Acta Marisiensis.

Seria Technologica
Vol. 16 (XXXIII) no. 1, 2019
ISSN 2668-4217, ISSN-L 2668-4217

10.2478/amset-2019-0001

EXPERIMENTS ON TEMPERATURE CONTROL USING

ON-OFF ALGORITHM COMBINED WITH PID ALGORITHM
Mircea DULĂU
University of Medicine, Pharmacy, Sciences and Technology of Târgu Mureș
Gheorghe Marinescu Street, no. 38, 540139, Târgu Mureș, Romania
mircea.dulau@umfst.ro

Abstract
The paper presents an automated system for the temperature control using a PIC
microcontroller, a digital temperature sensor with I2C and a MOSFET as actuator. The
control strategy is a proportional-integrative one, preceded by an on-off algorithm. The most
important data is presented on a graphical interface. In order to send the parameters of the
process and the command states to the computer, the serial communication is used.

Key words: temperature control; PIC microcontroller, PID controllers, adaptive control

1. Introduction Experiments of PID auto-tuning controllers,

In the industrial applications, the precise control of implemented on microcontrollers and used for the
the processes’ parameters implies the necessity of temperature control are developed in [6]. A set of
adequate strategies for monitoring. techniques which provide a systematic approach for
In this paper, the authors propose a strategy of automatic adjustment of controllers in real time, when
monitoring and control of the temperature, using a the parameters of the plant dynamic model are
PIC16F826 microcontroller and the discrete controller unknown and/or change in time are developed in [7].
implemented on the PC. This study refers to the design, The architecture of the temperature control system
implementation and testing of the control schematic. using the PIC18F4620 microcontroller, the
With adequate modifications, the schematic can be programming software and the PID control are
extended for the control of other parameters, such as presented in [8]. The paper [9] describes the
flow, level, pressure etc. functionality of a temperature controlling system using
This field’s literature presents the PIC18F45K22 microcontroller.
microcontrollers’ using problem for the industrial An example of embedded system design for the
applications and highlights similar studies and temperature control, which uses the PIC16F876A,
experiments. with different block diagrams and conditioning, are
So, the fundamentals of the PIC microcontrollers, presented in [10]. In the paper [11] the microcontroller
with hardware and software details, are very well based water level sensing is investigated in order to
presented in [1]. The elements of designing and manage the power consumption and the water
implementation with microcontrollers and overflow. The paper [12] relate the appliance of
microprocessors, the set of instructions, the organizing adaptive model predictive controller that is designed to
and addressing of the memory, the interruption system handle temperature control of continuous stirred tank
etc. are developed in [2,3]. The PID and the On-Off reactor.
classical control algorithms with the tuning criteria are In [13], an adaptive PID control algorithm is used
presented in [5]. The study from [4] presents an in order to track the process temperature. The control
automated system used for the temperature control of algorithm employs Lyapunov function based artificial
the fluid. The control algorithm is an on-off type with neural networks for online tuning of PID actions.
hysteresis one and it is implemented with a PIC The paper [14] deals with a PIC microcontroller to
microcontroller and a relay as its actuator. monitor and record the value of temperature, humidity,
© 2019 Published by University Press. This is an open access article licensed under the Creative Commons
Attribution-NonCommercial-NoDerivs License (http://creativecommons.org/licenses/by-nc-nd/3.0/).
5
soil moisture and sunlight of the natural environment. off control and the final conclusions are provided in the
In order to solve the problem of high overshoot and final section.
slow calculation speed in the design of temperature 2. The schematic control
control system, the paper [15] propose an adaptive The block diagram designed for the temperature
control algorithm based on characteristic model. Also, acquisition and the command of the heating elements
for the design and implementation of the temperature (electric resistance) consist of (Fig. 1):
control schematic presented in the paper, the  acquisition and signal processing block;
recommendations from the [16, ..., 18] data sheets  computing, recording and display block.
were considered. The acquisition block is an interface between the
In this section of the paper a short review of studies PC and the process (furnace) and ensure:
on the temperature control, from the literature is  the processing of the signal receiving from the
presented. In Section 2 the block diagram and the temperature sensor and sending to the PC;
electronic schematic for the temperature control are  the modulation of the voltage for the heating
presented. In Section 3 the control algorithm with element.
implemented subroutines and the graphical interface This block consists of (Fig. 2):
are detailed. Experiments of PI controlled systems,  MCP9803, digital temperature sensor with I2C
tuned by the Ziegler – Nichols criteria, preceded by on- interface [17];

Fig. 1: The block diagram of the temperature control

Fig. 2: The electronic schematic

6
  PIC16F876, 8 bits microcontroller [16];
 MAX232 driver driver/receiver for RS232 serial
interface [18];
 MOSFET transistor as actuator for high power
and low voltage;
 Optocoupler that provide electrical isolation
between the supply voltage and the lower
voltage;
 Quartz oscillator, signaling LEDs, connectors.
The microcontroller communicates with the
temperature sensor through I2C bus and receives 12
bit, binary data (temperature), then waits data from
RS232.
This data is sent to the PC. If there are no data, after
a specified time, the supply voltage of the heating
element is stopped, and the receiving subroutine is Fig. 3: The dependence of the average power by the PWM
reactivated.
After the data receiving, a test decides if data are in
the (0 - 100)% admissible domain, which is exactly the
amount of power to be supplied to the heating element
(expressed as a percentage). The control of the power
is based on the Pulse Width Modulation (PWM), with
a rectangular fixed-frequency signal and a variable
duty cycle.

3. The software application

The dependence of the average power by the on-
time PWM is presented in Fig. 3, with the time (T) of
the command signal, generated by the microcontroller,
of 10 seconds.
So, for a 1% power, a pulse of 0.1 sec./T is
generated, this value being the on-time resolution of
the PWM. The power can be percentage controlled in
the (0 - 100)% range.
The program in the PIC16F876 microcontroller is
structured in subroutines, for (Fig. 4):
 USART setting (Universal Synchronus
Asynchronus Receiver Transmiter) for serial Fig. 4: Flow chart
transmission, with 9600 baud, with one stop bit
and no parity bit; The graphical interface highlights (Fig. 5): buttons
 setting of the temperature sensor: 12 bits, (1,2,16) for start/stop and program exit; serial port
meaning 0.0625 oC resolution; selection (3); data load/save (4,5); data save mode (6);
 data reception/transmission from/to PC, without index display (7); setting of the output power (8);
locking, with 5 seconds time out; selection of the controller type (9); tuning parameters
 I2C/SMbus type communication; (10,11); display of the instant/average output power
 MSSP setting (Master Synchronus Serial Port), (12,13); temperature display (14); temperature set
for SMbus, with 100 kHz communication; point (15); sizing of Oy and Ox axis (17,18,19,20);
 pulse width modulation by software; reset (21); errors messages (22); display of the Oy and
 sensor (temperature) reading, two bytes: MSB Ox axis (23,24); temperature (25).
for the integer and LSB for the fractional value.
The program for the digital controller is written 4. The experimental results and conclusions
based on LabWindows/CVI and has the following During the process identification there may be
facilities: significant temperature increases, that could damage
• manual control of the heating element; the process components (sensors, physical structures
• using of the P, PI and PI with anti-windup etc.), so beforehand an on-off control is made at a
controllers; temperature value of 90% out of the set point value.
• using of the adaptive controller; During this on – off control procedure, the average
• save – load and display of the measured data. power is calculated by averaging of the command
pulses in time.

7
 Fig. 5: The graphical interface

If the average power variation does not exceed 2%

in the last 2 times (T), the on-off control stops and the
control system switches to a PI control and a constant
output power, which is equal with the calculated
average power.
E.g., for the set point of 50oC, the control starts
with a set point value of 45oC and an on-off algorithm,
then switches to a PI algorithm, followed by the
increase of the set point temperature value with 10%,
meaning 50oC (Fig. 6, Fig. 7).
Similarly, Fig. 8 presents the PI control with the set
point value of 55oC.
Fig. 7: Proportional-Integrative (PI) control with set point
of 50oC

Fig. 6: On-off control with set point of 45oC

Fig. 8: PI control with set point of 55oC

The tuning of the controllers is the most important

8
task that needs to be solved in order to achieve the set- International Conference on Computer as a Tool
up performances imposed in the closed-loop system. (EUROCON), vols. I, II, pp. 266-269.
The tuning parameters of the PI controller are [7] Landau, I.D. (2011), Adaptive Control,
calculated using the Ziegler-Nichols criteria, available Communications and Control Engineering,
for the slow processes with one time constant and Springer-Verlag London Limited.
delay time [4]. In turn, these parameters are obtained [8] Levărdă, B. and Budaciu, C. (2010), The Design
by offline identification, using the Cohen-Coon of Temperature Control System Using
graphical method [4]. PIC18F46201, „Gheorghe Asachi” Technical
Starting with an on-off controller and a reduced set University of Iaşi, Politehnica Bulletin, Tome
point value (90%) the output temperature has a LVI (LX), Fasc. 4, pp. 203-214.
hysteresis of about 2oC, but the process components [9] Rob, R., Tirian, G.O and Panoiu, C. (2016),
are protected. By switching to the PI algorithm and to Temperature Controlling System using
the “total” set point (100%), the steady state error is Embedded Equipment, International Conference
zero. on Applied Sciences (ICAS), vol. 163.
From practical point of view, using a serial bus [10] Sharma, S.R., and Dahikar, P.B. (2013),
reduces the cabling costs and prototyping time, making Embedded Design of Temperature Controller
it the ideal solution based on microcontrollers. Using PIC16F876A for Industries and
The major disadvantage is given by the significant Laboratories, International Journal of Innovative
reduction of the communication speed, between the Research in Computer and Communication
components of the system. This disadvantage is Engineering, vol. I, Issue 10, pp. 2414-2422.
diminished in the case of the microcontrollers where [11] Mikrajuddin, A. and Aprianti, N.A. (2007),
the most of the components are integrated in the same Learning Control at Undergraduate Level Using
circuit. The PWM has a major advantage, because the PIC16F877 Microcontroller-based Temperature
actuator loses less power, compared to the analog Controller, ASME/IASME Conference on
system. Mathematical Methods and Computational
Techniques in Research and Education.
Acknowledgement [12] Ratnakumari, U.V. and Triven, M.B. (2016),
This paper is based upon work supported by the Implementation of Adaptive Model Predictive
Energy and Electrotechnologies Management Controller and Model predictive control for
Research Centre. temperature Regulation and concentration
tracking of CSTR, IEEE International Conference
References on Communication and Electronics Systems
(ICCES), pp. 698-703.
[1] Katzen, S. (2010), The Essential PIC18
[13] Aftab, B.S. and Shafiq, M. (2015), Adaptive PID
Microcontroller, Springer Verlag London.
Controller based on Lyapunov Function Neural
[2] Duka, A. V. (2013), Sisteme cu microprocesoare
Network for Time Delay Temperature Control,
(Systems with Microprocessors), „Petru Maior”
IEEE 8th GCC Conference and Exhibition
University of Tîrgu Mureş Publishing House.
(GCCCE).
[3] Genge, B. and Haller, P. (2008), Proiectarea
[14] Singh, A.J., Raviram, P. and Shanthosh, K.
sistemelor dedicate şi încorporate cu
(2014), Embedded based Green House
microcontrolerul PIC (Design of the Dedicated
Monitoring System using PIC Microcontroller,
and Embedded Systems with PIC
IEEE International Conference on Green
Microcontroller), „Petru Maior” University
Computing Communication and Electrical
Publishing House, Tîrgu Mureş.
Engineering (ICGCCEE).
[4] Dulău, M. and Gligor, A. (2015), Introducere în
[15] Minkai, L., Sang J. and Yiming, Y. (2017),
Ingineria sistemelor automate (Introduction to
Design of temperature control system of adaptive
the Control Systems Engineering), „Petru Maior”
control algorithm based on characteristic model,
University Press, Tîrgu Mureş.
IEEE 32nd Youth Academic Annual Conference
[5] Dulău, M., Karoly, M. and Dulău, T. M. (2018)
of Chinese Association of Automation (YAC).
Fluid temperature control using heat exchanger,
[16] PIC16F87X Data Sheet, Microchip Technology
11th International Conference on
Inc., 2013.
Interdisciplinarity in Engineering (Inter-Eng
[17] MCP9800/1/2/3 Data Sheet, 2-Wire High-
2017), Procedia Manufacturing, vol. 22, pp. 498-
Accuracy Temperature Sensor, Microchip
505.
Technology Inc., 2004.
[6] Bolat, E.D., Erkan, K. and Postalcioglu, S.
[18] MAX232 Data Sheet, Texas Instruments Inc.,
(2005), Experimental Autotuning PID Control of
2014.
Temperature Using Microcontroller, IEEE