Sensors and Actuators

Introduction to sensors

Sander Stuijk
(s.stuijk@tue.nl)

Department of Electrical Engineering

Electronic Systems
2

TEMPERATURE SENSORS
(Chapter 3.5, 5.1, 5.4, 5.9-5.12, 16.1)
3 Applications
4 Applications
5 Temperature sensors

placement excitation physical effect material thermal

sensor
contact passive thermal expansion metal bimetal
contact active resistive effect metal RTD
contact active resistive effect semiconductor silicon
resistive
contact active resistive effect polymer or thermistor
ceramic
contact passive thermoelectric effect conductor thermocouple
contact active PN junction semiconductor
non-contact passive pyroelectric effect pyroelectric pyroelectric
non-contact active ultrasound piezoelectric acoustic

there are many other classification criteria: construction, linearity,

reference point, ...
6 Temperature sensors

 temperature sensors are deceptively simple

 resistive sensor – a conductor connected to a voltage source
 thermocouple – any two dissimilar materials welded together at
one end and connected to a micro-voltmeter

 temperature sensors can be used to measure other quantities, e.g.,

Quantity
S
e t Acceleration / Vibration Flow rate / Point velocity Force
n y
Resistive Mass-spring + strain gage Thermistor Strain gage
s p
o e Self-generating Mass-spring + piezoelectric Thermal transport + Piezoelectric sensor
r sensor thermocouple

 complex sensors for radiation, pressure, position, level, and

chemical reactions can be constructed on the basis of temperature
or temperature-difference sensors
7 Temperature sensors

placement excitation physical effect material thermal

sensor
contact passive thermal expansion metal bimetal
contact active resistive effect metal RTD
contact active resistive effect semiconductor silicon
resistive
contact active resistive effect polymer or thermistor
ceramic
contact passive thermoelectric effect conductor thermocouple
contact active PN junction semiconductor
non-contact passive pyroelectric effect pyroelectric pyroelectric
non-contact active ultrasound - acoustic
8 Resistance

V
 resistance of a material is defined as R 
i
l
 resistance depends on geometrical factors R  
a
 length of wire (l)
 cross-sectional area (a)

l m l
 resistance depends on temperature R    2
a ne  a
 number of free electrons (n)
 mean time between collisions (τ)

 changing dimensions affect resistance (piezoresistive effect)

 changing temperature affect resistance (thermoresistive effect)
 resistive sensor can be used to sense changes in these quantities
 only one quantity should be measured
 other quantity should be kept constant or corrected for
9 Thermoresistive effect
 specific resistivity of a material
m

ne 2
 temperature dependency through
 number of free electrons (n) in semiconductors
 NTC behavior: ↑T → ↑n → ρ↓
 mean time between collisions (τ) in conductors
 PTC behavior: ↑T → ↓τ → ρ↑
semiconductor (silicon) conductor (wolfram)
10 Resistance temperature detectors (RTDs)

 RTD is a temperature sensor build from metal (conductor)

 specific resistivity of metal depends on time between collisions
 positive temperature coefficient (PTC)
 increasing temperature leads to increasing resistivity

 relation between temperature and resistance

RT  R0 [1  1 T  T0    2 T  T0   ...   n T  T0  ]
2 n
PT100 RTD
 T0 – reference temperature
 R0 – resistance at T0

 coefficients can be found using calibration

 example – PT100 RTD
 α1 ≈ 3.89∙10-3/K, α2 ≈ -5.83∙10-7/K2, α3 ≈ 1.92∙10-7/K3
 almost linear relation between temperature and resistance
11 Temperature coefficient of resistance

 assume linear temperature/resistance relation for an RTD sensor

RT  R0 [1   T  T0 ]

 α is called the temperature coefficient of resistance (TCR)

RT  R0

(T  T0 ) R0
 TCR indicates relative change in resistance per unit temperature
(between temperature T and a reference temperature T0)
 TCR is often called relative sensitivity

 PTC devices: TCR decreases when temperature increase (why?)

 fractional increase in resistance decreases with increasing
temperature
 limits usability of PTC devices at higher temperatures
12 Temperature coefficient of resistance

example – PT100 sensor

 R0 = 100Ω, α0 = 0.00389 (Ω/Ω)/K at 0°C, RT=R0[1+ α0(T-T0)]
 what is the sensitivity of this sensor?
 sensitivity is the slope of the resistance-temperature curve
 resistance-temperature curve is straight line
 sensitivity S   0 R0   25R25  50R50  0.00389( / ) / K 100  0.389 / K
 what are TCR at 25°C and 50°C? (use 0°C as reference)
 TCR (at 25°C)
 0 R0  0 R0 0
 0 R0   25R25   25   
R25 R0 [1   0 (250 C  00 C )] 1   0 (250 C )
0.00389( / ) / K
  0.00355( / ) / K
1  0.00389 / K (250 C )
0
 TCR (at 50°C)  50   0.00326( / ) / K
1   0 (50o C )

 TCR decreases for increasing temperature

16 Self-heating in RTDs

 current must be passed through sensor to measure resistance

 power will be dissipated in the RTD creating heat (self-heating)
PD = I2R

I R

 effect of self-heating reduced by thermal dissipation to environment

 heat dissipation factor δ (W/K) depends on
 surrounding fluid
 velocity of the fluid
PD I 2 R
 temperature error given by T  
 

 self-heating error can be limited by dimensioning the current I

17 Self-heating in RTDs

example – PT100 sensor R(T)=R0[1+ α0(T-T0)]

 R0 = 100Ω, α0 = 0.00389 (Ω/Ω)/K at 0°C I R
 δ = 6mW/K (in air), δ = 100mW/K (in still water)
 sensor used in range [0°C, +100°C]
 what is the maximal current through the sensor to keep the
self-heating error below 0.1°C?

PD I 2R
 temperature PT100 above environment T  
 

 self-heating error maximal when resistance is maximal

 maximal self-heating at +100°C
 resistance R(100°C) = 100Ω[1 + 0.00389/°C ∙100°C] = 139 Ω
18 Self-heating in RTDs

example – PT100 sensor R(T)=R0[1+ α0(T-T0)]

 R0 = 100Ω, α0 = 0.00389 (Ω/Ω)/K at 0°C I R
 δ = 6mW/K (in air), δ = 100mW/K (in still water)
 sensor used in range [0°C, +100°C]
 what is the maximal current through the sensor to keep the
self-heating error below 0.1°C?

PD I 2R
 temperature PT100 above environment T  
 
T  
 relation between current and temperature I
R

(0.1o C )  (0.006W / K )
 max current in air I  2.1mA
139
(0.1o C )  (0.1W / K )
 max current in water I   8.5mA
139
19 Self-heating in RTDs
R1
example – PT100 sensor R(T)=R0[1+ α0(T-T0)]
Vr
 R0 = 100 Ω, α0 = 0.00389 (Ω/Ω)/K at 0°C
R2 vo
 sensor used in range [0°C, +100°C]
 δ = 6 mW/K (in air), Vr = 5 V, R2 = 1 kΩ
 what is the maximal self-heating error (resolution) of this
sensor?
PD I 2 R1
 temperature PT100 above environment T  
 
 current I depends on resistance R1 and temperature T
Vr Vr
I 
R1  R2 R0 1   0 T  T0   R2

 maximal current when T = 0°C, but minimal resistance

 temperature error depends on power dissipation
 maximal power dissipation when T = 100°C
 maximal self-heating error occurs when T = 100°C
20 Self-heating in RTDs
R1
example – PT100 sensor R(T)=R0[1+ α0(T-T0)]
Vr
 R0 = 100 Ω, α0 = 0.00389 (Ω/Ω)/K at 0°C
R2 vo
 sensor used in range [0°C, +100°C]
 δ = 6 mW/K (in air), Vr = 5 V, R2 = 1 kΩ
 what is the maximal self-heating error (resolution) of this
sensor?
PD I 2 R1
 temperature PT100 above environment T  
 
 maximal self-heating error occurs when T = 100°C

 
R 100 C 139 
V 2
1 V R1 1   T  0.5 C
2
T  r
 2
r
R22  R2  
R1  R2  
R1

dominates when R1 << R2

21 Lead-wire resistance
 lead wires are not perfect conductors (lead-wire resistance)
 resistance of the wires will affect measured voltage
RL i
sensor
VPT100 PT100 interface
Vout
RL circuit

 temperature error due to lead-wire resistance when interface circuit

provides constant current i
R  2RL  PT 100  R  2RL 
 2 RL
R   T 
T  S
S 
 example – PT100
 PT100 has resistance of 107.8Ω at 20°C and S = 0.389Ω/K
 assume RL = 1Ω
 ΔT = +5.1°C → interface circuit measures temperature of 25°C
 measured temperature 25% above actual temperature
22 Lead-wire resistance
 lead wires are not perfect conductors (lead-wire resistance)
 resistance of the wires will affect measured voltage
RL i
sensor
VPT100 interface
PT100 Vout Vout  VPT 100  2RLi
RL circuit

 lead wire resistance can be cancelled with 4-wire sensing method

 interface circuit has high impedance
RL i
sensor
RL i=0
interface current
VPT100 PT100
RL
Vout
circuit source Vout  VPT 100
i=0

RL

 4-wire sensing method requires stable current source

 6-wire sensing method can be used with stable voltage source
23 Summary - Resistive temperature detectors (RTDs)

 temperature-resistance relation RT  R0 [1   T  T0 ]
 α is called the temperature coefficient of resistance (TCR)
dRT dT R  R0
  T
RT (T  T0 ) R0
 TCR indicates relative change in resistance per unit temperature
(between two reference temperatures)
 TCR is not equal to the sensitivity

 several error sources influencing accuracy

 lead-wire resistance (RL)
interface circuit
 self-heating (RT)
RL
 non-linearity (RT) sensor
 loading effect (Rm) RT PT100 Rm Vr
RL
24 Interface circuits

 interface circuits can be used to

 increase sensor sensitivity
 cancel lead-wire resistance
 linearization of the sensor output
 limit self-heating
 compensate for error sources (e.g. strain or temperature)
 set output voltage at reference point
 amplify sensor output for use with AD-converter
 ...

+
-

Vout
25 Resistive sensors

 resistance of resistive sensor R  R0 f ( x)

 f(x) – fractional change in resistance (with f(0) = 1)

 resistance of linear resistive sensor R  R0 1  x 

 range of x depends on type of sensor
 [-1, 0] – linear potentiometer
 [1, 10] – RTDs
 [0.00001 , 0.001] – strain gauges
 [1, 100] – NTC thermistors
 [1, 10000] – switching PTC thermistors

 requirements on signal conditioners for resistive sensors

 electric voltage or current must be applied
 supply and output voltage/current are limited by self-heating
26 Resistive sensors – current excitation

 resistive sensor connected to Norton equivalent circuit

 sensor driven by current source

Io Ro R
 when does maximal self-heating error occur?
 dissipation maximal when R is maximal

 deflection measurement with current source + Vr

Ir -
 feedback loop enforces constant current
Vr
Ir  R
Rr vo
 output of a linear sensor

R0 1  x 
Vr
vo  I r R  Rr
Rr
 choose R0 = Rr then vo  Vr 1  x   Vr  xVr
 output consists of offset and (small) fluctuation around offset
27 Resistive sensors – current excitation

example – circuit for temperature measurement [20°C,100°C]

 measure temperature with 0.1°C resolution (self-heating < 0.1°C)
 PT 100 sensor (R0=100Ω and α=0.00389Ω/Ω/K at 0°C)
 dissipation factor δ = 40mW/K in 0.4m/s water
 reference voltage Vr = 5V
 what resistance should Rr have to get a sensitivity of 1mV/°C?
 temperature resolution limited by self-heating + Vr
2
Ir -
I R 2
V  R
  r   T  0.1 C
r 

  Rr   R vo
 maximal dissipation at 100°C, condition is thus
Rr  Vr
R100 
  (0.1 C ) 
 Rr
R100  R0 1   T  T0  
138.9
 1001  0.00389 100C    Rr  5V   932
 (40mW / K )  (0.1 C )
 138.9 
28 Resistive sensors – current excitation

example – circuit for temperature measurement [20°C,100°C]

 measure temperature with 0.1°C resolution (self-heating < 0.1°C)
 PT 100 sensor (R0=100Ω and α=0.00389Ω/Ω/K at 0°C)
 dissipation factor δ = 40mW/K in 0.4m/s water
 reference voltage Vr = 5V
 what resistance should Rr have to get a sensitivity of 1mV/°C?
 output voltage of the sensor + Vr
Ir -
vo  r R0 1  x   r R0 1  T 
V V
Rr Rr
R vo
 sensitivity is equal to

dvo Vr Vr R0 (5V )(100)(0.00389 / K )

S  R0  R    1945
dT Rr
r
S 1mV / K Rr
 sensor output has also an offset (output not 0V at 20°C)
Vr
R20  107.8  Voffset  R20  277mV
Rr
29 Interface circuits

 resistance of linear resistive sensor: R(x) = R0(1+x)

 range of x depends on type of sensor

 requirements on signal conditioners for resistive sensors

 electric voltage or current must be applied
 supply and output voltage/current are limited by self-heating

 current excitation
 maximal self-heating when R maximal
Io Ro R
 maximal sensitivity when R maximal
Ro
 voltage excitation
 when does maximal self-heating error occur?
 when is sensitivity maximal? Vo R
 when is non-linearity error minimized?
30 Voltage divider – self-heating error

 sensor driven by voltage source Rr

 sensor: R Vr
 load resistance: Rr
R vo
 when does maximal self-heating error occur?
 power consumption by sensor
2
 Vr 
P    R
 R  Rr 
 maximal power consumption occurs when
2 2
dP Vr  Vr  Vr   Vr  Rr  R
2 R        0  Rr  R
dR R  Rr R  Rr 2  R  Rr   R  Rr  R  Rr

 self-heating error is maximal when Rr = R

 power consumption is then equal to
2
 Vr  Vr2
P    Rr 
 Rr  Rr  4 Rr
31 Voltage divider – self-heating error
example – dimension voltage divider for temperature measurement
 measure temperature from 0°C to 100°C
 PT 100 sensor (R0=100Ω and α=0.00389Ω/Ω/K at 0°C)
 maximal power dissipation in sensor is 1mW
 voltage source Vr = 5V
 what resistance Rr must be used for this voltage divider?
 power dissipation in sensor
2
 Vr 
  R  1mW Rr
 R  Rr 
Vr
 maximal dissipation when R=Rr
R vo
5V   6.25k
2
 Vr  2
V 2
  Rr  1mW  Rr  r

 Rr  Rr  2  0.001W 2  0.001W

 sensor range is from 100Ω to 139Ω

 always R < Rr, thus power dissipation always below limit
32 Voltage divider – linearity
 measure fractional change in resistance x
 sensor: R = R0(1+x) Rr
 load resistance: Rr = R0k Vr
 output voltage of the circuit
R vo
R0 (1  x)
Vr  1  x Vr
R
vo  Vr 
Rr  R R0 k  R0 (1  x) k 1 x
 response becomes linear when Rr >> R (i.e. k >> 1+x)
33 Voltage divider – linearity

|vo/Vr| Rr
k=0.1
Vr
k=1
R vo

k=10

k=100

 increasing k is good for linearity, but what about sensitivity?

34 Voltage divider – sensitivity
 measure fractional change in resistance x
 sensor: R = R0(1+x) Rr
 load resistance: Rr = R0k Vr
 sensitivity
R vo
dvo d  1 x  (k  1  x)  (1  x) k
S   Vr   V  V
dx  k  1  x  ( k  1  x) (k  1  x)
2 r 2 r
dx
 maximal sensitivity use quotient rule
dS d  k  d j ( x ) j ' ( x ) h ( x )  j ( x ) h' ( x )
 0   0 
dk dk  (k  1  x)
V
2 r
 dx h( x) h( x)2
( k  1  x ) 2  2k ( k  1  x ) x 1 k
   0  k  x 1
( k  1  x) 4
( k  1  x) 3

 maximal sensitivity reached when R = Rr

 same situation as when self-heating error is maximal
 maximal transfer of power (at R = Rr) leads to
 maximal sensitivity and maximal self-heating
35 Voltage divider – sensitivity and linearity

S Rr
Vr
R vo
k=1

k=0.1 k=10
k=100

 for many sensors x < 1

 sensitivity largest for k = 1
 sensitivity may be considered constant if maximal value of x << 1
36 Voltage divider – output voltage

 maximal sensitivity when k = 1 Rr

 output voltage Vr
1 x 1 x R
vo  Vr  Vr vo
k 1 x 2 x

 offset voltage present in output