Research Journal of Applied Sciences, Engineering and Technology 13(10): 771-780, 2016

DOI:10.19026/rjaset.13.3351
ISSN: 2040-7459; e-ISSN: 2040-7467
© 2016 Maxwell Scientific Publication Corp.
Submitted: May 29, 2016 Accepted: September 14, 2016 Published: November 15, 2016

Research Article
Arduino and Labview Based Control for Efficient Drive of Cooling Fan System

Tareq S. El-Hasan, Mohammad Alia, Wasif Saluos and Ahmad Al-Janaideh

Electrical Engineering Department, Zarqa University, Zarqa 132222, Jordan

Abstract: This study is concerned with the development of PID/PWM control algorithm for use with Arduino Uno
and Labviewe for efficient control of a cooling fan drive system. It relies on the development and testing of a
regulatory temperature control system, where forced ventilation is required via cooling fans driven by DC motors. A
prototype that emulates the case of PC was developed in which a heater element was introduced in the case. An
electric fan is placed in the vicinity of the heated element so that cooled air is forced over it. The amount of heat
transfer from the element is directly proportional to the rate of air flow over it. In order to achieve higher efficiency
and steady-state stability, the fan speed is regulated by a software-based PID/PWM controller using Arduino Uno
controller. More over, Matlab/Simulink model for the PWM motor control is also developed to predict the speed of
the motor at various duty cycle to emulate the change in the temperature. In addition, PWM is realized with a high
switching frequency (33 KHz) in order to minimize the associated acoustic noise using Labview. Experimental tests
show acceptable results.

Keywords: Acoustic noise, Arduino control, cooling fan drive, Labview control, PID, PWM

INTRODUCTION system causes constant amperage spikes, excessive

noise and premature wear. Using soft-start technology,
During the past few years, the world has witnessed fans ramp up slowly, which eliminates harmful current
a phenomenal growth in microcontroller technology spikes. Thus the fan operates at the optimum fan rate as
and computer based control packages. In particular, the needed from 1 to 100% of its speed. This may be done
development of embedded systems has created using software-based controllers that offer an incredible
numerous possibilities to use a variety of new degree of controllability. With properly tuned controller
technology tools for many applications. The growth of it is possible to maintain systems well within 0.1°C of
these tools, their power, their variety and ease of use, set point. Unfortunately many off-the-shelf solutions
allow users to access to new techniques of control for temperature control may not be suitable because
beyond the traditional techniques. they were designed for heating or cooling hardware that
Electronic equipment has penetrated every aspect is very different. This leave the way opened for seeking
of our modern life. The reliability of the electronic alternative solutions.
components is a key factor in the overall reliability of Review on how to control cooling fan speed was
any electronic system. Inevitably, the current passing published by Burke (2003, 2004). Traditional on-off
through electronic components raise their surface temperature control strategy was tackled by Watlow
temperature and they become potential sites for Corp. (1995). More advanced P, PI and PID control
excessive heating. Unless properly designed and modes are discussed in a practical manner by
controlled, high amounts of heat generation result in Axiomatic Technologies Corporation (2003). The trend
high operating temperatures for electronic equipment, nowadays is to implement closed-loop control using
which threatens its safety and reliability. The software-based PWM techniques as published by
malfunction rate of electronic equipment increases Austriamicrosystem Co. (2010) and Texas Instruments
drastically with temperature. Therefore, thermal control (2009). In general, there are three common types for
has become increasingly important in the design and temperature control: ON-OFF control, linear (steady-
operation of electronic equipment. state) control and Pulse-Width Modulation (PWM) fan
Cooling fans are vital to systems that generate a control. When implementing two-position control, the
considerable amount of heat such as computers and cooling fan rotates at maximum speed or stops rotating,
other electronic components. Historically, cooling fans depending on the temperature set point. Main
run at 100% speed, even when less airflow is needed, disadvantages of this mode of operation, is that the
constantly turning ON and OFF as the temperature controlled variable oscillates in a continuous cyclic
changes. This reaction-based temperature control mode around the set point with an acceptable error

Corresponding Author: Tareq S. El-Hasan, Electrical Engineering Department, Zarqa University, Zarqa 132222, Jordan
This work is licensed under a Creative Commons Attribution 4.0 International License (URL: http://creativecommons.org/licenses/by/4.0/).
771
 Res. J. Appl. Sci. Eng. Technol., 13(10): 771-780, 2016

value determined by the hysteresis width of the output signal of a PID controller. In such a case, the fan
f
controller static characteristic. Another important speed can be adjusted to achieve a zero offset at the
drawback is that the cooling fan runs at maximum steady state by the virtue of the integral action. The
speed which generates high level of acoustic noise. derivative action is also very helpful, especially in slow
Another more accurate method is the linear fan motor processes such as temperature control. The derivative
drive.
ive. In this case the fan motor supply voltage is varied action accelerates the system performance
erformance and gives the
between its minimum and maximum values. For small closed-loop
loop better stability margin. Based on this, in i
power fans this may be accomplished using an this study, Arduino and Labview are both utilized to
adjustable voltage regulator or by using linear power develop PID/PWM control algorithm to drive a cooling
amplifiers. For high power fans, power operational fan DC motor. In addition, Matlab/Simulink model for
amplifiers such as OPA 512 may be used. However this PWM motor control is also developed to predict the
method of control is always associated with high power speed of a DC motor at various duty cycles to emulate
losses in the form of heat (Jacob, 1989).). Also, the fan is the change in the temperature around the motor.
designed to operate at a given voltage, whereas the In order to minimize the audible noise, the PWM
operation of the fan below this can shorten the life of generated signal will be set to a frequency of 33.3 kHz,
the motor (Leigh, 1988). which can be realized alized using the Arduino-Uno
Arduino
A better method is to use the PWM technique to controller.
control the amount of power going to the fan through a
switch mode D-type type amplifier (Grahame Holmes and MATERIALS AND METHODS
METHOD
Lipo,, 2003) which includes the power static switches in
its output stage. Such amplifiers feed the motor with the System description: Normally, the 4 wire PWM
rated voltage value where the average voltage is controlled fan is used to reduce the overall system
proportional to motor speed. Using switch mode type acoustics (Intel Corporation U.S.A.,, 2005). However, in
amplifiers excellent features are gained such as this research, a 3-wire
wire DC brushless fan motor is
increasing system efficiency and lowering fan excess
excessive controlled by adjusting pulse width of a high frequency
noise. Running the fan speed as a function of PWM in order to reduce the overall system acoustic. In
temperature means that the consumed power by the such implementation, the tachometer signal is readily
motor will be less than the consumed power at full available for closed loop speed control. Fan operating
speed. Moreover, as the amplifiers operate in a switch range is 12+1.2 V. Peak fan current during start-up
start
mode, losses in the cutoff mode or in the saturatio
saturation operation shall not exceed 2.0 A. The proposed single
mode will be minimized. This in turn unlocks the loop block diagram of the system is shown in Fig. 1.
potential for using switch mode power converters in Fan current spike during start up with 13.2 V is in the
controlling the fan speed. range of 10 A. Because of that an interface circuit is
When implementing the switch mode power supply added to enable delivering of 10 A current. The
with PWM driving signal, one has to consider system implemented motor drive circuit block diagram is
stability. If proportional control mode is used alone, a shown in Fig. 2.
definite value of system offset error will be available. As a temperature transducer, LM35 integrated
More promising results concerning system stability are circuit is already used by Texas Instruments (2016).
achieved by driving the pulse width modulator by the This transducer gives an output voltage

Fig. 1: Single loop Block diagram of temperature controlled system

Fig. 2: Implemented motor drive circuit block diagram

772
 Res. J. Appl. Sci. Eng. Technol., 13(10): 771-780, 2016

and the other is PNP transistor. Transistor 2n2222 NPN

is used as a buffer between the digital ON/OFF side and
the analog motor control side. An electrostatic
capacitor- based microphone is used as an acoustic
transducer for the evaluation of fan acoustic noise.

Software-based PWM/PID controller: By controlling

analog circuits digitally, system costs and power
consumption can be drastically reduced. What’s more is
that many controllers and DSPs, such as Arduino
already include on-chip PID control algorithm and
PWM configurable modulator. This makes their
implementation in any application very easy. The
complete code of the PWM signal and PID algorithm
are given in Appendix A. The experimental results
achieved through the complete control circuit with
Fig. 3: The schematic diagram for the power interfacing connection lines between the Arduino and the fan is
circuit presented in Appendix B at the end of this study.

Table 1: DC motor matlab simulation results vs different duty cycle Matlab simulation for DC motor PWM control: In
Motor voltage
Duty cycle % Speed (rpm) (V) Current (A) order to predict the motor speed at various
28 771 3.6 0.98 temperatures, the DC fan motor and its drive were
44.46 1700 5.33 0.964 modeled using Matlab/Simulink as shown in Fig. 4. The
80 2420 9.6 0.936
motor speed was controlled using PWM at three
proportional to temperature in Celsius Degrees (10 different levels of duty cycle (28, 44.46 and 80%,
mV/°C). The schematic diagram for the power respectively) to emulate the conditions of various
interfacing circuit between the Arduino and the fan is temperatures. PWM output signals for different values
shown in Fig. 3. of duty cycle (k) are shown in Fig. 5 whereas the
TIP147 and TIP142 is a Darlington pair of high simulation results obtained for the motor speed, voltage
power transistors (125 W). One is NPN type transistor and current are summarized in Table 1.

Fig. 4: Matlab/Simulink model for fan motor control using PWM

773
 Res. J. Appl. Sci. Eng. Technol., 13(10): 771-780, 2016

Fig. 7: Close view for the motor drive circuit and controller

(a)

Fig. 5: PWM output from Matlab/Simulink model for duty

cycle of 28%, 44.46% and 80%

(b)

Fig. 6: Prototype for the controlled DC fan motor

RESULTS AND DISCUSSI

DISCUSSION

Hardware and Experimental setup: In order to

implement the proposed control strategy, a simple
prototype was built as shown in Fig. 6. The prototype
which emulates the case of PC, consists of a box
encapsulating a heater element and a DC motor that is
driving
ing a fan through a controlled circuitry. A close
view for the electronic hardware and the controller is
shown in Fig. 7. After connecting the fan motor to the (c)
power interface circuit, the switching frequency was
adjusted to 33.33 KHz and the duty cycle wa
was set to 28, Fig. 8: Fan speed (rpm) vs temperature (C) with the
44.46 and 80%, respectively. corresponding PWM
774
 Res. J. Appl. Sci. Eng. Technol., 13(10): 771-780, 2016

Fig. 9: Motor speed; simulation vs experimental

Fig. 10: Motor voltage; simulation vs experimental

Fig. 11: Motor current; simulation vs experimental

Figure 8 shows the experimental results for PWM good match between experimental and simulation
signals and the corresponding speed, voltage and results are achieved which allows for using the Matlab
current values for different temperatures. Experimental model for further investigations and analysis.
results vs Matlab simulation results for motor speed,
voltage and current are graphically represented in Fig. 9 Labview setup: Virtual instruments have been
to 11 respectively. It can be seen from the graphs that extensively used in control as well as measurements of
775
 Res. J. Appl. Sci. Eng. Technol., 13(10): 771-780, 2016

(a)

(b)

Fig. 12: Noise level and noise density; (a): without using controller; (b): using controller

components and systems (Swain et al., 2001; Bachnak tested to measure the noise level and density without
and Steidley, 2002). For example, Labview has been using the fan controller and the results obtained are
used in severalmodern control applications (Bishop, presented in Fig. 12a, whereas the effect of controller
2012; Larsen, 2011; Resendez and Bachnak, 2003; on the noise level and density is presented in Fig. 12b.
Naghedolfeizi et al., 2002; Sokoloff, 1999). In this As can be seen from the measurement indicators,
study, to control the temperature of the system, when using fan controller, the Signal level is reduced
LABVIWE is programmed to read serial data from from 5.21 dB to 4.78 dB which is equivalent to 8%
Arduino as shown in Appendix C. whereas the THD and noise level was reduced from
Also, in order to evaluate the effect of the 82% to 57% which is equivalent to 30% reduction. The
controller on the noise level, a LabVIEW program was figure also shows that the noise amplitude and density
designed as shown in Appendix D. The system is first are effectively reduced.
776
 Res. J. Appl. Sci. Eng. Technol., 13(10): 771-780, 2016

CONCLUSION PID
//interface
This study proposes a software-based PID/PWM int timeCounter;
controller using Arduino Uno; which was developed double gap;
and utilized to drive a high efficiency low noise cooling int highfreq = 0;
fan. A prototype for the proposed system was void setup()
developed and experimentally tested and results {
obtained showed that using a high frequency PWM it // start serial port for temperature readings
was possible to effectively reduce the noise level. The Serial.begin(9600);
PID/PWM controller yields a zero steady state error and Serial.println("Start");
improves system stability margin by proper tuning of Setpoint = 28; //Inintialize desired Temperature in Deg
integral mode and differential mode gains. High C
switching frequency (33 KHz) is realized in order to //PID Setup
minimize the associated acoustic noise and the duty myPID.SetMode(AUTOMATIC);
cycle was made 28, 44.46 and 80% respectively. 16 CHAPTER 3 | IMPLEMENTATION
Matlab simulation model for the PWM DC motor //TCCR2B = TCCR2B & 0b11111000 | 0x01; //adjust
control showed a good match of the results as compared the PWM Frequency, note: this changes
to experimental results. timing like delay()
//Setup Pins
pinMode(FAN, OUTPUT); // Output for fan speed, 0 to
ACKNOWLEDGMENT
255
//interface
The authors would like to thank the deanship of
timeCounter=0;
scientific research at Zarqa University for supporting
//Setup LCD 16x2 and display startup message
this research.
lcd.begin(16, 2);
lcd.print(" Smart Fan");
Appendix A: PWM/PID codes lcd.setCursor(0,1);
lcd.print ("Starting Up");
The Code of PWM signal: delay (1000);
void setup() { lcd.clear ();
DDRD |=1<< PD3; Timer1.initialize(15); // set a timer of length 100000
} microseconds (or 0.1 sec - or 10Hz => the
void loop() { led will blink 5 times, 5 cycles of on-and-off, per
PORTD |= 1 << PD3; second)
delayMicroseconds (15); Timer1.attachInterrupt (timerIsr); // attach the service
The Code of PID: routine here
PORTD&= ~(1<<PD3); }
delayMicroseconds (15);} void loop()
#include <PID_v1.h> {
#include <LiquidCrystal.h> temp = analogRead (A0);
#include <TimerOne.h> tempvalue = temp*.488;
//Definitions timeCounter++;
#define FAN 9 // Output pin for fan //print out info to LCD
#define CRITICAL 28.00 //Critical temperature to lcd.setCursor (1,0);
ignore PID and turn on fans Serial.print ("Temp:");
int temp,tempvalue; Serial.println (tempvalue);
LiquidCrystal lcd(12, 11, 13, 5,6,7); //set up LCD lcd.setCursor (1,1);
//Setup PID Serial.println ("Set:");
double Setpoint, Input, Output; //I/O for PID Serial.println ((int)Setpoint);
double aggKp = 40, aggKi = 2, aggKd = 10; //original: 17 CHAPTER 3 | IMPLEMENTATION
aggKp = 4, //Compute PID value
aggKi = 0.2, aggKd = 1, Aggressive Turning, 50, 20, 20 if(tempvalue>Setpoint){
double consKp = 20, consKi = 1, consKd = 5; //original gap = abs(Setpoint-tempvalue); //distance away from
consKp = 1, setpoint
consKi = 0.05, consKd = 0.25, Conservative if(gap<1)
Turning,20,10,10 {
PID myPID(&Input, &Output, &Setpoint, consKp, //Close to Setpoint, be conservative
consKi, consKd, REVERSE); //Initialize myPID.SetTunings(consKp, consKi, consKd);}
777
 Res. J. Appl. Sci. Eng. Technol., 13(10): 771-780, 2016

else{ else
//Far from Setpoint, be aggresive analogWrite(FAN,255);}
myPID.SetTunings(aggKp, aggKi, aggKd);} else{ analogWrite(FAN,0);}}
myPID.Compute(); void timerIsr()
Serial.println(timeCounter); {if(highfreq==1)
Serial.println(tempvalue); {digitalWrite(3,HIGH);
analogWrite(FAN,255); highfreq=0;}
//Write PID output to fan if not critical else
if (tempvalue<CRITICAL) {digitalWrite(3,LOW);
analogWrite(FAN,Output); highfreq=1;}}

Appendix B: Complete control circuit with PWM experimental results

778
 Res. J. Appl. Sci. Eng. Technol., 13(10): 771-780, 2016

Appendix C: Labview front panel and block diagram for fan temperature control

Appendix D: Labview block diagram for fan noise measurement

REFERENCES Bachnak, R. and C. Steidley, 2002. An interdisciplinary

laboratory for computer science and engineering
Austriamicrosystem Co., 2010. Programmable PWM technology. J. Comput. Sci. Coll., 17(5): 186-192.
DC Motor Driver/controller. Datasheet # AS Bishop, R.H., 2012. Modern Control Systems with
8444/AS 8446, pp: 1-2. labVIEW. Tom Robbins, ISBN: 13-978-1-
Axiomatic Technologies Corporation, 2003. 934891-18-6.
Temperature Controller with Proportional Fan Burke, M., 2003. Programming the Automatic Fan
Drive, P/N: FC-EF-01. Technical Datasheet Speed Control Loop. Analog Devices, Application
#TD7004AX, pp: 1-2. Notes: AN-613, pp: 1-28.
779
 Res. J. Appl. Sci. Eng. Technol., 13(10): 771-780, 2016

Burke, M., 2004. Why and how to control fan speed for Resendez, K. and R. Bachnak, 2003. LabVIEW
cooling electronic equipment. Analog Dialog., programming for internet-based measurements. J.
38(2): 1-3. Comput. Sci. Coll., 18(4): 79-85.
Grahame Holmes, D. and T.A. Lipo, 2003. Pulse Width Sokoloff, L., 1999. LabVIEW implementation of
Modulation for Power Converters: Principles and ON/OFF controller. Proceeding of the ASEE
Practices. John Wiley and Sons, ISBN: 978-0-471- Annual Conference. Charlotte, NC, pp: 3561-3576.
20814-3. Swain, N.K., J.A. Anderson, M. Swain and R.
Intel Corporation U.S.A., 2005. 4-wire Pulse Width Korrapati, 2001. State-space analysis of linear,
Modulation (PWM) Controlled Fans. Datasheet time-invariant control systems using virtual
Revision 1.3, pp: 1-25. instruments. Proceeding of the ASEE Annual
Jacob, J.M., 1989. Industrial Control Electronics: Conference. Albuquerque, NM, pp: 8931-8937.
Applications and Design. Prentice-Hall Texas Instruments, 2009. Intelligent Temperature
International, Englewood Cliffs. Monitor and PWM Fan Controller. Datasheet #
Larsen, R.W., 2011. LabVIEW for Engineers. Prentice AMC6821-Q1, SBAS475, pp: 1-54.
Hall/Pearson, Upper Saddle River, N.J., ISBN: 13: Texas Instruments, 2016. LM35 Precision Centigrade
978-0-13-609429-6. Temperature Sensors. Datasheet SNIS159G, pp:
Leigh, J.R., 1988. Temperature Measurement and 1-31.
Control. Peregrinus, London, DOI: Watlow Corp., 1995. Temperature Control. The
10.1049/PBCE033E, ISBN: 9780863411113. Watlow Educational Series, Book 5, pp: 11-32.
Naghedolfeizi, M., S. Arora and S. Garcia, 2002.
Survey of LabVIEW technologies for building
Web/Internet-enabled experimental setups.
Proceeding of the ASEE Annual Conference and
Exposition. Montreal, CA, June 16-19, pp: 2248-
2258.

780