Engineering, 2010, 2, 263-269
doi:10.4236/eng.2010.24036 Published Online April 2010 (http://www.
Copyright © 2010 SciRes. ENG
Electrical Performance Study of a Large Area
Multicrystalline Silicon Solar Cell Using a Current
Shunt and a Micropotentiometer
Hala Mohamed Abdel Mageed1, Ahmed Faheem Zobaa2, Ahmed Ghitas3, Mohamed Helmy Abdel
Raouf1, Mohamed Sabry3, Abla Hosni Abd El-Rahman1, Mohamed Mamdouh Abdel Aziz4
1National Institute for Standards, Giza, Egypt
2University of Exeter, Exeter, UK
3National Research Institute of Astronomy and Geophysics, Helwan, Egypt
4Cairo University, Cairo, Egypt
E-mail:{halaabdelmegeed, mohammed_makka},,
Received November 30, 2009; revised February 12, 2010; accepted February 19, 2010
In this paper, a new technique using a Current Shunt and a Micropotentiometer has been used to study the
electrical performance of a large area multicrystalline silicon solar cell at outdoor conditions. The electrical
performance is mainly described by measuring both cell short circuit current and open circuit voltage. The
measurements of this cell by using multimeters suffer from some problems because the cell has high current
intensity with low output voltage. So, the solar cell short circuit current values are obtained by measuring the
voltage developed across a known resistance Current Shunt. Samples of the obtained current values are ac-
curately calibrated by using a Micropotentiometer (μpot) thermal element (TE) to validate this new measur-
ing technique. Moreover, the solar cell open circuit voltage has been measured. Besides, the cell output
power has been calculated and can be correlated with the measured incident radiation.
Keywords: Large Area Multicrystalline Silicon Solar Cell; Current Measurements; Calibration; Current
Shunt; Micropotentiometer; Short Circuit Current; Open Circuit Voltage
1. Introduction
There are many types of the solar cells that are used in
different life applications. The main important types of
the solar cells are manufactured by some back-contact
techniques [1-3]. Back-contact solar cells have some
advantages over ordinary solar cells due to their lower
cost and their higher efficiency [4,5]. Moreover, charac-
teristics of these back-contact solar cells are studied to
enhance their performance [6,7].
Short circuit current and open circuit voltage are two
main electrical parameters usually used to characterise
solar cells. Typically, these quantities are measured by
multimeters. During the solar cell current measurements
some troubles appeared, because the solar cell under test
produces high current intensity with low output voltage.
So, the multimeters are not suitable for the solar cell
output current measurements [8]. In that study, a Hall
sensor technique is applied in order to overcome the
problem of the effect of multimeters internal resistance
in these measurements. However, the Hall sensor has
some limitations and precautions; such as the depend-
ence of its performance on operating ambient tempera-
ture, the quality and stability of the supply voltage and
the linearity limitations through a specified current range
[8]. It is also very sensitive to external magnetic fields
and its offset is not stable and may vary with temperature
and time [9].
Current sensors play a vital role in our life. At present,
comprehensive research concerned current sensing tech-
nology has been conducted, including current shunts [10].
They are used in many applications to measure current
by measuring the voltage developed across known im-
pedance [11].
In this paper, a current shunt is simply used as a sensor
to get the solar cell current values by measuring the volt-
age developed across its known resistance. Hence, a new
technique using Holt HCS-1 current shunt is applied to get
the short circuit currents of a multicrystalline silicon solar
cell with back contact technology. Then, accurate calibra-
tions are carried out using a μpot thermal element to get
the actual precise values of the measured currents.
A comparison between computed versus experimen-
tally corrected and calibrated values of the cell short cir-
cuit currents is carried out; hence, the new current meas-
urement system is confirmed and verified. This study is
extended to measure the cell open circuit voltage; then,
the cell output power is computed to be correlated with
the incident radiation profile. All results are carefully
studied through some representing mathematical curves.
2. Solar Cell Under Test
The multicrystalline silicon solar cell with back contact
technology is shown in Figure 1. It has a large area of
21cm × 21cm. The module was installed in a tilted posi-
tion at the optimum tilt angle of the location of study
[12], in the outdoor. The cell current is collected by the
fine finger grid which is led to the back side through 25
holes. On the back side there are 25 soldering pads for
each polarity.
The outdoor cell electrical performance is studied by
measuring both short circuit current and open circuit
voltage in the tilted position at Helwan, Egypt.
3. Measurements and Calibrations
Measurement is a set of operations performed on a
physical object or system according to an established
documented procedure, for determining some physical
property of the object or system. Science and technology
progress is based on the development of measurements.
Calibration is achieved by comparing a measurement
device (unknown) against an equal or better standard.
3.1. Current Shunt Characteristics
In order to measure current with high accuracy, a current
shunt is commonly used. Four-terminal resistors' current
shunts are in wide use in the metrology community and
in industrial measurement applications. Such applica-
tions include the measurement of DC, and AC electric
currents [13].
They are commonly used in high current low voltage
applications. Shunts often have low resistance value and
low temperature coefficient of resistance and use Kelvin
terminals for improving measurement accuracy [10].
They are the most cost effective sensing elements, hav-
ing compact package profiles, suitable for DC or AC
measurement. These shunts have as their major design
goals adequate power dissipation and minimal resistance
changes with temperature. Also considerations are taken
to minimize thermoelectric errors of the 4-terminal re-
sistance [13].
Special resistance alloys, such as Constantan, Man-
ganin, and Zeranin, have been formulated to have very
small temperature coefficients of resistance to combat
Figure 1. Multicrystalline silicon solar cell in the out-
door located at optimum tilt angle with the radiation
the rise of temperature in shunts [14].
One feature of the current shunts is that it converts the
applied current to voltage drop across its terminals in a
linear manner [13].
3.2. Solar Cell Current Measurements
Using a Current Shunt Linearity
The cell current measurements are obtained by using a
data logger and a current shunt linearity curve. Then a
μpot thermal element is used to get the actual calibrated
current values accurately.
Different current shunt products, like Fluke Model
(A40, A40A, A40B) and Holt Model HCS-1 current
shunts are used to simplify the task of making precise
current measurements in the laboratory.
For this task, Holt HCS-1 current shunt 20 Ampere
range shown in Figure 2, is used to measure the cell
short circuit current (ISC). It is of a coaxial design; the
resistor being a web of wire arranged coaxially about the
axis of the shunt. The input terminal is a female UHF
connector at one end, and the potential terminal is a male
UHF at the other [15].
This device is used to obtain the equivalent voltage
drop across its resistance structure when the short circuit
current is applied. Then, this equivalent voltage is ap-
plied to a 14 bit data logger, which is in turn connected
to a PC to compute the corresponding current using the
shunt linearity curve. Figure 3 illustrates the linearity
curve of the 20A HCS-1 current shunt, which shows that
its output voltage is linearly proportional to its input cur-
The linearity equation that relates the input current to
the output voltage is:
 IV inout (1)
Where Vout is the output voltage drop across the
Copyright © 2010 SciRes. ENG
H. M. A. MAGEED ET AL. 265
Figure 2. Holt HCS-1 current shunt with 20 A range.
y = 0.0518x + 5E-05
0510 15 20
I/P Current (A)
O/P Volt (V)
Figure 3. linearity curve of the 20A HCS-1 current shunt.
current shunt resulted from applying input current source
The cell short circuit current (ISC) can be computed by
using the previously estimated linearity equation with the
voltage drop across the shunt element Vout as an input to
the equation. Computed daily profile of the solar cell
short circuit current ISC is depicted in Figure 4.
This current curve is obtained in terms of the shunt
voltage signals transferred to the PC through a 14Bit data
logger. These voltage signals are accurately measured to
get the actual voltage values experimentally, which are
used to obtain the corresponding actual current values
using the μpot thermal element.
Practically, samples of the voltages, which present the
short circuit currents of the cell, is measured experimen-
tally by using a precise digital multimeter (Fluke 8508A-
DMM) to get their actual calibrated values accurately.
Table 1 illustrates the voltage results obtained from
data logger and the actual calibrated results which meas-
ured at the same time by using the DMM.
Figure 5 shows the two voltage patterns, one of them
is for the data logger voltages and the other is for the
actual calibrated voltages with respect to the calculated
short circuit currents obtained from the current shunt
linearity equation.
The actual values of the cell short circuit currents can
be obtained accurately by calibrating them by using a
μpot thermal element and the calibrated voltages. There-
4:489:3614:24 19:12
Local Time
Isc (Ampere)
Figure 4. Computed cell short circuit current.
Table 1. Actual calibrated and data logger voltages.
Actual Calibrated Voltage
Data Logger
0.03300 0.02905
0.09720 0.09253
0.12320 0.11645
0.15940 0.15310
0.16000 0.15601
0.16130 0.16060
0.18645 0.18042
0.18745 0.18240
0.18845 0.18280
0.21845 0.21069
0.21940 0.21480
0.22550 0.21655
0.22850 0.22168
0.23655 0.22973
0.23745 0.23730
0.24445 0.23802
Computed SC Current (A)
Voltage (V)
Actual Calibrated Voltage
Data Logger Voltage
Figure 5. Data logger voltages and calibrated voltages used
to obtain ISC.
fore, the computed current values obtained in Figure 5,
are calibrated by the actual current values that obtained
by using the μpot thermal element.
Copyright © 2010 SciRes. ENG
t 5 mA.
3.3.Solar Cell Current Calibration Using a μpot
thermal element
The schematic diagram of the pot which consists of a
thermal element in series with a non reactive radial re-
sistor is shown in Figure 6. The radial resistor is se-
curely soldered into the output N-type coaxial connector
and screwed into the pot case. It has two outputs, one for
the output thermal electromotive force (e.m.f), and the
other for the output voltage where as in National Metrol-
ogy Institutes Pots are basically voltage sources [16].
The pot with single-range output resistor can be
connected as a pot, as a thermal voltage converter
(TVC), or as a thermal current converter (TCC) when
attached with a current shunt as demonstrated in Figure
7. Actually, this flexibility broadens the range of useful
applications of this device because its rated current is
The core of the pot is the thermal element that is
shown in Figure 8. It consists of a thin lament-heater
and a thermocouple inserted in an evacuated glass bulb.
The thermocouple thermally contacts the heater at its
midpoint using a bead made of electrically insulating
material such as glass or ceramics [17].
The basic measurement principle of the thermal ele-
ment is based on converting the electrical signal to a heat
power. In such converters, energy dissipated by a current
flowing through a heater resistor, raising its temperature
above the ambient, is compared to the energy dissipated
by the voltage flowing through the same heater.
The increase in the temperature of the heater at voltage
and current, proportional to the dissipated energy, is
measured using a thermocouple. Therefore, when voltage
or current is applied to the input of a thermal element it
gives e.m.f. At the same output e.m.f.s for both of the
two inputs, we can say that this applied input current is
corresponding to the applied input voltage.
Accordingly, the actual values of the voltages corre-
sponding to the cell short circuit currents that listed in
table 1 are applied to the pot thermal element by using
the Wavetek 9100-Calibrator. It is used in the voltage
mode as a traceable standard DC source. Then, the ther-
mal element output e.m.f.s are measured by using a pre-
cise digital multimeter (Fluke 8508A-DMM) as shown in
Figure 9.
In the second step, the current is applied from the
same calibrator, but in the current mode, to the pot
thermal element through the 20A current shunt to attain
the same output e.m.f.s obtained in the first step as dem-
onstrated in Figure 10.
Therefore, these currents represent the actual cali-
brated values of the corrected short circuit currents of the
solar cell as listed in Table 2.
The actual calibrated short circuit currents, the corre-
Figure 6. A Simple construction of the pot.
Figure 7. Thermal current converter (TCC) attached with a
current shunt.
Figure 8. Structure of a Thermal Element (TE).
Figure 9. Measurements of the thermal element output emfs
for the voltages corresponding to the ISC.
Copyright © 2010 SciRes. ENG
H. M. A. MAGEED ET AL. 267
Figure 10. Calibration System of the Short Circuit Current.
Table 2. Actual calibrated short circuit currents against
actual calibrated equivalent voltages of the shunt at the
same output e.m.f.s.
Actual calibrated
O/P emf
Actual calibrated
Current (A)
0.03300 0.01778 0.54810
0.09720 0.09492 1.83410
0.12320 0.14780 2.33680
0.15940 0.24196 3.03520
0.16000 0.24360 3.04660
0.16130 0.24705 3.07040
0.18645 0.32716 3.55440
0.18745 0.33570 3.57420
0.18845 0.33390 3.59380
0.21845 0.44532 4.17147
0.21940 0.44897 4.18907
0.22550 0.47372 4.30827
0.22850 0.48601 4.36527
0.23655 0.51979 4.51737
0.23745 0.52342 4.53517
0.24445 0.55364 4.66757
sponding computed short circuit currents (obtained from
data logger and current shunt linearity equation) and the
computed relative error between them in percentage are
recorded in Table 3 and are illustrated in Figure 11.
The percentage errors between actual and computed
short circuit currents don’t exceed 0.04% which means
that the computed results of the short circuit current are
closed to the actual calibrated results.
The consistency between the actual and the computed
current curves is clearly demonstrated in Figure 11.
Therefore, the system that consists of the current shunt
and the data logger is accurate, precise and reliable in the
solar cell current measurements, especially at such high
currents with low voltages.
4. Accurately Measured Solar Cell Electrical
The cell open circuit voltage signals are received by the
data logger to be transferred to the PC through the pre-
pared computer program. After the previously discussed
measurement calibration techniques, accurate solar cell
output power could be obtained. A daily variation of
Table 3. The actual calibrated short circuit currents and
the corresponding computed short circuit currents of the
Actual calibrated
Short Circuit Currents
Computed Short
Circuit Currents
Relative Error (%)
0.54810 0.55985 - 0.02143
1.83410 1.78533 0.02659
2.33680 2.24710 0.03838
3.03520 2.95463 0.02654
3.04660 3.01081 0.01175
3.07040 3.09942 - 0.00945
3.55440 3.48205 0.02035
3.57420 3.52027 0.01509
3.59380 3.52799 0.01831
4.17147 4.06641 0.02518
4.18907 4.14575 0.01034
4.30827 4.17954 0.02988
4.36527 4.27857 0.01986
4.51737 4.43397 0.01846
4.53517 4.58011 - 0.00991
4.66757 4.59415 0.01572
051015 20
Isc (A)
Number of readings
Actual Calibrated sc Current
Computed sc Current
Figure 11. Actual calibrated and computed short circuit
currents ISC.
large area 21cm 21cm multicrystalline silicon solar cell
short circuit current, open circuit voltage and power are
plotted in Figure 12.
As seen in this figure, the maximum open circuit
voltage value is nearly 0.5V, while the maximum short
circuit current at the same time is nearly 4.6A. This is the
main distinguishing property of this solar cell.
Figure 13 shows the cell electrical output power along
with the solar radiation intensity incident on its surface in
case of the optimum tilted orientation. The data has been
recorded in 8th March 2009 which corresponds to a tilted
angle 30о.
The incident radiation is recorded by using CMP3
Kipp&Zonen, which is also connected to the data logger
after signal amplification.
Copyright © 2010 SciRes. ENG
4:487:129:3612:0014:24 16:48 1
Local Time
Isc (A), V0c (V), Power (W
Figure 12. Daily variation of ISC , VOC and Power of the so-
lar cell.
4:487:129:3612:00 14:2416:48 19:12
Radiation (W/m2)
Power (W)
Local Time
Radiati o n
Figure 13. Daily variation of the incident solar radiation
against the solar cell output power.
5. Conclusions
A new current measurement technique of a back contact
large area multicrystalline silicon solar cell in outdoor
conditions is introduced. The 20 Ampere, Holt HCS-1
current shunt is used for this aim to avoid the other
measuring techniques’ problems. It represents accurate,
easy, cheap, and reliable way to get high current values
at low voltages. The concept of this current shunts; is ob-
taining the current values by using the corresponding
measured voltages developed across its known resistance.
In order to validate this new current measuring tech-
nique, samples of short circuit current values are accu-
rately obtained and practically calibrated to get their ac-
tual precise values by using a μpot thermal element. A
comparison between the accurate calibrated short circuit
current results and the computed results demonstrates an
excellent agreement between them to about 0.04% rela-
tive error. Then the cell electrical output power could be
computed easily.
. References
[1] W. Jooss, K. Blaschek, R. Toelle, T.M. Bruton, P. Fath
and E. Bucher, “17% Back Contact Buried Contact Solar
cells,” Proceedings of the 16th EPVSC, Glasgow, 2000,
pp. 1124-1127.
[2] D.L. Meier, H. P. Dabis, A. Shibata, T. Abe, K. Kinoshita,
C. Bishop, S. Mahajan, A. Rohatgi, P. Doshi and M. Fin-
negan, “Self Doping Contacts and Associated Solar Cell
Structures,” Proceedings of the 2nd WCPSEC, Vienna,
1998, pp. 1491-1494.
[3] K. Faika, M. Wagner, P. Fath and E. Bucher, “Simplifi-
cation of EWT (Emitter-wrap Through) Solar Cell Fabri-
cation Using Al-P-Codiffusion,” Proceedings of the 28th
IEEE PVSC, Anchorage, 2000, pp. 176-179.
[4] A. Kress, P. Fath, G. Willeke and E. Bucher, “Low-cost
Back Contact Silicon Solar Cells Applying the Emit-
ter-wrap Through (EWT) Concept,” Proceedings of the
2nd WCPSEC, Vienna, 1998, pp. 1547-1550.
[5] W. Neu, A. Kress, W. Jooss, P. Fath and E. Bucher,
“Low-cost Multicrystalline Back-contact Silicon Solar
Cells with Screen Printed Metallization,” Solar Energy
Materials & Solar Cells, 2002, pp. 139-146.
[6] J. H. Bultman, A. W. Weeber, M. W. Brieko, J. Hoonstra,
J. A. Dijkstra, A. C. Tip and F. M. Schuurmans, “Pin Up
Module: A Design for Higher Efficiency, Easy Module
Fabrication and Attractive Appearance,” Proceedings of
the 16th EPVSC, Glasgow, 2000, pp. 1210-1213.
[7] A. Schonecker, A. Weeber, W. Sinke, C. Zechner, A.
Kress and P. Fath, “Attacking Limiting Factors in 10 × 10
cm2 Multicrystalline Silicon, Emitter Wrap-through Solar
Cell Design and Processing,” Proceedings of the 2nd
WCPSEC, Vienna, 1998, pp. 1677-1680.
[8] A. Ghitas and M. Sabry, “Performance of a Large Area
Back Contact Silicon Solar Cell in a Desert Zone,” Pro-
ceedings of the Cairo 11th International Conference on
Energy and Environment, Hurghada, 2009.
[9] J. Trontelj, “Optimization of Integrated Magnetic Sensor
by Mixed Signal Processing,” Proceedings of the 16th
IEEE Instrumentation and Measurement Technology
Conference, Vol. 1, 1999, pp. 299-302.
[10] C. C. Xiao, L. Y. Zhao, T. Asada, W. G. Odendaal and J.
D. van Wyk, “An Overview of Integratable Current Sen-
sor Technologies,” IEEE Industry Application Confer-
ence, Vol. 2, 2003, pp. 1251- 1258.
[11] R. G. Jones, P. Clarkson and A. J. Wheaton, “Calibration
of Ac Current Shunts,” Precision Electromagnetic Meas-
urements Digest Conference, Sydney, 2000, pp. 492-493.
[12] H. K. Elminir, A. E. Ghitas, F. El-Hussainy, R. Hamid, M.
M. Beheary and K. M. Abdel-Moneim, “Optimum Solar
Flat-plate Collector Slope: Case Study for Helwan, Egypt,”
Energy Conversion & Management, 2006, Vol. 47, pp.
[13] D. E. Destefan, R. S. Stant, “AC and DC ShuntsCan
You Believe Their Specs?” IEEE Instrumentation and
Measurement Technology Conference, Vol. 2, 2003, pp.
[14] M. Handbook, “Stainless Steel Materials and Special
Purpose Metals,” 9th Edition, American Society for Met-
als, Detroit, pp. 640-645.
Copyright © 2010 SciRes. ENG
Copyright © 2010 SciRes. ENG
[15] T. E. Lipe, “Operation and Reference Manual for the
NIST Automated AC-DC Calibration Systems and Soft-
ware,” National Institute of Standards and Technology
Special Publication, May 2004.
[16] M. Klonz, G. Schliestedt, T. Funck, M. Stojanovic and B.
Stojanovic, “Millivolt Calibrations Using Micropotenti-
ometers with New Disc Resistors and Planar MJTCS
Aided by New Millivolt Ampliers,” IEEE Transactions
on Instrumentation and Measurement, Vol. 54, No. 2,
April 2005, pp.787-790.
[17] H. Sasaki, K. Takahashi, B. D. Inglis and M. Klonz, “A
Numerical Simulation of Thermoelectric Effects in Sin-
gle-Junction Thermal Converters,” IEEE Transactions on
Instrumentation and Measurement, Vol. 48, No. 2, April
1999, pp. 408-411.