ⋅ [x ] ≅ 1,1320 ⋅ 10

2
-3
⇒ ∆t ≅ 3,3645 ⋅ 10-2 error of the axis of intersection
∆

correlation coefficient r:

n ⋅ [y ⋅ x] − [x] ⋅ [y]
r = ≅ 0,9804 with [x]2 = [x]·[x]
(n[x ] − [x] ) ⋅ (n[y ] − [y] )
2 2 2 2
c) p1 ≅ 1,4431 ⋅ 10 −11 and p0 ≅ 0,0126

Problem 6:

balance equation: conservation of electrical charge Q

Qind − Qstore = 0 x e = ∆F x a = ∆U

Qind : amount of charge induced and released by F

Qstore : stored amount of charge, because quartz acts as capacitor

coupling relations: Q ind = k ⋅ ∆F

ε 0 εr A
Qstore = C ⋅ ∆U plate capacitor C =
d
ε 0 εr A
Qstore = ⋅ ∆U
d

ε 0 εr A
inserting in balance equation: k ⋅ ∆F = ⋅ ∆U
d

kd
standardize: ∆U = ⋅ ∆F
ε 0 εr A

kd
differential equation: xa = K ⋅ xe PT0 transmission element with K =
ε 0 εr A
Problem 7:

d
∆s = ∆ω ⋅ r → ∆s = ∆ω ⋅ r
dt
1 1
∆ω = ⋅ ∆s DT0 transmission element with KD =
r r

differential equation: x a = KD ⋅ x e

Problem 8:

balance of forces: Fc (∆s) + FD(∆s ) = ∆Fres (∆F)

coupling relations: Fc (∆s) = c ⋅ ∆s FD (∆s ) = k ⋅ ∆s

- for levers in equilibrium holds: the sum of the torques is equal to zero

     
∑M
i
i =0= ∑r ×F
i
i i that means a × Fres = b × F

 
( 
)
   
| a | Fres sin a, Fres + | b | F sin b, F = 0 ( )

aFres − bF = 0

- small path changes ∆s cause small force changes: ∆Fres , ∆F

b
- that means: ∆Fres ⋅ a = ∆F ⋅ b → ∆Fres = ⋅ ∆F
a

b
inserting in balance equation: c ⋅ ∆s + k ⋅ ∆s = ∆F
a

ac ka 
standardize: ∆F = ⋅ ∆s + ⋅ ∆s
b b

differential equation: x a = K ⋅ x e + KD ⋅ x e

ac ka
PDT0 transmission element with K = and KD =
b b
Solutions of the 5th Exercise

Problem 1L:

L3
deflection s(L) = F
3⋅E⋅I

force 1
general: spring model deflection = = F
spring cons tan t c

3⋅E⋅I
comparison provides: c =
L3

1 c
undamped spring pendulum: f = mw : effective mass
2π mw

mw = mp + msystem mp : mass of the deposited particles

Zeitpunkt t = 0 : keine Partikel am Filter, also mw = msystem

1 3EI
fsystem = 3
2π L msystem

3EI  1 1 
b) mp = mw − msystem = − 2
4π L  fw fsystem 
2 3  2

moment of inertia of an annular surface: I =

π 4
4
(
R − r4 )

mp =
( )
3E R 4 − r 4  1
−
1 
= κ
1
 −
1 
with κ=
(
3E R 4 − r 4 )
16πL3  fw2 fsystem
2 

 2 2 
 fw fsystem  16πL 3

1 1 
c) mp = κ 2 − 2
f 
 w fsystem 

∂mp
linearisation at the point fSystem : mp (fw + ∆fw ) = mp (fw ) + ⋅ ∆fw
∂fw fsystem

∂mp
∆mp = mp (fw + ∆fw ) − mp (fw ) = ⋅ ∆fw
∂fw fsystem
  2 2κ
∆mp = κ − 3  ⋅ ∆fw = − 3 ∆fw
 f  fsystem
 w  fsystem

3 3
fsystem fsystem
normalize: ∆fw = − ∆mp PT0 element with K = −
2κ 2κ

Problem 2:

∫
x a(t) = K ⋅ x e + K I x e ⋅ dt
0 t<0 0 t<0
jump function σ (t) =  ; Ramp function (rise) α (t) = 
1 t≥0 t t≥0

a) jump response: x e = 1 at t ≥ 0

∫
t ≥ 0 : h(t) = K ⋅ 1 + K I 1 dt = K + K I ⋅ t

b) ramp response: xe = t at t ≥ 0

t²
∫
t ≥ 0 : a(t) = K ⋅ t + K I t dt = K ⋅ t + K I ⋅
2

Problem 4L:

a) bilance equation: FD − FR = FB

coupling relationships:

pressure force: FD = A ⋅ ∆ p acceleration force: FB = m ⋅ ∆ c = ρ ⋅ A ⋅ L ⋅ ∆ c

friction force: FR = U ⋅ L ⋅ τ = 8 ⋅ π ⋅ L ⋅ ν ⋅ ρ ⋅ ∆ c
λ ⋅ ρ ⋅ ∆c2 64 ⋅ ν
U = π ⋅D , τ = , λ=
8 ∆c ⋅ D
 D2 D2
inserting in balance equation and normalize: ⋅ ∆ c + ∆ c = ⋅ ∆p
32 ⋅ ν 32 ⋅ L ⋅ ν ⋅ ρ

differential equation: xa + T1 ⋅ x a = K ⋅ xe

D2 D2 m2 ⋅ s
b) PT1 element, T1 = = 27,85 s; K= = 2,79.10− 4
32 ⋅ ν 32 ⋅ L ⋅ ρ ⋅ ν kg

 −
t   −
t 
   m
solution of the differential equation: ∆c = K ⋅ ∆p ⋅ 1 − e T1  = 2,79 ⋅ 10
−2
⋅ 1 − e 27,85 s 
    s
   

m m
at t → ∞ : ∆c∞ = 0,0279 and thus c∞ = c0 + ∆ c∞ = 0,0779
s s

Problem 5:

balance equation: equilibrium of forces FD-FR-FT=0

coupling relations: FT = m ⋅ a = m ⋅ s = m ⋅ v

FR = r ⋅ v
FD = p ⋅ A

m A
inserting in balance equation and normalize: v+ v= p
r r

differential equation: xa + T1 ⋅ x a = K ⋅ xe

m A
PT1 transmission element with T1 = and K =
r r

Problem 6:

m
 in − m
 out = m
 container

coupling equations: m
 out = b p − pa

 container = p ⋅ V

mcontainerRT = pV → m
R⋅T
m
 in = a ⋅ y;
inserting in balance equation:

V
a ⋅ y − b p − pa = ⋅ p
R⋅T

b V 
y = p − pa + p non-linear differential equation y = y(p, p )
a RTa

linearisation at the operating point (p 0 , p 0 ) :

∂y ∂y
y(p0 + ∆p, p 0 + ∆p ) = y(p0 , p 0 ) + ⋅ ∆p + ⋅ ∆p
∂p p ∂p p
0 ,p 0 0 ,p 0
 

∂y ∂y
∆y = y(p0 + ∆p, p 0 + ∆p ) − y(p0 , p 0 ) = ⋅ ∆p + ⋅ ∆p
∂p p ∂p p
0 ,p 0 0 ,p 0
 

b 1 V b 1 V
∆y = ⋅ ∆p + ⋅ ∆p = ⋅ ∆p + ⋅ ∆p
a 2 p − pa RTa p0 ,p 0 a 2 p0 − pa RTa
p0 , p 0

2 p0 − p a V 2a p0 − pa
normalise: ⋅ ⋅ ∆p + ∆p = ⋅ ∆y
b RT b

differential equation: T1 ⋅ x a + x a = K ⋅ x e

2a p0 − pa 2 p0 − p a V
→ PT1 transmission element with K = and T1 = ⋅
b b RT

Problem 7H:

a) balance equation: Maschensätze UC + UR − Ue = 0; Ua = UR

1
coupling equations: UR = I ⋅ R ; UC =
C ∫ I ⋅ dt
inserting in balance equation:

1 d
C ⋅R ∫U a ⋅ dt + Ua − Ue = 0
dt

 + U = C ⋅R ⋅U
C ⋅R ⋅U  ⇒ DT1 ; T1 = C ⋅ R ; KD = C ⋅ R
a a e

differential equation: T1 ⋅ x a + xa = KD ⋅ x e

b) KD = T1 = C ⋅ R = 10 µF ⋅ 100 Ω = 1 ms
Problem 8:

a) balance equation: Node point set with the assumption that voltage measurement takes place
without current consumption

applies to both blue and green marked nodes: IL = IC + IR

dQL dUC UR
coupling equations: IL = Ic = C IR =
dt dt R

Ua = UC = UR

dQL dUC UR
inserting in balance equation: =C +
dt dt R


 +U = R ⋅Q
b) CR ⋅ U ⇒ DT1 ; T1 = C ⋅ R; KD = R
a a L

t
−
KD T1
c) transition function: h(t) = ⋅e
T1
KD
at t=0 h(t = 0) =
T1
t
−
K T1 K
h (t) = − D2 ⋅ e ; h (t = 0) = − D2 ;
T1 T1

~ K K K  t 
tangente at t=0: h = D − D2 ⋅ t = D 1 − T 
T1 T1 T1  1 
T1
~ K −
K
at h = 0 ⇒ t = T1; h(T1) = D ⋅ e T1 = 0,368 ⋅ D
T1 T1
Problem 9H:

i⋅c ic(1 − ib) cb + ic cb c

a) z1 = = = = +i
1 + i ⋅ b (1 + ib)(1 − ib) 1 + b 2 1+b 2
1 + b2

c 2b 2 c2 c 2 (1 + b 2 ) c
| z1 |= [Re(z1 )]2 + [Im(z1 )]2 = + = =
(1 + b ) (1 + b )
2 2 2 2
(1 + b )
2 2
1 + b2

a + i ⋅ c (a + ic)(1 − ib) a + ic − iba + cb a + cb c − ba

b) z 2 = = = = +i
1 + i ⋅ b (1 + ib)(1 − ib) 1+b 2
1+b 2
1 + b2

(a + cb)2 (c − ba)2
| z 2 |= [Re(z 2 )]2 + [Im(z 2 )]2 = +
(1 + b )2 2
(1 + b ) 2 2

a2 + 2abc + b 2 c 2 + c 2 − 2abc + b 2 a2 (a2 + c 2 )(1 + b 2 ) a2 + c 2

| z 2 |= = =
(1 + b )
2 2
(1 + b ) 2 2 1 + b2

D D(1 − a2 − ib) D(1 − a2 − ib)

c) z 3 = = =
1 − a2 + i ⋅ b (1 − a2 + ib)(1 − a2 − ib) (1 − a2 )2 + b 2

D(1 − a2 ) bD
z3 = 2 2 2
−i
(1 − a ) + b (1 − a2 )2 + b 2

D2 (1 − a2 )2 + b 2D2 D2 [(1 − a2 )2 + b 2 ] D
| z 3 |= 2 2 2
= 2 2 2
=
(1 − a ) + b (1 − a ) + b (1 − a2 )2 + b 2
E E E[−bd − id(1 − a2 )]
d) z 4 = = =
id(1 − a2 + i ⋅ b) − bd + id(1 − a2 ) [−bd + id(1 − a2 )][−bd − id(1 − a2 )]

E[−bd − id(1 − a2 )] Ebd Ed(1 − a2 )

z4 = 2 2 2 2 2
=− 2 2 2 2 2
−i
b d + d (1 − a ) b d + d (1 − a ) b d + d2 (1 − a2 )2
2 2

E2b 2 d2 + E2 d2 (1 − a2 ) E2 d2 [(1 − a2 )2 + b 2 ] E
| z 4 |= 2 2 2 2 2
= 2 2 2 2
=
b d + d (1 − a ) d [b + (1 − a ) ] d b 2 + (1 − a2 )2