Psoc 3 and Psoc 5Lp - Temperature Measurement With An RTD

AN70698
PSoC 3 and PSoC 5LP Temperature Measurement with an RTD

Author: Praveen Sekar and Todd Dust
Associated Project: Yes
Associated Part Family: All PSoC 3 and PSoC 5LP Parts
Software Version: PSoC Creator 2.2 SP1 and
higher
Related Application Notes: AN75511, AN66477, AN60590
AN70698 explains the theory of temperature measurement using an RTD, and then shows how to do so with a single
PSoC 3 or PSoC 5LP without the need for external ADCs or amplifiers. To make it easy to calculate temperature from
ADC readings, PSoC Creator provides an RTD Component. Four example projects are included to demonstrate operation
with both low and high levels of accuracy and resolution.
Contents
Introduction ....................................................................... 2
RTD Theory of Operation ............................................... 2
RTD Resistance Measurement Methods........................... 3
Two-Wire Measurement ............................................... 3
Three-Wire Measurement............................................. 3
Four-Wire Measurement............................................... 4
Reference Resistor Method ............................................... 4
Reference Resistor Selection ....................................... 5
Offset Error Cancellation .............................................. 5
Gain Error Cancellation ................................................ 5
RTD Resistance-to-Temperature Conversion ................ 5
Positive Temperatures.................................................. 5
Negative Temperatures ................................................ 6
Choosing the Right Polynomial Order .......................... 6
RTD Component .......................................................... 6
RTD Temperature Measurement with PSoC ..................... 7
Hardware Used CY8CKIT-025 EBK .......................... 7
Project Description ............................................................ 8
Firmware Flow .............................................................. 9
Testing the Project ..................................................... 10
Interfacing Multiple RTDs ................................................ 10
Broken RTD Reconfiguration .......................................... 10
Performance Measures ................................................... 10
Temperature Resolution ............................................. 10
Temperature Accuracy ............................................... 11
List of all Errors .......................................................... 14
Test Results ............................................................... 14
Summary ......................................................................... 15
About the Author ............................................................. 15
www.cypress.com
Appendix A ...................................................................... 16
Appendix B ...................................................................... 18
Broken RTD reconfiguration ....................................... 18
Worldwide Sales and Design Support ............................. 26
Document No. 001-70698 Rev. *H
Introduction
Temperature is one of the most frequently measured
environmental variables. Temperature measurement is
typically done using one of four sensors: resistance
temperature detector (RTD), thermocouple, thermistor,

and diode. Table 1 compares these four sensor types.
Table 1. Comparison of Temperature Sensors

Parameter
RTD
Thermocouple
Thermistor
Diode
Temperature range
200 to +850
250 to +2350
100 to +300
50 to +150
Sensitivity at 25 C
0.387 /C
40 V/C (K-type)
416 /C
250 V /C
Accuracy
High
Medium to High
Medium
Low
Linearity
Good
Fair
Poor
Good
Typical cost (US $)
$3 $80
$3 $15
$0.2 $10
<$0.2
Typical distance of
sensing
Surface mount for

on-board temperature
<100 meters
Surface mount for

on-board temperature
On-board temperature
three- and four-wire up to

a few hundred meters
Leaded for <1 meter
Resource requirement
Excitation current,
amplifier, ADC, reference
resistor
Amplifier, ADC, voltage

reference, and another
temperature sensor for
cold junction
Excitation current, ADC,

reference resistor
Excitation current,
amplifier, ADC
Response time
Slow
Fast
Fast
Slow
Computational
complexity (best
possible accuracy)
High
Very high
Very high
Medium
Although they are more expensive than other sensor

types, RTDs have the best accuracy over a wide
temperature range.
RTDs are primarily made of platinum, which gives them
good linearity and repeatability. Commonly used RTDs
include PT100, PT500, and PT1000. PT stands for
platinum, and the number (100/500/1000) indicates the
resistance value at 0 C. This application note focuses on
measurements based on PT100, which has the best
linearity and repeatability and is the most widely used.
PSoC 3 and PSoC 5LP can fully integrate the hardware
required for RTD measurement. Not only can they
integrate the hardware, but PSoC 3 and PSoC 5LP
measure RTDs accurately. On top of this measurement
front end, you have all of the other resources of the PSoC
3 and PSoC 5LP available to complete the rest of your
product. PSoC 3 and PSoC 5LP are attractive devices for
reducing BOM cost and component count in your RTD
designs, while maintaining the accuracy and resolution
you need.
www.cypress.com
RTD Theory of Operation

An RTD has a positive temperature coefficient (PTC);
resistance increases as temperature increases.
The resistance-temperature relationship is not perfectly
linear. Various standards approximate this non-linearity.
Of them, IEC 60751 is one of the most widely used
standards. Equation 1 and Equation 2 define the
resistance to temperature relationship in IEC 60751.
Above 0 C, RTD temperature is specified by the RTD
resistance at 0 C (R0) and constants A and B.
RT R0 (1 AT BT 2 ) , T > 0
Equation 1
Where, RT is the resistance at T C

Below 0 C, in addition to the A and B, a third constant (C)
is involved, as shown in Equation 2.
RT R0 (1 AT BT 2 C (T 100)T 3 ),
T < 0
Equation 2
Equation 1 and Equation 2 are also referred to as

CallendarVan Dusen equations, and A, B, and C
coefficients are also known as CallendarVan Dusen
coefficients; the term is named after the scientists who
discovered and refined the equations.
The values of A, B, and C for PT100 RTD are specified in
IEC 60751 for standard industry-grade platinum and are:
Figure 2. Two-Wire RTD Temperature Measurement

Rw1
Is
VRTD
RRTD
A 3.9083*103 C 1
Rw2
B 5.775*107 C 2
In Figure 1 the Linear Approximation is made by assuming

the RTD changes by 0.385/C. A 1-ohm error in the
resistance measurement leads to a temperature error of
about 2.6 C(1/.385). Therefore, while measuring RRTD,
the RTD resistance has to be determined with an accuracy
of <0.0385 for the temperature error to be <0.1 C.
C 4.183*1012 C 4
and the resistance at 0 C,
R0 100
(PT100 RTD)
Plotting the resistance versus the temperature yields a

nearly linear curve, which is shown in Figure 1. However,
it is not a perfectly straight line. Figure 1 also shows the
straight line approximation imposed on the RTD curve.
Figure 1. RTD Resistance versus Temperature
Because of the length of the wire connected in series with

the RTD, the RTD wire resistances (Rw1 and Rw2) get
added to the RTD resistance. This wire resistance
contributes to an error in the final temperature
measurement. Equation 3 shows the error components,
Rw1 and Rw2, in the voltage equation.
VRTD IS * (RRTD RW1 RW 2)

Equation 3
Rw1 and Rw2 vary based on the length of the wire. For very
short wires, they could be in the milliohms; for very long
wires, 10s of ohms.
One way to minimize the effect of wire resistance is to use
the three-wire method described in the next section.
Three-Wire Measurement
RTD temperature measurement involves two steps:
1.
Measure the RTD resistance accurately
2.
Convert the measured resistance accurately to

temperature.
The following sections describe the two tasks.
RTD Resistance Measurement

Methods
With a three-wire measurement, the error due to the wire

resistances is eliminated by measuring the wire
resistances and subtracting it from the measured RTD
resistance (see Equations 4, 5, and 6).
In Figure 3, the switch is set to position 1 and the voltage
across the RTD, VRTD, is measured. Then, the switch is set
to position 2, and the voltage across the wire resistance,
Vwire, is measured.
VRTD = Is * (RRTD + Rw1 + Rw3)
Equation 4
Vwire = Is * (Rw2 + Rw3)
Equation 5
If Rw1 = Rw2,
The techniques commonly used to measure resistance,

and their effectiveness for RTDs, are described in the
following section.
RRTD = (VRTD Vwire) / Is
Equation 6
Two-Wire Measurement
In a two-wire measurement, a current is passed through
the RTD and the voltage across the RTD is measured, as
shown in Figure 2.
www.cypress.com
Figure 3. Three-Wire RTD Temperature Measurement
To overcome the gain/offset error caused by the ADC and

the current DAC (IDAC), add a reference resistor to your
design.
Rw1
Reference Resistor Method
2
Is
VRTD
Rw2
RRTD
The reference resistor makes the measurement error

independent of both the current source accuracy (gain
error,), and the ADC accuracy (gain error). The
measurement error depends primarily on a reference
resistance.
Vwire
Rw3
C a ve a t
Although this method is better than the previous method,
the three-wire method gives accurate results only if Rw1 =
Rw2.
To avoid all error associated with the wire resistances, use
the four-wire measurement described next.
The schematic of this method is shown in Figure 5. A

constant current is passed through a known accurate
reference resistance in series with the RTD. The voltage,
Vref, across the reference resistor and the voltage, VRTD,
across the RTD are measured. The RTD resistance is
given by Equation 8:
RRTD
Four-Wire Measurement
The four-wire measurement, as shown in Figure 4, greatly
reduces any error caused by wire resistances.
VRTD
* Rref
Vref
Equation 8
Figure 5. RTD Temperature Measurement
Figure 4. Four-Wire RTD Temperature Measurement

IDAC
Rw1
Rw2
IS
VRTD
Rref, 100 ohm

0.1%
RRTD
Rw3
Rw1
ADC
Rw3
Rw4
RRTD
In this method, a known constant current is passed

through the RTD. The voltage across it is measured using
a separate sensing path. The separate sensing path
ensures that the voltage drop across the wire resistances,
Rw1 and Rw4, does not affect the RTD voltage measured.
There will be little voltage drop across resistances Rw2 and
Rw3, which are in the ADC measurement path, because
there is negligible current flow into the high-input
impedance terminals of the ADC.
The RTD resistance in this method is given by Equation 7.
RRTD
VRTD
I RTD
Equation
To find the RTD resistance, the current source and the

ADC measuring the voltage must be accurate.
Specifically, the current source and the ADC should be
free from offset, gain and non-linearity errors. Even a
small error in voltage measurement can result in a large
temperature error at higher temperatures.
www.cypress.com
Rw4
Rw2
PSoC
The current path is shown by the bold blue line, and

voltage measurement paths are shown in red. The current
does not flow through wire resistances Rw3 and Rw4,
because of the high-input impedance of the ADC. The rest
of the application note discusses this method and the
associated project based on it. The following sections
discuss how to choose a reference resistor, and how the
offset error and the gain error of the ADC and IDAC
become null.
Reference Resistor Selection

Choose a small reference resistor that doesnt load the
IDAC. This project uses a reference resistance of 100
ohms. The IDAC has a compliance voltage of VDDA-1V.
This means that the voltage at the IDAC cannot exceed
VDDA-1V. If the maximum RTD resistance is 400 ohms,
the reference resistor is 100 ohms, the internal PSoC
routing is ~400 ohms and a 1-mA current is passed
through the three in series, then a voltage of 900 mV will
be produced (1 mA * 400 ohms + 100 ohms + 400 ohms).
This is well below the compliance voltage.
In addition, choose a reference resistor that uses the
same ADC range as the RTD. This will reduce the time it
takes to measure the RTD, because the ADC will not need
to be reconfigured. Again, if the maximum RTD resistance
is 400 ohms, and 1 mA is passed through it, the RTD will
produce a voltage of 400 mV. The DelSig ADC has an
input range of 512 mV. Keep the voltage across the
reference resistor in this 512-mV range. If it is outside this
range, you will have to reconfigure the ADC every sample;
reconfiguration takes time.
Offset Error Cancellation

In this application note, offset cancellation is done by
correlated double sampling (CDS). In CDS, the offset is
measured and then, in firmware, it is subtracted from the
other voltage measurements. Using this method reduces
the ADC sample rate by 50%, because offset is measured
every other sample.
You can use CDS to measure offset in various ways. See
AN66444 PSoC 3 and PSoC 5LP Correlated Double
Sampling for details.
For this application note, offset is measured by passing
zero current through the IDAC and measuring the voltage
directly across channel 0 or 1.
With CDS, the equation for resistance measurement
becomes:
RTD VRTD V0 * Rref

Vref V0
Equation 9
Where, VRTD is the voltage measured across the RTD

Vref is the voltage measured across the 100- resistor
V0 is the offset voltage measured when the IDAC current is
set to zero.
The offset current of the IDAC does not cause any error
because it would be nulled by the difference in Equation 9.
VRTD = (I+I0) * RRTD + ADC offset
V0 = I0 * RRTD + ADC offset
www.cypress.com
Subtracting V0 From VRTD removes ADC offset. The

difference nulls the I0 term, which is caused by the IDAC
offset. If CDS is done regularly, then offset drift is also
canceled out.
Gain Error Cancellation

Assume that the ADC has a gain error of k and the DAC
has a gain error of k. These errors reflect as multiplicative
factors in the voltage measurements, VRTD and Vref.
Because Equation 9 includes a ratio, the multiplicative
errors k and k cancel out.
RTD =
k * k ' * (VRTD V0 )
* Rref
k * k ' * (Vref V0 )
Now the error depends primarily on the accuracy of the

reference resistor, Rref.
This method also removes any errors associated with gain
drift, because the ratio metric measurement is being taken
every time.
Using the reference resistor method, you can determine
the RTD resistance accurately. The next step is to convert
the RTD resistance to temperature. This conversion is
discussed in the following section.
RTD Resistance-to-Temperature
Conversion
The straightforward method to obtain temperature from
resistance is to use the CallendarVan Dusen equations.
But Equations 1 and 2 show resistance in terms of
temperature; you need to know temperature in terms of
resistance. The solution is given below.
Positive Temperatures
Solving Equation 1 for T,
R
A A2 4 B1 T
R0
T
2B
Equation 10
The other solution of the quadratic equation is eliminated

using the known point (T = 0; R =100).
Using Equation 10 involves a square root, which requires
a math library and about 5800 8051 CPU cycles for
computation. This is a lot of time to spend doing a single
calculation.
Instead of Equation 10, use a polynomial to calculate
temperature. Compute the polynomial by first constructing
a resistance versus temperature table and then using
curve-fitting techniques on the table. Excel is a good tool
to use for curve fitting.
Polynomial computation, which does not require a math

library, executes faster. The temperature accuracy
increases as the order of the polynomial increases. Using
[1]
a fifth-order polynomial can reduce the conversion error
to <0.002 C, but, higher-order polynomials require more
CPU cycles for execution.
Negative Temperatures
Solving Equation 2 for T is not straightforward, because it
is a fourth-order equation. Again, approximate the
temperature-resistance relationship using a polynomial.
Using a fourth-order polynomial reduces the conversion
error to < 0.002 C.
A single polynomial can apply to both negative and
positive temperatures. However, doing so would require a
polynomial order greater than 10 to reduce the conversion
error to <0.002 C. Therefore, use two polynomials, one
each for positive and negative temperatures.
Choosing the Right Polynomial Order
Table 2. Accuracy and Number of Cycles Required for

Computation in Using Polynomials of Different Orders
Number of
Cycles
Required for
Computation*
Accuracy in
C (-200C to
850 C
Range)**
Accuracy in
C (-50C to
150 C
Range)**
First order
70
< 20.2
<0.55
Second
order
110
< 1.7
<0.007
Third order
150
< 0.17
<0.0001
Fourth order
190
< 0.018
Fifth order
230
< 0.002
Note: (*) The number of cycles required for computation is calculated from
the PSoC 5LP-based code in the attached project. The project uses
floating-point arithmetic to compute temperature. You can greatly reduce
the number of cycles by using special algorithms for floating-point
multiplications.
(**) This only includes the accuracy of the resistance to temperature
conversion, not the whole system accuracy
If your temperature range or accuracy requirements are

lower, you can use a lower-order polynomial and decrease
A fifth-order polynomial for temperature in terms of resistance is

5
4
3
2
in the form T = a5R + a4R + a3R + a2R + a1R + a0, where a5,
a4,a0 are constants determined by curve-fitting techniques.
www.cypress.com
Choosing and creating the appropriate polynomial is mathintensive and time-consuming. Cypress simplifies this task
by providing an RTD Component in PSoC Creator that
creates a polynomial of the required order based on your
temperature range and the accuracy required.
The component automatically calculates the required
order of the polynomial and polynomial coefficients. For
resistance-to-temperature conversion, the component
provides an API, which uses the computed polynomial
coefficients to find the temperature.
RTD Component
Figure 6 shows the RTD component. The RTD component
supports PT100, PT500, and PT1000 RTDs.
Figure 6. RTD Component
The temperature range and accuracy determine the order

of the polynomial. As discussed, a fifth-order polynomial
can reduce the error to <0.002 C in the -200 C to 850 C
range, but it requires more CPU cycles for computation.
Table 2 provides the number of cycles required for
temperature computation and the temperature error
resulting from polynomials of different orders for the
-200 C to 850 C range.
Polynomial
Order
the number of cycles required for temperature

computation. If your range is just -50 C to 150 C, you
can use a second-order polynomial to achieve a
temperature error of <0.007 C.
Double-clicking the component yields the configuration

dialog box shown in Figure 7.
Figure 7. RTD Component Customizer
Select the RTD type, enter the maximum and minimum

temperatures to which the RTD will be subjected, and
select the calculation error budget. The configuration
dialog returns an appropriate polynomial order for the
chosen error budget, the temperature calculation error
versus temperature graph, the maximum error for the
chosen temperature range, and the number of cycles
required for computation.
For example, if you are using PT100 RTD, your range is

component automatically chooses the appropriate
-100 C to 400 C, and you require 0.05 C temperature
polynomial for you.
calculation accuracy, select PT100 RTD, enter the
Figure 8 shows that a third-order polynomial meets that
maximum and minimum temperatures in their respective
requirement.
text boxes, and select 0.05C temperature accuracy. The
Figure 8. Customizer for 100 C to 400 C Range
In your code, use the RTD_GetTemperature(int32 res) API

to calculate temperature. The API takes resistance (in
milliohms) and computes the temperature using the
polynomial coefficients and returns the temperature (in
1/100th of C).
The temperature error depends on many factors in
addition to resistance-to-temperature conversion error.
The RTD customizer shows only the error due to
resistance-to-temperature conversion. To accommodate
other errors, ensure that the resistance-to-temperature
conversion error is less than one-tenth of the total error
budget. The other errors are discussed in Accuracy of the
Measurement section.
Hardware Used CY8CKIT-025 EBK

The PSoC Precision Analog Temperature Sensor EBK
(CY8CKIT-025) is used in the example project. The kit
provides four sensorsthermocouple, thermistor, RTD,
and diodefor measuring temperature. In addition,
interface slots let you plug in your own thermocouple,
thermistor, RTD, and diode. You can connect the EBK to
the CY8CKIT-030 PSoC 3 Development Kit (DVK), or to
the CY8CKIT-050 PSoC 5LP DVK. Figure 9 shows the kit
and Figure 10 shows the RTD portion. For more details on
the kit, go to www.cypress.com/go/Cy8CKIT-025.
Figure 9. PSoC Precision Analog Temperature Sensor
EBK
RTD Temperature Measurement with

PSoC
This application note has four example projects
(RTD_HighEnd, RTD_MidEnd, RTD_LowEnd, and Broken
RTD reconfiguration) to showcase RTD temperature
measurement using PSoC. The first three projects display
the RTD temperature on an LCD. These projects have the
same signal chain. This section describes the
RTD_MidEnd project in detail. Differences between the
projects are explained in detail in the Performance
Measures section. The fourth project, Broken RTD
Reconfiguration, is explained in Appendix B.
www.cypress.com
Figure 10. RTD Section of the EBK
The CY8CKIT-025 does it this way to share the reference

resistance for diode temperature measurement. Because
the reference resistance is not in series with the RTD, the
current through the reference resistance can differ from
the current through the RTD. If the currents through the
reference resistance and RTD are not equal, you can get
a temperature error. Manual analog routing is used to
avoid a significant error. See Appendix A for a detailed
explanation.
If you are designing your own board for RTD
measurement, use the reference resistor in series with the
RTD, as shown in Figure 5. You can avoid the 0.7 C error
at 850 C (see Appendix A). You require only one pin from
IDAC, and the current mux shown is not required (see
Figure 11). Use an IDAC direct connection (P0_6, P0_7,
P3_0, P3_1) to connect the series combination of RTD
and reference resistor to the IDAC. The PSoC Creator
schematic for that project is shown in Figure 12.
Project Description
The RTD_MidEnd PSoC Creator project associated with
this application note displays the RTD temperature directly
on an LCD. The schematic of the project, which uses the
CY8CKIT-025 PSoC Precision Analog Temperature
Sensor EBK, is shown in Figure 11. This figure is similar to
Figure 5 except that the reference resistance is in a
separate path, and is not in series with the RTD.
Figure 11. PSoC Creator Top Design Schematic
www.cypress.com
Figure 12. PSoC Creator Top Design Schematic for Reference Resistance in Series with RTD
Firmware Flow
The following chart shows the firmware flow for the low- and mid-end RTD temperature measurement projects associated with
this AN. The projects differ only in the ADC resolution and the filter attenuation factor used.
Read voltage across RTD sensor

at zero current for CDS
Start
Read voltage across reference

resistor at zero current for
Correlated Double Sampling (CDS)
Start IDAC and set to source 1-mA

current
Connect IDAC current mux to

Reference resistor and measure
voltage V1.
CDS V1 and filter (IIR Filter)
Connect IDAC current mux channel

to RTD and measure voltage V2.
CDS V2 and filter
Measure RTD resistance

from V2 and V1
RRTD = (V2/V1)*Rref
Calculate temperature using the

RTD_GetTemperature(uint32 res)
API of the RTD component
END
www.cypress.com
If the reference resistance is in series with the RTD, the

firmware flow remains as shown, except that the IDAC
output is not multiplexed.
Testing the Project

1.
Plug CY8CKIT-025 into PORT

CY8CKIT-030 or CY8CKIT-050.
of
the
2.
For
the
CY8CKIT-030,
choose
the
CY8C3866AXI-040. For the CY8CKIT-050,
choose the CY8C5868AXI-LP035
3.
Build the attached project and program the PSoC

device.
4.
The LCD displays the RTD temperature.
Interfacing Multiple RTDs

The easiest way to interface to multiple RTDs is to wire
them in series with one another. However, make sure you
dont violate the compliance voltage of the IDAC, as
discussed earlier.
Alternatively, you can easily interface as many as four
RTDs with PSoC 3 / PSoC 5LP using the four IDACs in
the device.
Broken RTD Reconfiguration

If one of the RTD wires is broken, you can reconfigure the
four-wire RTD to a three-wire RTD and continue to show
the temperature without a considerable loss of accuracy.
Appendix B describes the broken RTD reconfiguration in
detail. A project (Broken RTD reconfiguration) is also
associated with this application note.
Temperature Resolution
The temperature resolution depends on three factors:
The temperature range to be measured

The IDAC current
The ADC resolution
High End
Lets calculate the resolution required for measuring
-200 C to 850 C with a 0.01 C resolution. First, we need
to translate the temperature resolution to voltage
resolution. From Equations 1 and 2, you can calculate the
resistance change per unit change in temperature .
A change from 850C to 851C causes a 0.292- change
in resistance.
A change from -200C to -201C causes
change in resistance.
a 0.432-
Take the smaller of the two: 0.292 for the calculations.

Use an IDAC current of 1 mA; the current used in the
associated project, for the calculations.
The voltage resolution required for 1 C = 0.292 * 1 mA =
292 V.
If you can resolve 2.92 V, you will achieve a 0.01 C
resolution.
This means the IDAC (operating at 1-mA current) and the
ADC together should have a noise of less than 2.92 V.
Next, calculate the full-scale voltage range of the RTD:
Resistance at -200 C = 18.52
(using Equation 2)
Resistance at 850 C = 390.481 (using Equation 1)

Voltage output at -200 C = 18.52 * .001 = 18.52 mV
Performance Measures
Voltage output at 850 C = 390.481 * .001 = 390.481 mV
The RTD-based temperature-sensing market can be

categorized into three performance segments: high end,
mid end, and low end (see Table 3).
Table 3. RTD Performance Ranges
Market segment
Resolution* (C)
Accuracy* (C)
High
0.01
0.1%
Mid
0.1
0.2 0.5%
Low
>0.1C
>0.5C
Resolution is generally specified only for temperatures > 0C. Accuracy

doesnt include the sensor accuracy and is usually specified with a fixed
offset, such as 0.1% or 1C, whichever is greater.
The ADC has an input range of 512 mV; this range fits
well with the output voltages of the RTD.
Using a 20-bit ADC with an input range of 512 mV, each
bit represents 1.024 V/2^20 = 0.9765 V.
According to the DelSig ADC datasheet, the ADC has an
RMS noise of ~1 count in the 512-mV range. This
indicates the noise is below the required 2.92 V.
However, when the voltage is close to a temperature
transition, the noise causes flicker in the output.
To eliminate flickering bits, the attached project adds a
small software-based IIR filter to the code. See AN2099,
Single-Pole IIR filter for a detailed description of a
software-based IIR filter.
In this section, we will see how to use PSoC 3 and

PSoC 5LP to address all three segments. Unit test results
appear at the end.
www.cypress.com
10
Mid End
For mid-end applications, the resolution requirement is
0.1 C. Using the same logic we used before, we need to
resolve only 29.2 V to achieve a resolution of 0.1 C.
For the mid-end project, the same ADC configuration is
used as the one in the high-end project. The only
difference is that the filter attenuation is lower in the midend project.
temperature (PVT). PSoC 3 ADC has an INL of 32 LSb;

32 LSb corresponds to 64 V for 20-bit resolution and
1.024-V range. The ADC INL has the same analog value,
64uV, in 0.512 V range. Lets calculate the error due to
INL at 850 C for IDAC current of 1 mA.
Low End
Vref I * Rref 1*100 100mV
For low-end applications, a resolution of 1 C is often

sufficient. Based on our earlier calculations, we need to
resolve only 292 V.
Using a 12-bit ADC in the 512-mV range gives a voltage
resolution of 250 V. The 250 V is below the requirement
of 292 V. Using a 12-bit ADC allows us to move into
lower-cost PSoC devices.
The project RTD_LowEnd associated with this application
note demonstrates how to use a 12-bit ADC to measure
an RTD.
Temperature Accuracy
You can calculate the temperature accuracy by summing
all possible individual errors, which fit into one of two
categories:
1.
Error due to the measurement system
2.
Error due to RTD
Error Due to the Measurement System

The error due to the measurement system is due to the
circuit shown in Figure 5. Consider Equation 8, which is
used to obtain RTD resistance.
RRTD
VRTD
* Rref
Vref
Using equation 8
Resistance error
390.481
390.481 0.064
*100
*100
100
100
0.064
Substituting the worst-case INL at the numerator, we get
0.064- resistance error corresponds to a temperature
error of 0.16 C(.064/.385). The INL is at its worst when it
applies only to the numerator.
Note that we have taken the worst-case INL across PVT
and substituted worst-case positive INL at the numerator.
In practical application, the error due to INL is much lower.
For example, for an INL of +8 LSb at the numerator and
0 LSb at the denominator, the temperature error at 850 C
is 0.054 C.
Error Due to Reference Resistance
Tolerance
In Equation 8, we substituted 100 for the value of the
reference resistor, Rref. But the actual value of Rref will
change because of its tolerance and temperature
coefficient. Therefore, the value Vref, which is measured
across the reference resistance, will be erroneous.
Assume that the tolerance of Rref is 0.1 percent and the
temperature coefficient is 10 ppm/C.
VRTD I * RRTD
VRTD is the voltage measured across the RTD
Vref I * Rref
Vref is the voltage across the 100- resistance

Rref is the reference resistance
As discussed, the only major source of error using this
method is the accuracy of the reference resistance. The
offset error is nulled by the difference, and the gain error is
nulled by the ratio. The other source of measurement error
is non-linearity in the ADC.
E r r o r D u e t o AD C I n t e g r a l N o n - L i n e a r i t y
The integral non-linearity (INL) of an ADC at any point is
the difference between the ideal ADC count and the actual
ADC count at that point after gain and offset corrections
have been completed. The datasheet specifies the
maximum INL of all points across process, voltage, and
www.cypress.com
VRTD I * RRTD 1* 390.481 390.481mV (The

RTD resistance at 850 C = 390.481, using Equation 1)
Substituting the value of Rref with the tolerance and

temperature coefficient yields Equation 11.
Vref I *100(1 0.001 0.00001* (sysTemp 25)

Equation 11
The measured value of the RTD resistance, Rmeas, is:
Rmeas
RRTD
(1 0.001 0.00001* (sysTemp 25))
Equation 12
where the 0.1 percent resistance tolerance contributes to

the additive factor 0.001, and the temperature coefficient
11
(10 ppm/C) contributes to the additive factor

0.00001*(sysTemp-25). Note that the temperature
coefficient is usually specified with reference to 25 C.
Therefore, the effect of the temperature coefficient is zero
at 25 C, but it increases as the temperature deviates from
25 C.
Figure 14. Temperature Error Due to Reference Resistor

Temperature Tolerance
The measured temperature can be calculated from the

measured resistance using Equations 1 and 2.
Figure 13 shows the temperature error due to the
reference resistor error at different RTD temperatures and
ambient temperatures. Ambient temperature (see legend)
is the temperature of the reference resistance on the
printed circuit board (PCB) while RTD temperature (x axis)
is the actual temperature to which the RTD is exposed.
Figure 13. Temperature Error Due to Reference Resistor
Accuracy and Temperature Tolerance
The temperature coefficient of the reference resistor would
still cause an error if the ambient temperature is much
higher or lower than 25 C. The ambient temperature can
be measured using the on-chip temperature sensor in
PSoC 3 / PSoC 5LP, and the error due to the temperature
coefficient of the reference resistor can be corrected.
For this method, the reference resistor needs to be
measured at manufacturing at several temperatures. This
approach will not be explained any further.
RTD Error
This is divided into two:
The orange line shows that for a 0.1 percent reference

resistor, the error is <1.3 C for an ambient temperature of
25 C and RTD temperatures from -200 C to 850 C.
The error contributed by the reference resistor can be
calibrated out using two methods. For the first method, the
value of the reference resistor should be measured at
25 C in the factory using an accurate ohmmeter and then
stored in the PSoC 3 / PSoC 5LP EEPROM. Use this
value (not 100 ) for Rref in Equation 8.
For the second method, replace the RTD with a known
precision resistor. Measure the value of the resistor.
Compute the ratio of this measured resistance against the
actual resistance. Store this ratio in flash. Multiply all
subsequent resistances by this ratio. This method will be
described more later.
By calibrating out the initial tolerance of the reference
resistor, the remaining error is only due to the temperature
coefficient. That error looks like the following:
1.
RTD interchangeability error
2.
RTD self-heating error
RTD Interchangeability Error

This error is caused by replacing one RTD with another
one of the same part number (process variation). This is
defined in RTD datasheet. IEC 60751 defines two primary
classes of RTD tolerances, as shown in Table 4.
Table 4. RTD Tolerance Classes
Tolerance
Class
Tolerance (C)
0.15 + 0.002 |t|
0.30 + 0.005 |t|
At no point would the temperature error of a class B RTD

exceed the value 0.3 + 0.005 |t|
Where |t| is the absolute value of temperature in C
The RTD on CY8CKIT-025 PSoC Precision Analog
Temperature Sensor EBK, PTS080501B100RP100, is a
class B RTD.
Figure 15 shows the graph of class B tolerance RTD. The
errors at 25 C and 800 C due to class B RTD are
calculated below.
www.cypress.com
12
Figure
15.
Temperature
Interchangeability Error
Error
due
to
RTD
Next, the project asks for the calibration temperature. The

CapSense slider on the CY8CKIT-030 or CY8CKIT-050
can be used to adjust the temperature by 1 C. The
CapSense buttons can be used to adjust the temperature
by 0.01 C. After the temperature has been entered, press
SW2.
The firmware will calculate the resistance for the entered
temperature. Next, it will measure the resistance of the
RTD. Finally it will compute the ratio and store it in
EEPROM.
If an accurate and stable temperature cannot be achieved,
replace the RTD with a known resistor. Using a known
resistor will calibrate out only the error due to the
reference resistor. It will not calibrate out the RTD
interchangeability error.
At 25 C, the worst-case temperature error = 0.3 + 25 *

0.005 = 0.43 C
At 800 C, the worst-case temperature error = 0.3 + 800 *
0.005 = 4.3 C.
If this error is not acceptable, the RTD has to be
calibrated. For most users seeking a high-performance
RTD (0.1% or 1C), calibration will be required.
Ideally, the resistance at 0 C, R0, is expected to be 100
for a PT100 RTD. But, practically, it will have a tolerance
that causes the interchangeability error. Going back to
Equations 1 and 2, note that a deviation in the value of R0
causes a scale error in the measured temperature.
Therefore, performing a scale correction corrects the RTD
interchangeability error.
For RTD calibration, perform the following steps.
Note that the reference resistance tolerance also causes a

fixed scale error in the measured resistance. Hence, when
the above six steps are completed, the reference
resistance also is calibrated. The scale error value can be
stored in PSoC EEPROM and retrieved each time the
device goes through a power cycle.
The 100 comes because you measured at 0C and 100C,
for a difference of 100. Now multiply all temperatures by
this scale factor.
This process is not implemented in the project.
Figure 15 and Table 4 show that the temperature error has
an offset and a gain. Thus, it may become necessary to
perform a two-point temperatue calibration. For example,
place the RTD at 0 C and measure the temperature
(Toffset). Subtract this measured temperature from
subsequent temperature readings.
1.
Make sure that all other sources of error are nulled

(Offset should be nulled by correlated double
sampling, and the gain error is automatically nulled by
the ratio).
Second, place the RTD at 100 C and measure the

temperature (TGain). Next, compute a scale factor using
the equation.
2.
Adjust the RTD to a known temperature, T1, for this

method T1 should be close to 0 C. Validate and
measure that temperature, T1, with an accurate
thermometer.
The 100 comes because you measured at 0 C and

100 C, for a difference of 100. Now multiply all
temperatures by this scale factor.
Scale = TGain Toffset / 100.
3.
Measure the RTD resistance, Rmeas, at T1.
This process is not implemented in the project.
4.
Calculate the actual RTD resistance at T1, Ractual,

using Equations 1 and 2.
5.
The scale error is given by
RTD Self-Heating Error

An RTD self-heating error is the increase in temperature of
the RTD due to the current flowing through the RTD.
R
scale actual
Rmeas
6.
As a result of self-heating, the RTD can show a

temperature slightly higher than ambient. This error can be
found in the RTD datasheet. For PTS080501B100RP100,
the value is specified at <= 0.8 C/mW.
Equation 13
Multiply this scale error by the measured RTD

resistance.
This process is done in the project RTD_HighEnd. When

the project starts, it asks the user if he wants to calibrate.
If he does, he must press SW2.
www.cypress.com
At 150 C, the RTD resistance is 157.325 . If 1-mA

current is passed, the power dissipation is 390 W or
0.39 mW. This corresponds to RTD self-heating of
<0.13 C. Note that the power dissipation constant of SMD
RTDs are generally higher (about 1C/mW).
13
List of all Errors

Table 5 shows the temperature error due to various components at 150 C. As seen from the table, RTD interchangeability
and reference resistance tolerance are the biggest sources of error. By comparison, the other errors are negligible.
Table 5. Possible Errors in RTD Temperature Measurement at 150 C

Error Source
Error Value at 150 C (0.1%

Reference Resistor, class B
RTD)
Error Value at 150C (Both Reference Resistor

and RTD Calibrated)
Signal Chain Error

Offset Error/drift
0 C
0 C
Gain Error/drift
0 C
0 C
ADC INL*
0.2 C
0.2 C
RTD self-heating error

(PTS080501B100RP100 SMD RTD)
< 0.13 C
<0.13 C
Error due to reference resistor (Ambient

Temperature = 25 C)
0.43 C
Limited only by calibration accuracy and reference

resistor temperature coefficient(very accurate)
Error due to reference resistor (Ambient

Temperature = 80 C)
0.6 C
~0.2 C unless temperature calibration is

completed
Sensor Error
Error due to RTD interchangeability (Class B

RTD)
1.05 C
Limited only by calibration accuracy (very

accurate)
Polynomial fit error (fifth-order polynomial)
0.0003 C
0.0003 C
Note(*): worst-case INL is used at the numerator. Typically, the error will be < 0.1 C.
Test Results
An RTD was simulated by a potentiometer and the signal chain accuracy was tested in the whole temperature range of the
RTD. A potentiometer was connected in the external RTD slot of CY8CKIT-025 and the temperature shown in the LCD was
noted. The potentiometer resistance was measured by a precision multimeter and the resistance value was noted. The
resistance was converted into temperature manually using a fifth-order polynomial. The test results are shown in Table 6. As
you can see, the RTD signal chain is highly accurate.
Table 6. RTD Signal Chain Accuracy
Resistance Value ()
Expected Temperature (C)
Measured Temperature (C)
Error (C)
27.285
-179.6
-179.6
0.0
32.47
-167.3
-167.4
0.1
66.58
-84.3
-84.4
0.1
80.21
-50.2
-50.3
0.1
118.296
47.1
47
0.1
149.464
129.0
128.9
0.1
218.374
317.9
317.8
0.1
325.314
636.3
636.3
0.0
www.cypress.com
14
Summary
About the Author
When high accuracy is critical in measuring temperatures,

an RTD is the sensor of choice. PSoC 3 / PSoC 5LP have
the necessary hardware integrated in the device to
achieve high accuracy. The PSoC Creator RTD
Component makes designing with RTD easier.
www.cypress.com
Name:
Praveen Sekar
Title:
Applications Engineer
Background:
Praveen holds a Bachelors degree in

Electronics and Communication from
the College of Engineering, Guindy,
Chennai. He focuses on analog
modules in PSoC
Contact:
pfz@cypress.com
Name:
Todd Dust
Title:
Applications Engineer Staff
Background:
Todd holds a Bachelors degree in

Electrical Engineering from Seattle
Pacific University.
Contact:
tdu@cypress.com
15
Appendix A
This section describes the effect of having the RTD reference resistance in a separate path (not in series with the RTD).
The IDAC current droops by about 1 percent as the IDAC load voltage increases from 0 V to Vdda-1 V. Each of the four IDACs
in PSoC 3 / PSoC 5LP can connect to one pin directly and to other pins through the analog globals. The direct connection is
through a single switch, which has a resistance of about 200 . When the IDAC connects to a pin using an analog global, it
connects through two switches, each with a resistance of about 200 . For a detailed description of the PSoC 3 or PSoC 5LP
analog routing, see AN58827 Internal Routing Considerations for PSoC 3 and PSoC 5LP Analog Designs.
For IDAC3, the direct connection is to pin 3_1. The RTD on the EBK is connected to pin 3_1 of PSoC, and the calibration
channel is connected to pin 3_4 of PSoC (see Figure 16).
The Top Design schematic, shown in Figure 11 requires separate IDAC multiplexer channels: the RTD channel and the
calibration channel. In building the project, PSoC Creator can route the RTD channel through the direct switch and the
calibration channel through the analog global. In this case,
IDAC load resistance in RTD channel = 200 + RTD resistance
IDAC load resistance in calibration channel = 500 (400 + calibration resistance 100 ).
The different load resistances result in different IDAC load voltages, which produce different IDAC currents in both channels.
The error in RTD resistance caused by the different currents is calculated by Equation 16.
I
R RRTD * 1 RTD
I
Re f
Equation 9
The resistance error increases as the RTD resistance (temperature) rises. If the ratio, IRTD/IRef, is equal to 1, the resistance
error falls to zero, as expected.
Lets calculate the error in temperature at 25 C due to the different IDAC currents.
The IDAC current used = 1 mA
The IDAC load voltage across the RTD channel = Load resistance * IDAC current = (200 + 109.732) * 1 mA = 309.732 mV
The IDAC load voltage across the calibration channel = 500 * 1 mA = 500 mV
As the load voltage rises from 309.732 mV to 500 mV, the IDAC current drops by about 0.5 A (See IDAC user module
datasheet graph)
From Equation 16,
1000
R 109.734 * 1
0.055
999.5
This corresponds to a temperature error of 0.14 C at 25 C.
www.cypress.com
16
Figure 16. IDAC to Pin Routing Resistances

AGR 4 AGR 5
PSoC
200
Temp Sensor EBK

Pin 3_4
200
Pin 3_1
Calibration
Resistance
200
200
IDAC3
200
RTD
To avoid that error, use manual analog routing to force the router to use the analog global path for both channels. Now the
load resistance in both channels remains about the same. The only factor that would affect the load resistance is the change in
RTD resistance due to temperature. At 850C, the RTD resistance would be 400 , while the calibration resistance would be
100 . That 300- difference causes a 300-mV difference in the load voltages. For that voltage difference of 300 mV, the
current change would be about 0.5 A. (The current droop with load voltage is available in the IDAC user module datasheet)
As a result, the worst-case error would be 0.7 C at 850 C.
Also, you should force the IDAC to IDAC3 by using a force placement directive in the directives tab of the design-wide
resources (.cydwr) file, as shown in Figure 17.
Figure 17. Force Placement Directive on IDAC
www.cypress.com
17
Appendix B
Broken RTD reconfiguration
If one of the four wires of the RTD breaks, PSoC can automatically detect a broken wire and reconfigure four-wire RTD
connection to three-wire connection and display temperature with minimal degradation to accuracy (< +/- 0.5C). Broken RTD
reconfiguration involves three steps.
1.
Detect broken wire connection
2.
Reconfigure the analog routing to change four-wire connection to three-wire connection
3.
Compensate for the additional wire resistance due to the three wire mode and display temperature.
The PSoC Creator project (Broken RTD Reconfiguration) associated with this application note demonstrates this feature.
Detecting Broken RTD wire
A four-wire RTD connection to PSoC is shown in Figure 18. The current is passed through pin 3_1 into RTD wire 1 and it is
grounded through RTD wire 4 (RTD wire 4 is not connected to a PSoC pin). The ADC differential inputs are connected to pins
4_0 and 4_1, which are connected to wires 2 and 3 of the RTD. The pin choices are made according to the connections in
CY8CKIT-025 PSoC precision analog temperature sensor EBK
Figure 18. Four-Wire RTD Connected to PSoC
IDAC
Pin 3_1
Rw1
R1
Pin 4_0
Rw2
RTD
R2
Pin 4_1
ADC +-
Rw3
R3
PSoC
Kit025
External 4-wire
RTD
Rw4
A broken wire can be detected by using PSoCs GPIO structure. PSoC GPIO can be configured to source vdd through a pull
up resistor while simultaneously sensing the pin state through its digital input buffer. For example, pin 3_1 can be configured
as shown in Figure 19.
Figure 19. Pin Configured in Resistive Pull-Up Mode and Digital Input Mode
PSoC
VDDIO
> 3.5k
Pin
State
Digital Input
Buffer
To analog
global
To analog
Mux Bus
To detect the broken RTD wire connected to pin 3_1, pin 3_1 is configured in the resistive pull-up mode and the pin state is
sensed back.
www.cypress.com
18
When wires 1 and 4 are not broken, the RTD resistance forms a resistor divider with the internal pull-up resistor and the
voltage across the RTD is sensed back as the pin state.
The pull-up resistance has a minimum value of 3.5 k and the RTD can have a maximum resistance of 390 (at 850 C).
Assuming wire resistances (Rw1 and Rw4) = 5 each, we get a maximum value of 400 .
Figure 20. Detecting Wires 1 and 4 for Breakage
VDDIO
PSoC
> 3.5k
Pin 3_1
Pin
State
Rw1
Digital Input
Buffer
To analog
global
Rw2
RTD
To analog
Mux Bus
Rw3
External 4-wire
RTD
Rw4
Voltage sensed by pin = Voltage across RTD+Rw1+Rw4

= VDDIO *
RRTD R w1 R w 2
RRTD R w1 R w 2 R p
= VDDIO *
400
3900
< 0.1 * VDDIO

VIL of the pin
< 0.3 * VDDIO
Therefore, when pin 3_1 is configured as resistive pull up (with a high voltage forced through the pin) and when no RTD wire is
broken, the pin state will be low. When either RTD wire 1 or 4 is broken, the pin state will be high.
Similarly, we can detect if RTD wires 2 and 3 are broken by configuring the respective pins to resistive pull-up modes and
reading the pin state back.
To find which RTD wire is broken, follow these steps:
1.
Disconnect pins 3_1, 4_0 and 4_1 from ADC and DAC
2.
Configure pin 3_1 in resistive pull up mode
3.
Drive high through pin 3_1
4.
Read the pin state of pin 3_1
5.
Repeat steps 2, 3, and 4 for pin 4_0 and pin 4_1
Let the pin states of pin 3_1, pin 4_0 and pin 4_1 be stored in variables A, B, and C respectively. Based on different values of
A, B, and C, we can have eight states, as shown in the following table.
www.cypress.com
19
Result
No Wire Broken
Wire 3 broken
Wire 2 broken
Wires 2 and 3 broken
Wire 1 broken
Wire 4 broken
The previous table also shows the result of each combination of A, B, and C.
If any of A, B and C is equal to 0, then wire 4 is not broken.
If A, B and C are all equal to 1, wire 4 is definitely broken. Apart from wire 4 any other wire can also be broken. But in such a
case reconfiguration is not possible. Reconfiguration is possible only if one of the wires is broken.
R e c o n f i g u r i n g f o u r -w i r e R T D t o t h r e e - w i r e
After detecting the broken RTD wire, we have to reconfigure the four-wire RTD to three-wire RTD eliminating the broken wire.
The flexible analog routing structure of PSoC makes it very easy to reconfigure four-wire RTD to three-wire RTD eliminating
the broken wire. The reconfiguration routes are shown below.
RTD Wire 1 Broken
If RTD wire 1 is broken, the current path is opened. To close the current path, the routing is reconfigured such that the current
is forced through the ADC pin 4_0 as shown in Figure 21.
Figure 21. RTD Wire 1 Broken
IDAC
Pin 3_1
Rw1
R1
Pin 4_0
Rw2
RTD
R2
Pin 4_1
ADC +-
Rw3
R3
PSoC
Kit025
External 4-wire
RTD
Rw4
R1 = Routing resistance from IDAC to pin

R2 = Routing resistance from ADC (positive) to pin
R3 = Routing resistance from ADC (negative) to pin
In this case, the IDAC current flows through R1, Rw2, through RTD to ground. Since Rw2 is in the measurement path of the
ADC, the RTD resistance and wire 2 resistance is also measured. This wire resistance can be eliminated through calibration
explained in one time wire resistance computation section.
www.cypress.com
20
RTD Wire 2 Broken

If RTD wire 2 is broken, the path from RTD to ADC positive terminal is opened. To close the path, we connect ADC positive
terminal to pin 3_1 as shown in Figure 22 .
IDAC
Pin 3_1
Rw1
R1
Pin 4_0
Rw2
RTD
R2
Pin 4_1
ADC +-
Rw3
R3
PSoC
Kit025
External 4-wire
RTD
Rw4
In this case, the ADC measures RTD wire resistance 1 in addition to the RTD resistance.
RTD Wire 3 Broken
If RTD wire 3 is broken, the path from RTD to ADC negative terminal is opened. To close the path, we connect ADC negative
terminal to ground as shown in Figure 23.
IDAC
Pin 3_1
Rw1
R1
Pin 4_0
Rw2
RTD
R2
Pin 4_1
ADC +-
Rw3
R3
PSoC
Kit025
External 4-wire
RTD
Rw4
In this case, the ADC measures RTD wire resistance 4 in addition to the RTD resistance. Also, any difference in potential
between the two grounds (kit-025 ground and the chip internal ground) adds to measurement error. One time offset correction
eliminates both the wire resistance error and the ground difference error.
RTD Wire 4 broken
If RTD wire 4 is broken, the current path from RTD to ground is opened. The ADC input terminal is Hi-Z and no current flows
through the ADC input. To close the path, we provide the ground path by configuring the pin in open drain low mode as shown
in Figure 24.
IDAC
Pin 3_1
Rw1
R1
Pin 4_0
Rw2
RTD
R2
Pin 4_1
ADC +-
Rw3
R3
R4
PSoC
Kit025
External 4-wire
RTD
Rw4
In this case, the ADC measures RTD wire resistance 3 in addition to the RTD resistance.
www.cypress.com
21
One-time Wire Resistance Computation

When reconfiguring a four-wire RTD into a three-wire RTD, wire resistances affect the RTD temperature measurement
accuracy. 1-ohm wire resistance can cause 3C error in measured temperature. To eliminate the error due to the wire
resistances, we perform a one-time wire resistance computation before making the RTD measurements.
The steps followed to compute the wire resistances are given below
1.
Configure the RTD in four-wire mode as shown in Figure 18 and find the RTD resistance (R0)
2.
Configure the RTD in three-wire mode (wire 1 broken) as shown in Figure 21 and calculate the resistance (R1)
3.
Compute additional wire resistance, CompRes1 = (R1 R0). When wire 1 breaks and the RTD is reconfigured as shown in
Figure 21, CompRes1 should be subtracted from the measured resistance.
4.
5.
6.
7.
Compute additional wire resistance, CompRes3 = (R3 R0). When wire 3 breaks and the RTD is reconfigured shown in
8.
9.
Project Description
The Project (Broken RTD reconfiguration) has been built to work with CY8CKIT-025. The project detects a broken RTD wire
and automatically reconfigures a four-wire RTD connection to a three-wire RTD connection and continues to display
temperature with a broken alert. The project also performs one-time wire resistance compensation so that the temperature
displayed is accurate.
The PSoC Creator schematic of the project is shown in Figure 25.
Figure 25. PSoC Creator Schematic of Broken RTD Reconfiguration Project
www.cypress.com
22
This project is similar to the RTD temperature measurement project except that there are more IDAC and ADC channels to
support four-wire to three-wire reconfiguration. Manual analog routing has been used to ensure that the ADC positive
connection always happens through analog mux bus. This way we ensure that the ADC and IDAC connections are through
two different paths until the pin. Each pin connects to the analog network through two paths: Analog mux bus and Analog
globals (see Figure 19 and Figure 20). See PSoC3 Datasheet I/O System and Routing section for a detailed description of pin
structure. By making sure that the ADC connects to the pin through the analog mux bus and that the DAC connects to the pin
through one of the analog globals, we eliminate any additional resistance that can add to RTD resistance.
Figure 21 shows separate IDAC and ADC paths until pin. Figure 26 shows IDAC and ADC paths merged before pin. In this we
have resistance R4 in addition to Rw2. This additional resistance will also get subtracted out when we perform resistance
compensation. But having separate current and voltage paths until the pin is a better way of doing a three wire RTD. If the
compensation fails due to some reason, the additional resistance error will be in the order of ohms when you have separate
IDAC and ADC paths whereas it will be in the order of hundreds of ohms when you do not have separate IDAC and ADC
paths until the pin.
Figure 26. Wire 1 Broken IDAC and ADC Paths Merged Before Pin
IDAC
Pin 3_1
Rw1
R1
Pin 4_0
Rw2
R4
RTD
R2
Pin 4_1
ADC +-
Rw3
R3
PSoC
Kit025
External 4-wire
RTD
Rw4
Firmware Flowchart
The firmware flowchart is shown in Figure 27. There is a step in which the firmware reconfigures the routing to four-wire mode
on a switch press. Whenever a wire is broken, the RTD reconfigures the routing and remains in that routing even if the broken
wire is fixed. The user can press the Switch, SW2, in the DVK and the firmware reconfigures the routing to four wire mode.
When one of the RTD wire breaks, the RTD resistance will drop down to less than 10 . This condition is used in the project to
trigger the checking of wires.
www.cypress.com
23
Figure 27. Firmware Flowchart

Start
Find additional wire resistances in
each of the four cases, wires 1, 2, 3,
and 4 broken
Set the analog routing to 4-wire

RTD configuration
Is SW2
pressed?
Calculate RTD resistance, R, and

subtract additional wire resistance
Is R<10?
N
Compute resistance
and display
temperature
Find broken
wire
Is Broken wire
= {1,2,3,4}
1
Reconfigure
analog routing for
wire 1 broken
Reconfigure
analog routing for
wire 2 broken
Reconfigure
analog routing for
wire 3 broken
Reconfigure
analog routing for
wire 4 broken
Testing the Project

1. Connect CY8CKIT-025 PSoC precision analog temperature sensor EBK to port E of CY8CKIT-030 PSoC3 DVK or
CY8CKIT-050.
2.
Connect an external four wire RTD to CY8CKIT-025
3.
Program the Broken RTD Reconfiguration project to the PSoC3 device
4.
Follow the one-time calibration instructions. It requires disconnecting RTD wire 4 and connecting it back
5.
The LCD will display the RTD temperature
6.
Disconnect wire 1. The RTD temperature will be about the same value and the LCD displays a message W1 Broken
7.
Reconnect wire 1 and press SW2. You will see the message W1 Broken disappear and the RTD is reconfigured back in
four wire mode
8.
Repeat steps 6 and 7 for RTD wires 2, 3, and 4.
www.cypress.com
24
Document History
Document Title: AN70698 PSoC 3 and PSoC 5LP Temperature Measurement with an RTD
Document Number: 001-70698
Revision
ECN
Orig. of
Change
Submission
Date
Description of Change
**
3458038
PFZ
12/12/2011
New Application note
*A
3490797
PFZ
1/12/2012
MEH Review Feedback in CDT#116240
*B
3520653
PFZ
2/8/2012
Updated project. No change to document
*C
3689958
PFZ
08/09/2012
RTD component includes support for PT100, PT500 and PT1000 RTDs
A new section on broken RTD reconfiguration has been added
Other minor changes
*D
3740378
PFZ
09/11/2012
Updated associated project files.
*E
3818484
PFZ
11/21/2012
Updated title to read as PSoC 3 and PSoC 5LP Temperature Measurement

with an RTD.
Updated Associated Part Family as All PSoC 3 and PSoC 5LP Parts.
Updated Related Application Notes as AN75511, AN66477, AN60590.
Updated Introduction.
Updated RTD Resistance-to-Temperature Conversion (Updated Choosing the
Right Polynomial Order (Updated description), updated RTD Component
(Updated Figure 6, Figure 7, Figure 8)
Updated Project Description (Updated Figure 11 and Figure 12)
Updated Appendix B (Updated Broken RTD reconfiguration (Updated Project
Description (Updated Figure 25))).
Replaced PSoC 5 with PSoC 5LP in all instances across the document.
*F
4057734
TDU
07/11/2013
Added two projects and discussed different performance ranges.
*G
4152296
TDU
10/09/2013
Updated attached Associated Project.

Completing Sunset Review.
*H
4202789
www.cypress.com
TDU
11/26/2013
Fixed formatting errors.
25
Worldwide Sales and Design Support

Cypress maintains a worldwide network of offices, solution centers, manufacturers representatives, and distributors. To find
the office closest to you, visit us at Cypress Locations.
PSoC Solutions
Products
Automotive
cypress.com/go/automotive
psoc.cypress.com/solutions
Clocks & Buffers
cypress.com/go/clocks
PSoC 1 | PSoC 3 | PSoC 5LP
Interface
cypress.com/go/interface
Lighting & Power Control
cypress.com/go/powerpsoc
cypress.com/go/plc
Memory
cypress.com/go/memory
Optical Navigation Sensors
cypress.com/go/ons
PSoC
cypress.com/go/psoc
Touch Sensing
cypress.com/go/touch
USB Controllers
cypress.com/go/usb
Wireless/RF
cypress.com/go/wireless
Cypress Developer Community

Community | Forums | Blogs | Video | Training
Technical Support
cypress.com/go/support
PSoC and CapSense are registered trademarks of Cypress Semiconductor Corp. All other trademarks or registered trademarks referenced herein are
the property of their respective owners.
Cypress Semiconductor
198 Champion Court
San Jose, CA 95134-1709
Phone
Fax
Website
: 408-943-2600
: 408-943-4730
: www.cypress.com
Cypress Semiconductor Corporation, 2011-2013. The information contained herein is subject to change without notice. Cypress Semiconductor
Corporation assumes no responsibility for the use of any circuitry other than circuitry embodied in a Cypress product. Nor does it convey or imply any
license under patent or other rights. Cypress products are not warranted nor intended to be used for medical, life support, life saving, critical control or
safety applications, unless pursuant to an express written agreement with Cypress. Furthermore, Cypress does not authorize its products for use as
critical components in life-support systems where a malfunction or failure may reasonably be expected to result in significant injury to the user. The
inclusion of Cypress products in life-support systems application implies that the manufacturer assumes all risk of such use and in doing so indemnifies
Cypress against all charges.
This Source Code (software and/or firmware) is owned by Cypress Semiconductor Corporation (Cypress) and is protected by and subject to worldwide
patent protection (United States and foreign), United States copyright laws and international treaty provisions. Cypress hereby grants to licensee a
personal, non-exclusive, non-transferable license to copy, use, modify, create derivative works of, and compile the Cypress Source Code and derivative
works for the sole purpose of creating custom software and or firmware in support of licensee product to be used only in conjunction with a Cypress
integrated circuit as specified in the applicable agreement. Any reproduction, modification, translation, compilation, or representation of this Source
Code except as specified above is prohibited without the express written permission of Cypress.
Disclaimer: CYPRESS MAKES NO WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, WITH REGARD TO THIS MATERIAL, INCLUDING, BUT
NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. Cypress reserves the
right to make changes without further notice to the materials described herein. Cypress does not assume any liability arising out of the application or
use of any product or circuit described herein. Cypress does not authorize its products for use as critical components in life-support systems where a
malfunction or failure may reasonably be expected to result in significant injury to the user. The inclusion of Cypress product in a life-support systems
application implies that the manufacturer assumes all risk of such use and in doing so indemnifies Cypress against all charges.
Use may be limited by and subject to the applicable Cypress software license agreement.
www.cypress.com
26

Psoc 3 and Psoc 5Lp - Temperature Measurement With An RTD

Uploaded by

Document Informationclick to expand document informationAN70698_001-70698_0H

Document Informationclick to expand document information

Copyright:

Available Formats

Psoc 3 and Psoc 5Lp - Temperature Measurement With An RTD

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Psoc 3 and Psoc 5Lp - Temperature Measurement With An RTD

Uploaded by

Copyright:

Available Formats

AN70698

PSoC 3 and PSoC 5LP Temperature Measurement with an RTD

Document No. 001-70698 Rev. *H

PSoC 3 and PSoC 5LP Temperature Measurement with an RTD

temperature detector (RTD), thermocouple, thermistor,

Table 1. Comparison of Temperature Sensors

Typical cost (US $)

Surface mount for

Surface mount for

three- and four-wire up to

Leaded for <1 meter

Amplifier, ADC, voltage

Excitation current, ADC,

Although they are more expensive than other sensor

RTD Theory of Operation

Where, RT is the resistance at T C

Document No. 001-70698 Rev. *H

PSoC 3 and PSoC 5LP Temperature Measurement with an RTD

Equation 1 and Equation 2 are also referred to as

Figure 2. Two-Wire RTD Temperature Measurement

In Figure 1 the Linear Approximation is made by assuming

Plotting the resistance versus the temperature yields a

Because of the length of the wire connected in series with

VRTD IS * (RRTD RW1 RW 2)

Measure the RTD resistance accurately

Convert the measured resistance accurately to

The following sections describe the two tasks.

RTD Resistance Measurement

With a three-wire measurement, the error due to the wire

Vwire = Is * (Rw2 + Rw3)

The techniques commonly used to measure resistance,

RRTD = (VRTD Vwire) / Is

Document No. 001-70698 Rev. *H

PSoC 3 and PSoC 5LP Temperature Measurement with an RTD

Figure 3. Three-Wire RTD Temperature Measurement

To overcome the gain/offset error caused by the ADC and

Reference Resistor Method

The reference resistor makes the measurement error

The schematic of this method is shown in Figure 5. A

Figure 5. RTD Temperature Measurement

Figure 4. Four-Wire RTD Temperature Measurement

Rref, 100 ohm

In this method, a known constant current is passed

To find the RTD resistance, the current source and the

The current path is shown by the bold blue line, and

Document No. 001-70698 Rev. *H

PSoC 3 and PSoC 5LP Temperature Measurement with an RTD

Reference Resistor Selection

Offset Error Cancellation

RTD VRTD V0 * Rref

Where, VRTD is the voltage measured across the RTD

Subtracting V0 From VRTD removes ADC offset. The

Gain Error Cancellation

Now the error depends primarily on the accuracy of the

The other solution of the quadratic equation is eliminated

Document No. 001-70698 Rev. *H

PSoC 3 and PSoC 5LP Temperature Measurement with an RTD

Vref I 100(1 0.001 0.00001 (sysTemp 25)