Management Systems

in 2018, Volume 26, Issue 3, pp. 168-171

Production Engineering

Date of submission of the article to the Editor: 03/2018

Date of acceptance of the article by the Editor: 08/2018
DOI 10.1515/mspe-2018-0027

THE COMPARING OF THE SELECTED TEMPERATURE SENSORS COMPATIBLE

WITH THE ARDUINO PLATFORM

Emil ŠKULTÉTY, Elena PIVARČIOVÁ*, Ladislav KARRACH

Technical University in Zvolen

Abstract:
One of the most frequently measured quantity is temperature, which is also one of the most important physical
quantities. Temperature has influence on the almost all states and processes in the nature as well as in tech-
nique. A wide range of temperature sensors is currently available on the market. They use different measure-
ment principles and exist in many designs. According to the location of the sensing element in the measured
environment, they are divided into two main groups: contact and non-contact. Further, we can divide the
temperature sensors according to the physical principle on which they work. The article deals with the analysis
and comparison of selected Arduino-compatible contact temperature sensors. The temperature measurement
of machine functional nodes and its diagnostics are part of maintenance and engineering diagnostics. At pre-
sent, NC and CNC machine diagnostics are an important trend in machine condition monitoring and machine
status prediction to maintain production quality. Machine status monitoring allows reducing of machine ser-
vice costs as well as maintaining the high production quality.

Key words: temperature, sensor, measurement, Arduino

INTRODUCTION object and by early detection of physical changes and pro-

The temperature is one of the most important thermody- cesses [7]. At the same time, temperature is the most
namic properties which determine the state of matter and common malfunctioning factor in measuring physical and
is presented in many physical laws. There are very few chemical quantities, so it is often necessary to compen-
properties of substances that are not temperature de- sate its influence on the measurement of the requested
pendent [1]. quantity, which requires temperature measurement [4].
For a quantitative description of different thermal states, Sensors with different physical principles are used for
it is necessary to assign the numerical values to them and temperature measurement. According to the principle
so define a temperature scale. Two scales were standard- they use, we can divide them into resistive, thermoelec-
ized to measure the temperature: the Kelvin and the Cel- tric, semiconductor, monocrystalline, thermistors, dilata-
sius temperature scale [2]. tion, optical, chemical etc. Furthermore, they are divided
The current level of progress is influenced by advanced in- into contact and non-contact sensors, according to
formation technologies [3]. In automation technique, the whether or not they have contact with the measured en-
temperature belongs to the most monitored quantities in vironment. Despite the increasing importance of non-con-
production and also non-production processes and the tact measuring methods, industrial temperature is meas-
quality and production cost of many products and services ured by contact methods [8].
often depend on its accuracy [4]. For example, the accu-
rate measurement of the temperature in metallurgy is es- METHODOLOGY OF RESEARCH
sential mainly due to the high quality requirements for There are available many types of sensors and devices on
casting products. Quality can be achieved only by strict the market for temperature measurement. It is not easy
compliance to the technological regulations, where to select the most appropriate equipment from such a
knowledge of temperature and other physical and chemi- wide variety. [9]. The sensors, which we have chosen for
cal properties of the melt play a major role [5]. comparison, comply with following conditions: they can
High reliability of technical systems is a prerequisite for be connected to Arduino platform, for processing the in-
their efficient use and safety [6]. Reliability as one of the formation from the sensors it is not necessary to program
key features of the quality of each technical device is sub- complicated programs and they don’t need several other
ject to effective monitoring of the technical state of the components and ICs for correct functioning. For example,
the resistance temperature sensor Pt100, which is used in

Unauthentifiziert | Heruntergeladen 22.10.19 07:38 UTC

E. ŠKULTÉTY, E. PIVARČIOVÁ, L. KARRACH – The comparing of the selected temperature sensors… 169

various applications, needs Wheatstone bridge circuit for humidity sensor. The difference between these sensors is
evaluation. Temperature can be measured using deflec- in measurement range, accuracy and communication in-
tion or compensation method. terface. The basic parameters are shown in Table 3.
On the base of above mentioned conditions we have se-
lected the sensors LM35DZ, DS18B20 and AM2320, which Table 3
are often used in applications based on Arduino platform Parameters of the AM2320
e.g. meteo stations, temperature control in the flat or the Temperature measurement range –40-80°C
intelligent greenhouse. Accuracy ±0.5°C
The reference temperature was verified using a universal Resolution 0.1°C
measuring instrument, ALMEMO 2590-4AS. The first sen- Humidity measurement range 0-99.9%
sor in comparison is LM35DZ. It is semiconductor temper- Accuracy ±3%
ature sensor with analog output. The temperature meas- Resolution 0.1%
urement with this sensor can be more accurate than the Supply Voltage 3.1-5.5 V
Communication interface I2C
measurement with thermistors [10]. The LM35 device is
rated to operate over a −55°C to 150°C temperature
range. The important advantage against the other sensors The data from thermal sensors are processed with help of
is that, the analog output is directly proportional to tem- Arduino UNO R3. The Microcontroller ATmega 328P rep-
perature in °C [11]. The basic parameters are shown in Ta- resents the basic part of it and it works on 16 MHz fre-
ble 1. quency. The microcontroller board contain 6 analog pins
and 14 digital input/output pins of which 6 can be used as
Table 1 PWM outputs. It contains also power supply pins +5 V,
Parameters of the LM35DZ +3.3 V and 3 GND pins. Board is powered through USB or
Temperature-Voltage scale factor +10 mV/°C using AC-to-DC adapter 7–16 V.
Measurement range –55-150°C The connection of sensors can be seen in Fig. 1. Sensor
Supply voltage 4-30 V LM35DZ has 3 pins. Pin Vcc is connected to pin +5 V, GND
Current drain 60 μA to ground and Vout to the analog input A0. For sensor
Self-heating 0.08°C DS18B20 was selected “parasite” mode. Pins Vcc and GND
Accuracy ±3/4°C of sensor are connected to ground pin of Arduino. Middle
Package TO – 92 pin DQ is connected to digital pin D4 and resistor 4k7 is
connected on one end to pin DQ of sensor and on other
The second sensor in comparing is DS18B20. This digital end to pin +5 V of Arduino. Sensor AM2320 has 4 pins. Pin
thermometer provides digital output and communicates Vcc is connected to pin +5 V and pin GND is connected to
over a 1-Wire bus that by definition requires only one data ground of Arduino. Pins SDA and SCL are connected to an-
line (and ground) for communication with a central micro- alog pins A4 and A5 of Arduino UNO, which are used for
processor. The DS18B20 has a unique 64-bit serial code, I2C communication on this programmable controller
which allows multiple DS18B20s to function on the same board. In wiring there is also included push-button with
1-Wire bus [12]. The analog-to-digital converter (ADC), 10 kΩ resistor. Button is connected to digital pin D5.
coupled to the temperature sensing unit, converts tem-
perature to 9-12 bit word. The measurement range of sen-
sor is from –55°C to 125°C [13]. In addition, sensor can
operate in normal or ”parasite” power supplying mode. It
means, that the DS18B20 can derive power directly from
the data line (“parasite power”), eliminating the need for
an external power supply. The basic parameters are
shown in Table 2.
Table 2
Parameters of the DS18B20
Operating Temperature Range –55-125°C
Fig. 1 Wiring scheme of sensors LM35DZ, DS18B20 and AM2320
Supply Voltage 3-5.5 V
Accuracy ±0.5°C
Programmable Resolution 9-12 bits The program for data reading from temperature sensors
Communication interface 1 – Wire was created in Arduino IDE 1.8.2. The program consists of
Pin-Package TO – 92 2 functions: function setup() and loop(). Function setup()
is executed once at program startup. In this function
The last one in comparing is sensor AM2320. This sensor speed of communication and pin setup is performed. Af-
contains beside thermometer also hygrometer. It com- ter function setup() is completed, the function loop() is ex-
municates via I2C interface. Temperature measurement ecuted in infinite loop. This function waits until push-but-
range is from –40°C to 80°C. The further similar sensors, ton is pressed. When push-button is pressed, ten values
which we can find in use, are sensors AM2302, DHT11,
DHT12. Those contain except a temperature sensor also a

Unauthentifiziert | Heruntergeladen 22.10.19 07:38 UTC

 170 Management Systems in Production Engineering 2018, Volume 26, Issue 3

are read from temperature sensors. For efficient imple-

mentation of program there were used freely available li-
braries for sensors DS18B20 and AM2320.
In Fig. 2 there is block scheme of measuring workplace,
where we have tested accuracy of temperature sensors.
In experiment was used model of drying-plant at the de-
partment of institute. The model consists of a metal con-
struction that represents a model of a hot-air drying-plant
where temperature control is taking place. The spiral and
cooling fan are used as operational components. The con-
trol system is realized with help of programmable logic
controller (PLC) OMRON CP1L with touch panel [14]. Fig. 5 Course of the measured temperatures from the AM2320
Measurement was carried out by adjusting the tempera- sensor
ture from 25°C to 75°C stepwise, by step 5°C. At each step
10 measurements were made. The upper limit has been DISCUSSION
selected to prevent damage of the AM2320 sensor. The analyse of variance has shown, that sensors under-
rated the measured data. Sensor LM35DZ was the least
accurate. Using regression analysis for all sensors were as-
certained regression equations of linear regression mod-
els were ascertained relation between the real and meas-
ured temperature, which we obtained by sensors. We de-
termined the following parameters of regression equation
for sensor LM35DZ: a = 0.9725 and b = 1.177. In case of
sensor DS18B20 are parameters of regression equation: a
= 0.9985 and b = 0.1295. In case of sensor AM2320 the
Fig. 2 Block scheme of the measurement workplace parameters of regression equation were: a = 0.9956 and b
= 0.2327. The mathematical model of the sensors
RESULTS OF RESEARCH DS18B20 and AM2320 is closer to reality than in case of
The measured data were processed by Microsoft Excel sensor LM35DZ. The gradient of regression function of
2013. We analyzed the measured data and using the anal- sensors DS18B20 and AM2320 is closer to the value 1 and
ysis of variance we determined the basic statistical char- y-intercept constant is nearer to 0, than with sensor
acteristic. Then a regression analysis was performed. The LM35DZ.
results are shown in charts in Fig. 3-5. It is obvious from variance and regression analysis of the
measured data that sensors DS18B20 and AM2320 meas-
ure more accurate than sensor LM35DZ. For more precise
measurement the regular calibrating is necessary and also
using of further methods and approaches for increasing of
measurement accuracy.
Advantages of LM35DZ sensor are output voltage linearly-
proportional to the temperature and simple wiring. The
accuracy and speed of processing depends on the temper-
ature calculation formula in °C from the value we have got
from A/D converter. In case of DS18B20 sensor is the
measurement accuracy inversely proportional to the
time, which the sensor needs for A/D conversion. In com-
Fig. 3 Course of the measured temperatures from the LM35DZ parison to the LM35DZ sensor, the sensor communicates
sensor via the 1-Wire bus while supporting the “parasitic” mode.
The AM2320 sensor has a disadvantage of small measur-
ing range compared to the previous sensor. The ad-
vantage is that it contains, in addition to the thermometer
also a hygrometer and we can communicate with the I2C
bus. All examined sensors have their pros and cons. It is
up to the user, which of these sensors is the most suitable
for his application.

CONCLUSION
Temperature measuring and regulation is a dynamic field
of industrial automation. The sensors for temperature
measuring are produced in many sizes, shapes and preci-
Fig. 4 Course of the measured temperatures from the DS18B20 sions to fit requirements of specific tasks. Temperature
sensor

Unauthentifiziert | Heruntergeladen 22.10.19 07:38 UTC

E. ŠKULTÉTY, E. PIVARČIOVÁ, L. KARRACH – The comparing of the selected temperature sensors… 171

sensor must be periodically calibrated to provide accuracy 1 – Systémové pojetí automatizace. Brno: Computer
of measured data. The performance of individual parts of Press, 2012.
technological equipment could be optimised by calibrating [9] R. Strnad. “Měření teploty – porozumění vlastno-
with correct values of uncertainty and thereby make his op- stem měřicího přístroje”. Automa, vol. 15, no. 6, pp.
eration more effective [9]. 31-38, 2009.
The current trend in temperature sensors lies in wider use [10] S. Surya and S.S. Chauhan. “Water Level Indicator
of microelectronic technology e.g. in implementation of with Temperature Sensor”. IOSR Journal of Electrical
semiconductor temperature sensors on a single chip with and Electronics Engineering, vol. 10, no. 3, pp. 65-71,
analogue and digital circuits enabling the connection of 2015.
sensors to signal busses, in the development of new types [11] P.D. Patil and R.D. Patil. “Designing Multisensor Em-
of heat radiation detectors, in the integration of tempera- bedded System Using PSoC”. International Journal of
ture sensors into intelligent sensors for automatic correc- Current Advanced Research, vol. 8, no. 4, pp. 271-
tion of variations caused by temperature fluctuations [4]. 274, 2015.
Practical use of temperature measurement and tempera- [12] G. Gricius, D. Drungilas, A. Andziulis, D. Dzemydiene,
ture regulation is in manufacturing engineering in the field M. Voznak, M. Kurmis and S. Jakovlev. “Advanced
of maintenance and technical diagnostics. High demands Approach of Multiagent Based Buoy Communica-
on reliability and early fault diagnosis are placed on modern tion”. The Scientific World Journal, vol. 2015, 2015.
machining technology. Each CNC machine has a standard [13] H. Jing. “Design and Development of the Tempera-
operating diagnostics. The typical quantities, which are ture Detection System”. International Journal of
tracked, are engine winding temperatures and coolant Control and Automation, vol. 8, no. 2, pp. 409-416,
temperature. Measurement of temperature is important 2015.
for example in the diagnostics of electrical drives [15], con- [14] Ľ. Naščák and P. Koleda. “Regulácia teploty modelu
tact stress analysis [16], at determination of residual teplovzdušnej sušiarne programovateľným automa-
stresses [17], in vibration signals of machinery [18]. tom”. Acta facultatis technicae, vol. 1, pp. 127-133,
2014.
ACKNOWLEDGEMENTS [15] I.V. Abramov, A.I. Abramov, Z.R. Nikitin, E. Sosno-
This paper was prepared within the work on a research vich, P. Božek and V. Stollmann. “Diagnostics of Elec-
project KEGA 003TU Z-4/2016 “Research and education la- trical Drives”, in Proc. of International Conference on
boratory for robotics”, KEGA 001TU Z-4/2016 "Support Electrical Drives and Power Electronics, 2015, pp.
of Teaching for Heat and Mass Transfer in Technical Edu- 364-367.
cation" and VEGA 1/0086/18 “Researching Temperature [16] P. Frankovský, O. Ostertag, F. Trebuňa, E. Osterta-
Fields in a Set of Shaped Heat Transfer Surfaces”. gová and M. Kelemen. “Methodology of contact
stress analysis of gearwheel by means of experi-
REFERENCES mental photoelasticity”. Applied optics, vol. 55, no.
[1] M. Kreidl. Měření teploty – Senzory a měřící obvody. 18, pp. 4856-4864, 2016.
Praha: BEN – technická literatura, 2005. [17] K. Masláková, F. Trebuňa, P. Frankovský and
[2] P. Beneš, J. Chlebný, J. Král, J. Langer and M. Mar- M. Binda. ”Applications of the strain gauge – for de-
tinásková. Automatizace a automatizační technika 3 termination of residual stresses using ring-core
– Prostředky automatizační techniky. Brno: Compu- method”. Procedia Engineering, vol. 48, pp. 396-401,
ter Press, 2014. 2012.
[3] L. Chybowski, K. Gawdzińska and B. Wiśnicki. “Qual- [18] T. Stejskal, J. Kovac and S. Valencik. “Mechanism of
itative importance measures of systems components randomness in vibration signals of machinery”, in
– a new approach and its applications”. Manage- Proc. of International Conference on Industrial, Ser-
ment Systems in Production Engineering, vol. 24, no. vice and Humanoid Robotics, 2012, pp. 257-262.
4, pp. 237-246, 2016.
[4] J. Šturcel. Snímače a prevodníky. Bratislava: Vyda-
vateľstvo STU, 2002. Ing. Emil Škultéty
[5] P. Koleda, P. Koleda and S. Grúbel. “Analysis of tem- Assoc. Prof. Elena Pivarčiová, PhD.
peratures in the mould area during the process of Ing. Ladislav Karrach
engine cylinder heads casting”. Acta facultatis tech- Technical University in Zvolen
nicae, vol. 1, pp. 31-40, 2016. Faculty of Environmental and Manufacturing Technology
[6] S. Gierej. “Big data in the industry – overview of se- Department of Manufacturing and Automation Technology
lected issues”. Management Systems in Production Študentská 26, 960 53 Zvolen, Slovakia
Engineering, vol. 25, no. 4, pp. 251-254, 2017. e-mail: xskultetye@tuzvo.sk
[7] M. Kreidl and R. Šmíd. Technická diagnostika – Sen- pivarciova@tuzvo.sk (corresponding author)
zory, metody, analýza signálu. Praha: BEN – tech- karrach@zoznam.sk
nická literatura, 2006.
[8] P. Beneš, J. Janeček, J. Král, G. Künzel, B. Lacko, J. Se-
merád, P. Souček, L. Šmejkal, R. Voráček, L. Maixner
and B. Šulc. Automatizace a automatizační technika

Unauthentifiziert | Heruntergeladen 22.10.19 07:38 UTC