Arduino based temperature and humidity control for

condensation on wettability engineered surfaces

Rohit Gupta* Chayan Das Ankit Roy
Department of Power Engineering Department of Power Engineering Department of Power Engineering
Jadavpur University Jadavpur University Jadavpur University
Kolkata 700098, India Kolkata 700098, India Kolkata 700098, India
*Corresponding author- rohit.ju.pe@gmail.com

Ranjan Ganguly Amitava Datta

Department of Power Engineering Department of Power Engineering
Jadavpur University Jadavpur University
Kolkata 700098, India Kolkata 700098, India

Abstract—Condensation is a ubiquitous phenomenon which applications, a high condensation heat transfer coefficient is
depends on several factors, ranging from the degree of desired to improve energy efficiency. Typically it has been
subcooling to the relative humidity of the condensing seen that condensation is highly dependent on
environment. Characterizing condensation in experimental physicochemical characteristics of condenser surface, and
setups, therefore, requires a proper control strategy of the thermophysical and conditions of the condensing
operating parameters. Although industrial vapor chambers are medium [1]. Primarily there are two modes of condensation
available for carrying out condensation experiments under heat transfer, dropwise condensation (DWC), which
precisely controlled environments, these setups are predominantly occurs on hydrophobic surfaces, and
prohibitively expensive. Here we have developed a low-cost, filmwise condensation (FWC), which occurs on hydrophilic
prototype vapor chamber that is equipped with advanced surfaces. While the literature on influence of condenser
control systems. Environmental control aspects of a Styrofoam-
surface wettability on DWC and FWC in pure steam has
made condensation chamber are developed in-house for testing
been extensively studied over several decades [2], the
condensation on wettability engineered surfaces. Peltier-heat
sink arrangement is used to cool the condenser surface under literature is still replete with controversial results when it
study and the desired relative humidity is maintained within comes to the study of DWC in presence of noncondensable
the chamber by means of spraying deionized water with a gases (NCG) [3]. For characterizing such condensation
nebuliser. In order to automate the process, an Arduino process on novel surfaces, controlled temperature and
Duemilanove board is amalgamated with the setup. humidity environment chambers (EC) equipped with precise
Temperature is controlled by an ON/OFF trigger-driven control systems are required. Majority of these studies have
mechanical relay connected to the Arduino environment, which used off-the-shelf, proprietary ECs with complex
in turn generates an opportunely amplified signal to control the Proportional-Integral-Derivative (PID) control algorithm for
supply voltage of the Peltier element. A K-type thermocouple is controlling the ambient conditions. While these ECs offered
interfaced to the Arduino board with the help of MAX31855K reasonably good control, inquiry by us revealed that all
thermocouple amplifier for measuring the plate temperature. these ECs were prohibitively expensive. Here, we attempted
For humidity and chamber-temperature monitoring, an to develop a low-cost EC that is custom-designed for
SHT35D sensor is used. The relative humidity of the chamber is condensation experiments under NCG-vapor condition.
maintained by a mechanical relay-driven spray arrangement. Such an arrangement would warrant a precise control of the
The time-domain plots of humidity and plate temperature cooling plate (usually a Peltier cooler) and the temperature
response indicates that the temperature fluctuations are within and relative humidity (RH) inside the chamber.
0.25°C and RH fluctuations are within 0.5% about the set-
point. Transient response of the temperature and RH data are Several researchers have previously fabricated air de-
monitored by the Serial Monitor of Arduino software, which humidifiers, based on PID that controlled the voltage to the
indicates that the set values of temperature and RH are Peltier cooler by Pulse Width Modulation (PWM) of the
obtained approximately within 0 to 1000 seconds. signal [4,5,6,7,8]. But the Peltier elements need to be
Keywords—Condensation; Temperature Control; Humidity operated near their maximum efficiency point.
Control; SHT35D; MAX31855K Implementation of PID will increase or decrease the applied
voltage too much about the Peltier set-point, which will
I. INTRODUCTION affect their performance.
Condensation is a phenomenon that is not only To circumvent this problem, the authors have developed
omnipresent in nature, but it also plays an important role in a robust, low cost, yet accurate EC which can perform
wide range of industrial applications including power condensation experiments over a wide range of RH (from 60
generation, water desalination, industrial processes, heating to 98 %) and a surface temperature ranging from 3 to 15 °C.
ventilation and air conditioning (HVAC). In most of these The chamber is fabricated in such a way that it enables
optical imaging and Schlieren imaging of the condensation removed by an aluminum heat sink of and a brushless DC
surface in order to study the influence of vapor-NCG (BLDC) fan. The thermal contact between the Peltier
boundary layer profile on DWC or FWC heat transfer. The element and heat sink is also established by the same
whole environment control process is automated and thermal paste. The arrangement (see Fig 1) is such that the
monitored in real-time by an Arduino processor. The paper heat sink is exposed to the atmosphere whereas the TEG is
is structured as follows: in section II the proposed placed within the thick wall of the EC. An aluminum
experimental setup and its subcomponents are described. substrate (Dimension - 6.5×6.5 cm2) of known wettability is
The control strategy and connection diagrams are detailed in mounted (using thermal paste) on the inner face of the
section III. In Section IV the accuracy of the adopted control copper block, which experiences a drop in temperature as
algorithm is discussed and a set of successful condensate the Peltier element is powered. When surface temperature
collection data is also tabulated. Finally in section V and VI on the test plate drops below the dew point temperature
the observed results are summarized and some future corresponding to the environment inside the EC,
possible developments are suggested, respectively. condensation of water vapor begins on the test plate. A
fiber-glass window is installed on the opposite (to the test
II. EXPERIMENTAL SETUP plate) wall of the EC for optical imaging of the test plate.
Two windows in Line of Sight (LOS) grazing over the
condenser surface are also provided for Schlieren imaging
purpose. An effervescent atomizer-based nebulizer (Model
NE C25S) and a mixing fan arrangement is used to supply
required amount of deionized water within the EC in order
to emulate different relative humidity (RH). A K-type
thermocouple (Make - Omega) is used for measuring the
plate temperature whereas SHT35D (Make - Sensirion)
sensor module provides the temperature and RH data of
ambient condition within the EC. For comparing the
accuracy of the temperature readings of the copper block,
another K-type thermocouple is connected to a standard
DAQ (Make - Agilent). The temperature and humidity
control process is automated and monitored by an Arduino
Duemilanove development board. The details and working
principle of the different components of the controller will
be discussed in the following sub-sections.

A. ArduinoDuemilanove Development Board

The controller board used here is an Arduino
Duemilanove development board equipped with
ATMega328 running at a clock speed of 16 MHz. The
communication protocols feature UART/USART TTL (5V)
Serial communication, SPI, I2C. The board also has 14
digital I/O pins among which 6 pins can provide PWM
outputs of 8-bit/16-bit and 6 pins have a special feature of
taking analog inputs of 10-bit resolution. An
FTDI FT232RL on the board channels this serial
communication over USB.
Fig. 1. Schematic of the experimental setup

In the proposed experimental setup for characterizing

condensation, the cold side of the TEC-12706 (Make –
Hebei I.T co. ltd) Peltier element (having dimensions
4×4×0.39 cm3) is attached to a copper block (Dimension –
6×6×5 cm3) using a thermal paste (Omegatherm 201,
Omega). The opposite face of the copper block is exposed to
the interior of a Styrofoam chamber of inner dimension
31×35×22 cm3 and 6 cm wall thickness. The four side-faces
of the copper blocks are insulated using double-sided Teflon
tapes. In order to maintain the cold side of the copper block
at a nearly isothermal condition (thus ensuring one-
dimensional steady heat conduction through the block), the
Fig. 2. Arduino Duemilanove development board.
heat generated by the hot side of the TEG is continuously
B. SHT35D RH and temperature sensor D. Peltier Element
The SHT35D (see Fig. 3) is a combined temperature and The copper block is cooled by two Peltier elements
relative humidity sensing breakout board which (TEC-12706). It has maximum current and voltage ratings
communicates with ATMega328 using I2C protocol which of 6.4 Amps and 16.4 Volts. Heat is continuously extracted
can speed up to 1 MHz. It has a typical accuracy is ±1.5 % from the hot side of the Peltier by dual-fin arrangement
for RH and ±0.2°C for temperature readings. The SDA and equipped with a brushless DC Fan.
SCL pins of the breakout board are connected to the A4 and
A5 pins of the Arduino board respectively. The sensor E. Relay Modules
monitors the ambient temperature and RH % inside the Two relay breakout boards (see Fig. 5) are used, each
vapor chamber. An SF2 filter cap is used to prevent having two relays. One of the breakout board controls the
penetration of water droplets inside the sensor. supply to the Peltier elements based on the temperature set-
point whereas the other board controls the stirring fan and
fogger arrangement. The Peltier relays control inputs from
two DC power supplies which are operated at 7 volts and
can supply up to 5 Amps. The Mixing fan relay bypasses 12
volt, 500 mA from another DC supply module whereas the
fogger control relay directly bypasses the 230 V, 50 Hz
household AC supply. All the relays are capable of handling
the different supply lines.

Fig. 3. SHT35-D RH and temperature sensor module (by Sensirion)

C. K-type thermocouple and MAX31855K amplifier

In order to monitor and control the temperature of the
wettability-engineered test surface of condenser, a K-type
thermocouple is installed at the interface of cold side
exposed to the vapor chamber and test surface with the help
of thermal paste. The thermocouples provide a differential
voltage across its terminals for each degree of temperature
change. In order to decode that change a MAX31855 (see
Fig. 4) cold-Junction Compensated Thermocouple-to-
Digital Converters by Maxim Integrated were used to
Fig. 5. Relay Module.
acquire the generated Seebeck voltages. This converter
exhibits a 14-bit resolution (or 0.25°C) and a thermocouple
accuracy of ±2°C for temperatures ranging from -200°C to F. Power supply modules
+700°C for K-type. In order to perform ground isolation for obtaining
The communication between ATMega328 and the uninterrupted thermocouple readings, the development
MAX31855K module is based on SPI (Serial Peripheral board, sensors, and logic trigger side of the relays are
Interface), a synchronous interface ideal for short-distance powered by a 5 volt, 10000 mAh battery. The supply trigger
communication. In our case we have a single master - single sides are connected to two separate variable DC supplies.
slave configuration where the master is the Arduino UNO The mixing fan relay is fed with a 12 volt, 500 mA DC
board while the two slaves are the two MAX31855K, supply. The Heat sink cooling fan is driven by a 12 volt, 1
selected acting on the respective ChipSelect pin. Every amp DC adapter.
second, using the SPI interface of the Arduino Duemilanove
board, a vector of temperature values is acquired and used III. CONTROL STRATEGY
for further computations in the control loop.
A. Connection layout and control sequence :-
The detailed discussion in previous section indicates that
the SHT35D should be interfaced to the I2C bus,
MAX31855 to the SPI bus and the relays will use digital
output pins of ATMega328 for their logical voltage trigger
side. Fig. 6 represents the connection diagram. The adopted
control algorithm is represented in Fig. 7. The control loop
for surface temperature is a single layer cascaded system
whereas for humidity control it is a double layer cascaded
control loop.
Fig. 4. MAX31855K thermocouple voltage level booster.
 IV. RESULTS AND DISCUSSION

A. Time-domain plot of condenser surface temperature:-

Fig. 6. Detailed connection layout.

Fig. 8. Time domain response of temperature.

The time response (see Fig. 8) of condenser surface

temperature sensor output was plotted at different set-points
at which the condensation experiments are to be performed.
The sensor shows a nearly constant ramp down rate of
2.1°C/min for wide range of set-points. It also quantifies the
time at which the condenser surface reaches the set-point
from the start of experiment. This time can be termed as the
start-up time for the condensation experiments.

B. Time domain plot of chamber RH:-

Fig. 9. Time domain response of RH without condensation : Effect of

mixing.

For maintaining desired RH within the chamber, two

methods are adopted and compared with the help of time
domain response of RH. First method is by directly spraying
a fog of deionised water within the chamber and allowing
diffusive transport to homogenize the vapor concentration
within the EC. In the second method, a low-speed BLDC
Fig. 7. Process control algorithm. fan is used to augment mixing of the sprayed fog droplets
within the EC. The fog-inlet port and the mixing fan are Table- 1
installed far away from the condensing surface in order to
Surface Time Plate Chamber RH (%) Water
have minimal flow disturbance near the condenser plate and Type of Temp Temp (°C) collecti
ensure natural convection at the vicinity of condenser plate. test (°C) on
Clearly, the time domain curve (see Fig. 9) indicates that (hrs) (gms.)
mixing fan arrangement increases the chamber humidity at a Hydrophilic 4 8± 18±0.4 70±0.5 0.12
much faster rate and also minimizes the large amplitudes of CA-81°± 1 0.2
RH overshoot about the set-point. Therefore, the mixing fan
arrangement is chosen for the final experimental setup. Hydrophilic 4 8± 18±0.4 85±0.5 1.3
CA-81°±1 0.2
Hydrophilic 4 8± 18±0.4 95±0.5 1.68
CA-81°±1 0.2

V. CONCLUSION
A simple, yet accurate control system without
implementing PID algorithm has been implemented to
design a low cost controlled humidity environment
chamber, custom-designed to perform condensation
experiments at user defined surface temperature and
humidity set-points. An Arduino Serial Monitor acquired
the real time data of the chamber humidity, chamber
temperature and condenser surface temperature. The
accuracy and stability of the experimental setup for
characterizing condensation depended on the environment.
The time response plots indicate that the temperature
fluctuations are within 0.25° C and RH fluctuations are
Fig. 10. Time domain response of RH : Effect of condensation (With within 0.5% about the set-point, which is adequate for the
Mixing) experimental accuracy for the study of condensation on
wettability engineered surface in presence of
The comparison of chamber humidity response for runs noncondensable gases.
with and without condensation (see Fig. 10) is made to
ensure that the start-up time is minimal even when there is a
constant consumption of water (i.e., condensation takes VI. FURTHER DEVELOPMENT
place) while the surface temperature is ramping down to The future aspects of improving the environment
reach the set-point. For an RH set-point of 70% the chamber include varying the ambient dry bulb temperature
increment of charging time is nearly 50 seconds when there with a help of a convective strip heater with the help of PID
is simultaneous humidity charging and plate temperature algorithm. Further, for on-board process parameter
ramp down. The same for an RH set-point of 95% is monitoring, a 16×2 LCD module can be interfaced with the
approximately 400 seconds, which is negligible with respect Arduino. Lastly for real time condensate collection
to total experimental time of 4 hours. A few random time measurement a strain gauge load cell is to be installed.
sampling was done for measuring the time duration of spray
(water addition time) and condensation (water consumption
time) with a stopwatch. The spray time is nearly equal to 20 Acknowledgment
seconds irrespective of the surface temperature but the The authors gratefully acknowledge the funding from DAE-
consumption time is a strong function of the difference of BRNS through the Project No. 36(1)/14/24/2016-BRNS.
chamber and surface temperature (the degree of
subcooling). Increasing the degree of subcooling will References
decrease the water consumption time and will result in
frequent humidification of the chamber.
[1] D. Attinger, C. Frankiewicz, A. R. Betz, T. M. Schutzius, R.
Ganguly, A. Das, C-J. Kim, C. M. Megaridis, “Surface engineering
C. Condensate collection data:- for phase change heat transfer: A review”, MRS Energy &
Condensation experiments were performed at different Sustainability: A Review Journal, 2014, vol-1, pages 1-40.
environment and surface wettability (varying contact angle) [2] N. Miljkovic, E.N. Wang, Condensation heat transfer on
superhydrophobic surfaces, MRS Bulletin, 38, 397-406, 2013.
conditions. The measured quantity to characterize [3] Das, C., Roy, A., Datta A., Ganguly, R., Mukhopadhyay, D.,
condensation under various conditions is the amount of Bandyopadhyay, K., "Effect of surface wettability on condensation
condensate collected over a particular time. The area of the heat transfer in presence of non-condensable gas," Proc. 24th Nat.
condensation surface is 6.5×6.5 cm2. The salient data are 2nd Int. ISHMT-ASTFE Heat Mass Transfer Conf. (IHMTC), 27-30
tabulated in Table-1. 2017, Paper no. IHMTC2017-13-1264
[4] G. Andria, G. Cavone, C. G. G. Carducci, M. Spadavecchia, A.
Trotta, “A PWM temperature controller for themoelectric generator
characterization.” Proc. IEEE Metrol. Aerosp. (MetroAeroSpace),
Florence, Italy, Jun. 2016, pp. 295–300.
[5] J. Rusia, A . Naugarhiya, S. Majumder, B. Acharya, S. Verma.
“Remote temperature & humidity sensing through ASK modulation
technique”. ICT in Business Industry & Government (ICTBIG).
Proceedings OF 2016 International Conference on ICT in Business
Industry & Government (ICTBIG), Indore, India, pp. 1–6.
[6] M. A. M. Monel, M. M. Kamal, M. H. Omar . “Heat Sink Design of
Thermoelectric Module for Cooling System”. 2016 6th IEEE
International Conference on Control System, Computing and
Engineering, 25–27 November 2016, Penang, Malaysia .
[7] H. G. Ramos, F. Assuncao, A. Ribeiro, P. M. Ramos, “A low-cost
temperature controlled system to test and characterize sensors”,
IEEE Africon 2004, Gaborone, Botswana, vol. 1, pp. 457-460, Sept.
2004.
[8] A. Rocha et al. “Air Dehumidifier Controlled by Arduino Using
Peltier”. Recent Advances in Information Systems and
Technologies, Advances in Intelligent Systems and Computing 569.