RTD Measurement
Step-by-step Design Procedure
October 2014
Joachim Wurker
Systems & Applications Manager
Precision Delta-Sigma ADCs
RTD
Resistance Temperature Detector
• Principle of Operation:
Predictable resistance change
• Mostly made of Platinum
– linear resistance-temperature relationship
– chemical inertness
• Pt100 most common device used in industry
– Nominal Resistance R = 100Ω @ 0°C
– Sensitivity = 0.385Ω/°C (typ.)
• Slowly replacing thermocouples in many industrial applications below 600°C
due to higher accuracy, stability and repeatability
• 2-, 3-, 4-wire types
Advantages
Disadvantages
 High accuracy: < ±1°C
 Expensive
 Best stability over time
 Excitation required
 Temperature range: -200°C to +850°C
 Self-heating
 Good linearity
 Lead resistance
 Low sensitivity
R RTD
What is an RTD made of?
Resistivity
(Ohm/CMF)
• Platinum (Pt)
Metal
• Nickel (Ni)
Gold (Au)
13
Silver (Ag)
8.8
• Copper (Cu)
• Have relatively linear change in resistance over temp
• Have high resistivity allowing for smaller dimensions
• Either Wire-Wound or Thin-Film
Copper (Cu)
9.26
Platinum (Pt)
59
Tungsten (W)
30
Nickel (Ni)
36
(*)Images
from RDF Corp
Why use an RTD?
RTD Resistance vs. Temperature
Callendar-Van Dusen Equations
T  0 : RRTD (T)  R0  [1  A  T  B  T 2 ]
T  0 : RRTD (T)  R0  [1  A  T  B  T 2  C  T 3  (T  100)]
RTD Resistance vs. Temperature
450
405
Ideal  R0    T
IEC60751 Constants:
360
RTD( Temp)
RTD linear (Temp)
Resistance (Ohms)
315
R0 = 100Ω
A = 3.9083 ∙10-3
B = -5.775 ∙10-7
C = -4.183 ∙10-12
270
225
180
135
Pt100:
α = 0.385 @ 0°C
R0 = 100Ω @ 0°C
90
45
0
 200
 95
10
115
220
325
430
Temperature (C)
535
640
745
850
Different RTD Types – why do they exist? (I)
• 2-Wire RTD
–
–
–
–
2-Wire measurements are simplest to implement
Good for close proximity to RTD (RL is small)
RTD lead resistance is included in the result
Tradeoff:
• Accuracy: Error = 2∙RL∙IEXC
• Cost = Cheapest!
RL
RRTD
Red
RL
RL
• 3-Wire RTD
– Allows for RL cancellation and remote RTD
placement
– Tradeoff:
• Accuracy = Better
• Cost = More expensive
White
White
RRTD
Red
RL
Red
RL
Different RTD Types – why do they exist? (II)
• 4-Wire RTD
– Kelvin Connection:
Isolates the excitation path from the sensing path
• 2 wires carry the excitation current,
• 2 wires connect to high-impedance
measurement circuitry
– Useful when RL matching becomes difficult to
implement
– Tradeoff:
• Accuracy = Most accurate
• Cost = Most expensive
RL
White2
RL
White1
RRTD
Red1
RL
Red2
RL
Voltage, Non-ratiometric
Voltage, Ratiometric
Current, Ratiometric
RTD EXCITATION METHODS
RTD Excitation Methods (I)
Voltage, Non-ratiometric
3.3 V
0.1 mF
VREF
AVDD
RBIAS
RLEAD1
AIN0
3-wire RTD
PGA
RLEAD2
AIN1
24-bit
ΔΣ ADC
Mux
RLEAD3
AIN2
Internal
Reference
RREF
AIN3
AVSS
RTD Excitation Methods (I)
Voltage, Non-ratiometric
• Step 1: Measure V3 to determine excitation current (IEXC = V3/RREF)
• Step 2: Measure V2 to determine lead resistance
• Step 3: Measure V1 to determine RTD resistance
3.3 V
0.1 mF
VREF
AVDD
RBIAS
RLEAD1
3-wire RTD
RLEAD2
+
V2
–
AIN0
+
V1
–
PGA
AIN1
24-bit
ΔΣ ADC
Mux
RLEAD3
AIN2
IEXC
RREF
+
V3
–
Internal
Reference
AIN3
AVSS
RTD Excitation Methods (I)
Voltage, Non-ratiometric
• Step 1: Measure V3 to determine excitation current (IEXC = V3/RREF)
• Step 2: Measure V2 to determine lead resistance
(RLead = V2/IEXC)
• Step 3: Measure V1 to determine RTD resistance
3.3 V
0.1 mF
VREF
AVDD
RBIAS
RLEAD1
3-wire RTD
RLEAD2
+
V2
–
AIN0
+
V1
–
PGA
AIN1
24-bit
ΔΣ ADC
Mux
RLEAD3
AIN2
IEXC
RREF
+
V3
–
Internal
Reference
AIN3
AVSS
RTD Excitation Methods (I)
Voltage, Non-ratiometric
• Step 1: Measure V3 to determine excitation current (IEXC = V3/RREF)
• Step 2: Measure V2 to determine lead resistance
(RLead = V2/IEXC)
• Step 3: Measure V1 to determine RTD resistance
(RRTD = V1/IEXC - RLead)
3.3 V
0.1 mF
VREF
AVDD
RBIAS
RLEAD1
3-wire RTD
RLEAD2
+
V2
–
AIN0
+
V1
–
PGA
AIN1
24-bit
ΔΣ ADC
Mux
RLEAD3
AIN2
IEXC
RREF
+
V3
–
Internal
Reference
AIN3
AVSS
RTD Excitation Methods (II)
Voltage, Ratiometric
3.3 V
0.1 mF
VREF
AVDD
RBIAS
RLEAD1
AIN0
3-wire RTD
PGA
RLEAD2
AIN1
Mux
RLEAD3
AIN2
REFP0
Reference
Mux
RREF
REFN0
AVSS
24-bit
ΔΣ ADC
RTD Excitation Methods (II)
Voltage, Ratiometric
• Step 1: Measure V2 to determine lead resistance
• Step 2: Measure V1 to determine RTD resistance
3.3 V
0.1 mF
VREF
AVDD
RBIAS
RLEAD1
3-wire RTD
RLEAD2
+
V2
–
AIN0
+
V1
–
PGA
AIN1
Mux
RLEAD3
AIN2
IEXC
REFP0
Reference
Mux
RREF
REFN0
AVSS
24-bit
ΔΣ ADC
RTD Excitation Methods (II)
Voltage, Ratiometric
• Step 1: Measure V2 to determine lead resistance
• Step 2: Measure V1 to determine RTD resistance
3.3 V
0.1 mF
VREF
AVDD
RBIAS
RLEAD1
3-wire RTD
RLEAD2
+
V2
–
AIN0
+
V1
–
PGA
AIN1
Mux
RLEAD3
AIN2
IEXC
REFP0
Reference
Mux
RREF
REFN0
AVSS
24-bit
ΔΣ ADC
RTD Excitation Methods (II)
Voltage, Ratiometric
• Step 1: Measure V2 to determine lead resistance
• Step 2: Measure V1 to determine RTD resistance
• Code ~ RRTD/RREF
3.3 V
0.1 mF
VREF
AVDD
RBIAS
RLEAD1
3-wire RTD
RLEAD2
+
V2
–
AIN0
+
V1
–
PGA
AIN1
Mux
RLEAD3
AIN2
IEXC
Code VRTD  Gain

=
2n
2  VREF

I  RRTD  Gain RRTD  Gain

2  I  RREF 
2  RREF
REFP0
Reference
Mux
RREF
REFN0
AVSS
24-bit
ΔΣ ADC
RTD Excitation Methods (III)
Current, Ratiometric
RTD Excitation Methods (III)
Current, Ratiometric
• Step 1: Measure V1 to determine RTD resistance
• Code ~ RRTD/RREF
RTD Excitation Methods (III)
Current, Ratiometric
• Step 1: Measure V1 to determine RTD resistance
• Code ~ RRTD/RREF
Code VRTD  Gain I  RRTD  Gain RRTD  Gain



2n
2  VREF
2  (2 I  RREF )
4  RREF
RTD Excitation Methods (III)
Current, Ratiometric
• Step 1: Measure V1 to determine RTD resistance
• Code ~ RRTD/RREF
RLEAD Cancellation:
IIDAC1  IIDAC2  I
RLEAD1  RLEAD 2  RLEAD 3  RLEAD
VIN  I  (RLEAD  RRTD )  2I  (RLEAD  RREF )
VIN  3I  RLEAD  2I  RREF  I  RRTD
VIN-  I  RLEAD  2I  (RLEAD  RREF )
VIN-  3I  RLEAD  2I  RREF
V1  VIN - VIN-  I  RRTD  VRTD
3-wire RTD Measurement
Step-by-Step Design Procedure
ADS1247 EXAMPLE
TIPD120
Step-by-Step Example with ADS1247
• System Requirements:
–
–
–
–
RTD Type: 3-Wire Pt100
Temperature Range: -200°C to 850°C
Supply Voltage: AVDD = 3.3V
50/60Hz Rejection
• Design Considerations:
–
–
–
–
–
Data Rate
IDAC magnitude
RREF
Gain
Low-pass Filters
Selecting Data Rate
• At Data Rate = 20SPS or less, the ADS1247 offers simultaneous
50/60Hz rejection
Selecting IDAC Value
• Larger values produce larger signals – better resolution! (good, right?)
• Causes for concern with large IDAC values:
1. Self-heating of RTD
2. IDAC compliance voltage
Self-Heating Error of an RTD vs. Exc itation Current
10
• Start with 250μA – 1mA
1
Temperature (C)
0.1
0.01
Error65mW( I)
Error2.5mW( I)
3
110
4
110
5
110
6
110
7
110
5
110
3
4
110
110
I
Excitation Current (A)
0.01
Selecting RREF
• Select RREF such that VREF is ~40% to 50% of AVDD:
VREF 
AVDD
2
RREF 
VREF
2  IIDAC
• RREF should be chosen with tight tolerance and low temperature drift
– DC errors in RREF directly affect the uncalibrated measurement gain error
– Typical drift for RREF = 5 – 20ppm/°C
• Calculation:
AVDD  3.3V, IIDAC  1mA
VREF 
3.3V
= 1.65V
2
RREF 
1.65 V
 820 
2  1mA
Selecting Gain
• The largest gain will yield the best resolution per °C
• Choose gain such that ADC input signal is still less than VREF at the
max temperature
• Calculation:
RRTD @ 850 C  390 .48 
VRTD @ 850 C  390 .48   1mA  0.39048 V
Gain 
VREF
VRTD @ 850C

1.64V
 4.2
0.39048V
Input and Reference Filters (I)
Input and Reference Filters (II)
• Used to filter high-frequency noise from aliasing into ADC passband
• At 20 SPS, ADS1247 has -3 dB bandwidth of 14.8 Hz.
• f-3dB_Dif ≈ 10 x f-3dB_DR
• RTD input and reference filters should have matching cutoff
frequencies
– If noise appears on the RTD input, but not the reference, it will not be
cancelled in the ratiometric configuration
– The RTD will change resistance over temperature;
use the mid-point of the temperature range to calculate input filter
– App Note: http://www.ti.com/lit/pdf/sbaa201
IDAC Compliance
Input Common-Mode Voltage
DESIGN CHECKS
Check IDAC Compliance
Voltage Compliance  AVDD  0.7 V  2.6 V
VIDAC1  VREF  VRTD  2.6 V
RRTD @ 850 C  390 .48 
VRTD @ 850 C  390 .48   1mA  390 .48mV
VIDAC1  1.64 V  390 .48mV  2.03 V  2.6 V

Check IDAC Compliance
Voltage Compliance  AVDD  0.7 V  2.6 V
VIDAC1  VREF  VRTD  2.6 V
RRTD @ 850 C  390 .48 
VRTD @ 850 C  390 .48   1mA  390 .48mV
RP1  1k
VRP1  1k  1mA  1V
VIDAC1  1.64 V  390 .48mV  1V  3.03 V  2.6 V

Check Input Common-Mode
VCM (MIN )  VREF  1.64 V
VCM (MAX )  VREF 
VRTD (MAX )
0.1V 
2
 1.64 V 
390.48mV
 1.84 V
2
390.48mV  4
390.48mV  4
 VCM  3.3V  0.1V 
2
2
881mV  VCM  2.32V

ADS1247 - Things to be Aware of
 Place bypass cap on VREFOUT pin.
 Turn internal VREF ON, otherwise IDACs will not work.
 Check IDAC compliance is met.
 Check PGA common-mode voltage range is met.
Single-ended measurements are NOT possible using a unipolar supply!
 Start a measurement only after the input signal has settled.
Especially important when MUX'ing channels.
 Configure SPI interface for SPI Mode 1.
 Convert twos-complement correctly.
 Don't use absolute IDAC value to calculate RRTD.
Otherwise measurement is not ratiometric anymore.
RRTD 
Code 4  RREF

2n
Gain
OPEN SENSOR DETECTION
Open Sensor Detection, Lead 1
3.3 V
0.1 mF
AVDD
IDAC
RP1
RF1
3-wire RTD
RLEAD2
RF2
IDAC1
CCM1
AIN0
CDIF1
Mux
PGA
AIN1
CCM2
RLEAD3
RP2
RF3
IDAC2
CCM3
REFP0
IIDAC2
RREF
RF4
Reference
Mux
CDIF2
VREF = ½ VREF_NOM
REFN0
CCM4
AVSS
24-bit
ΔΣ ADC
+FS
Reading
Open Sensor Detection, Lead 2
3.3 V
0.1 mF
AVDD
IDAC
RP1
RLEAD1
RF1
3-wire RTD
RF2
IDAC1
CCM1
AIN0
CDIF1
Mux
PGA
AIN1
CCM2
RLEAD3
RP2
RF3
IDAC2
CCM3
REFP0
IIDAC1
RREF
RF4
Reference
Mux
CDIF2
VREF = ½ VREF_NOM
REFN0
CCM4
AVSS
24-bit
ΔΣ ADC
-FS
Reading
Open Sensor Detection, Lead 3 (I)
3.3 V
0.1 mF
AVDD
IDAC
RP1
RLEAD1
RF1
RLEAD2
RF2
3-wire RTD
IDAC1
CCM1
AIN0
CDIF1
Mux
PGA
AIN1
CCM2
RP2
RF3
RREF
RF4
IDAC2
CCM3
REFP0
Reference
Mux
CDIF2
VREF = 0V
REFN0
CCM4
AVSS
24-bit
ΔΣ ADC
Ouput
undertemined
Open Sensor Detection, Lead 3 (II)
• Option1: Measure VREF
3.3 V
– Reconfigure MUX1 [2:0] to
measure VREF
– Detect that VREF is below certain
threshold.
0.1 mF
AVDD
IDAC
RP1
RLEAD1
RF1
RLEAD2
RF2
3-wire RTD
IDAC1
CCM1
AIN0
CDIF1
Mux
PGA
AIN1
CCM2
• Option 2: REFP0 as GPIO
– Diagnostic cycle – stop
conversions, set REFP0 as GPIO
– If = high, reference is present
– If = low, reference is absent
**Make sure GPIO theshold is met
RP2
RF3
RREF
RF4
IDAC2
CCM3
REFP0
Reference
Mux
CDIF2
VREF = 0V
REFN0
CCM4
AVSS
24-bit
ΔΣ ADC
Ouput
undertemined
IDAC CHOPPING
Errors due to IDAC Mismatch
• IDACs exhibit initial mismatch and drift mismatch
• Two potential errors:
– Gain Error
Only one IDAC is flowing through RTD but both IDACs flow through RREF.
Calculation assumes both IDACs are equal.
– No 100% RLEAD cancellation
Improved Implementation
• Eliminates gain error due to IDAC mismatch.
• Same current that excites RTD, is flowing through RREF.
• Harder to maintain IDAC compliance voltage,
especially when using a 3.3V supply
3.3 V
0.1 mF
AVDD
IDAC1
REFP0
Reference
Mux
RREF
REFN0
RLEAD1
IIDAC1
AIN0
3-wire RTD
Mux
RLEAD2
IIDAC2
RLEAD3
PGA
AIN1
IDAC2
RBIAS
AVSS
24-bit
ΔΣ ADC
IDAC Chopping
• IDAC “Chopping”
– Two measurements with IDACs swapped are taken and averaged
– Improves RLEAD compensation
3.3 V
0.1 mF
AVDD
VIN(1)  IIDAC1  (RLEAD1  RRTD )  IIDAC2  RLEAD2
IDAC1
VIN( 2 )  IIDAC2  (R LEAD1  RRTD )  IIDAC1  RLEAD2
REFP0
Reference
Mux
RREF
REFN0
RLEAD1  RLEAD2  RLEAD
RLEAD1
IIDAC1
AIN0
RLEAD2
IIDAC2
AIN1
3-wire RTD
VIN 
VIN(1)  VIN( 2 )
2

IIDAC1  RRTD IIDAC2  RRTD

2
2
Mux
RLEAD3
PGA
IDAC2
RBIAS
AVSS
IIDAC1  RRTD
I
 RRTD
 Gain IDAC2
 Gain
VIN  Gain
R
 Gain
2
2


 RTD
2  VREF
2  IIDAC1  RREF
2  IIDAC2  RREF
2  RREF
• Note: Input filters need to settle before beginning a new conversion
24-bit
ΔΣ ADC
Achievable Resolution with ADS1248
• LSB size:
2  VREF
2  1.65 V

 49.2nV
224  Gain
224  4
• Input referrerd Noise:
3.83uVpp
• Pt100 Sensitivity:
1mA x 0.385Ω/°C = 0.385mV/°C
• Temperature Resolution per Code:
49.2nV / 0.385mV/°C = 0.0001°C
• Noise Free Temperature Resolution: 3.83uVpp / 0.385mV/°C = 0.01°C