Time domain technique to measure solar cell

capacitance
Cite as: Review of Scientific Instruments 74, 3516 (2003); https://doi.org/10.1063/1.1582391
Submitted: 04 February 2003 . Accepted: 17 April 2003 . Published Online: 24 June 2003

R. Anil Kumar, M. S. Suresh, and J. Nagaraju

ARTICLES YOU MAY BE INTERESTED IN

Facility to measure solar cell ac parameters using an impedance spectroscopy technique

Review of Scientific Instruments 72, 3422 (2001); https://doi.org/10.1063/1.1386632

Measurement of solar cell ac parameters using the time domain technique

Review of Scientific Instruments 75, 2732 (2004); https://doi.org/10.1063/1.1777380

Impedance spectroscopy of thin-film CdTe/CdS solar cells under varied illumination

Journal of Applied Physics 106, 044507 (2009); https://doi.org/10.1063/1.3204484

Review of Scientific Instruments 74, 3516 (2003); https://doi.org/10.1063/1.1582391 74, 3516

© 2003 American Institute of Physics.

REVIEW OF SCIENTIFIC INSTRUMENTS VOLUME 74, NUMBER 7 JULY 2003

Time domain technique to measure solar cell capacitance

R. Anil Kumar and M. S. Suresh
ISRO Satellite Centre, Bangalore 560 017, India
J. Nagarajua)
Department of Instrumentation, Indian Institute of Science, Bangalore 560 012, India
共Received 4 February 2003; accepted 17 April 2003兲
The solar cell capacitance is one of the important parameters for design of a reliable and efficient
switching charge controller. In this article a time domain technique is used to measure the solar cell
capacitance 共charge equivalent兲. The measurements carried out on GaAs/Ge and BSFR silicon solar
cells at different cell voltages are presented and compared with the charge equivalent capacitance
derived from the impedance spectroscopy technique. It is observed that the time domain technique
is simple, reliable, and can be used for on-line measurements of different types of solar cell/panel
capacitance. © 2003 American Institute of Physics. 关DOI: 10.1063/1.1582391兴

I. INTRODUCTION

The capacitance of a solar cell is one of the important

C q⫽
Q
⫽
1
Vd Vd
冕0
Vd
i dt 共3兲

parameters in the design of high speed switching charge con- also C q is

trollers suitable for a photovoltaic power system. Anil Kumar
et al.1 measured the cell capacitance of GaAs/Ge and silicon
共BSR and BSFR兲 solar cells and shown that the solar cell
C q⫽
1
Vd
冕
0
Vd
Cp d␯. 共4兲
capacitance is a function of operating voltage and tempera-
In this article a technique to measure solar cell capaci-
ture. The solar cell capacitance is nonlinear2,3 and can be
tance 共charge equivalent capacitance兲 using the time domain
measured by a frequency domain technique or a time domain
共transient兲 technique is presented along with the measure-
共transient兲 technique. Anil Kumar et al.4 measured the cell
ments carried out on space quality GaAs/Ge and BSFR sili-
capacitance using impedance spectroscopy a frequency do-
con solar cells. These measured values are compared with
main technique by applying a small ac signal about the op-
the charge equivalent capacitance derived from cell capaci-
erating point over a wide range of frequencies. In time do-
tance, measured using the impedance spectroscopy
main technique, the cell voltage or current decay/buildup is
technique.3
monitored from short circuit to open circuit or vice versa.
The total charge stored in any nonlinear capacitor 共such
as solar cell兲 depends on the cell voltage and can be written II. DESCRIPTION OF THE TEST SETUP
as
The experimental setup shown in Fig. 1 consists of a dc

冕 Vd power supply, digital multimeter, digital storage oscilloscope

Q⫽ C pd ␯ , 共1兲 interfaced to a personal computer 共PC兲, custom-built shunt
0
switch, and the solar cell whose capacitance is to be mea-
sured. The solar cell is kept under dark conditions and is
where Q is the total charge stored, C p is solar cell capaci-
biased by the dc power supply, as measured by the digital
tance 共function of voltage兲 measured using impedance spec-
multimeter. The current that passes through the solar cell is
troscopy, and V d is voltage across capacitance.
measured by voltage drop across the noninductive sense re-
The current due to stored charge is

dQ
i⫽ . 共2兲
dt

Equations 共1兲 and 共2兲 explicitly recognize the change in

both capacitance and voltage as a function of time. A linear
capacitor (C q ), which stores the same amount of charge as
the nonlinear cell capacitor at the operating voltage V d can
be defined by Eq. 共3兲. This is called charge equivalent ca-
pacitor of the cell and given by

a兲
Author to whom correspondence should be addressed; electronic mail:
solarjnr@isu.iisc.ernet.in FIG. 1. Experimental setup.

0034-6748/2003/74(7)/3516/4/$20.00 3516 © 2003 American Institute of Physics

Rev. Sci. Instrum., Vol. 74, No. 7, July 2003 Time domain capacitance measurement 3517

TABLE I. Measured capacitor value of network-1.

Capacitance

Cell Standard capacitor

voltage Time domain value
共V兲 共␮F兲 共␮F兲 % error

1 0.117 ⫺17
0.8 0.118 ⫺18
0.6 0.122 0.10 ⫺22
0.4 0.127 ⫺27
0.2 0.136 ⫺36

FIG. 2. Standard network for calibration.

共charge equivalent兲 is estimated/calculated. To minimize the
error due to noise, the measurements at a particular voltage
sistor (R s ) and acquired into the digital storage oscilloscope are repeated several times and the average value is taken. To
共Tektronix, Model: TDS 220兲. A pulse generator with driver reduce errors due to lead resistance and inductance, a
circuit is used to switch ‘ON’ the metal–oxide– shielded coaxial cable of small length is used for connecting
semiconductor field effect transistor, which shunts the solar the solar cell.
cell at high speed. The resistor (R 1 ) limits the power supply
current when the electronic switch is closed and the blocking
III. CALIBRATION OF EXPERIMENTAL SETUP
diode (D 1 ) blocks the discharge of the solar cell capacitor
through the power supply. The digital storage oscilloscope The test setup is calibrated using a simulated solar cell
interfaced to the PC through RS-232 communication port network shown in Fig. 2 with passive components. The dc
acquires the transient current verses time data when the bias voltage from 0.2 to 1 V in steps of 0.2 V is applied and
switch is closed. From these data the solar cell capacitance measurements are carried out. The variation in current with

FIG. 3. Transient response of standard network.

3518 Rev. Sci. Instrum., Vol. 74, No. 7, July 2003 Kumar, Suresh, and Nagaraju

TABLE II. Measured capacitor value of network-2. cell temperature of 22⫾1 °C under dark condition. The mea-
Capacitance surements on GaAs/Ge solar cell were carried at 0.3–0.9 V
in steps of 0.1 and also at 0.87 V 共maximum power point
Cell Standard capacitor
voltage Time domain value
voltage兲 and 0.95 V. For the BSFR silicon solar cell the bias
共V兲 共␮F兲 共␮F兲 % error voltage was set from 0.2 to 0.6 V in steps of 0.1 V and at
0.45 and 0.55 V. At each voltage multiple measurements are
1 1.087 ⫺4.5
0.8 1.088 ⫺4.6 made and the average value of cell capacitance is estimated.
0.6 1.09 1.04 ⫺4.8 Figures 4共a兲– 4共d兲 shows the variation in cell current
0.4 1.09 ⫺4.8 with time measured on GaAs/Ge and BSFR silicon solar
0.2 1.20 ⫺15.3 cells at different cell voltages. The finite and measurable rise
tome of current indicates the existence of inductance in the
time at different bias voltages when the electronic switch is cell discharge path. However, this does not affect the mea-
closed is measured using the oscilloscope. The variation in sured capacitance. The area under the curve gives the total
current at 0.4 and 1 V for network-1 is shown in Figs. 3共a兲 charge 共Q兲 accumulated in the solar cell and dividing it with
and 3共b兲, where as for network-2 the variation in current at the corresponding cell bias voltage gives the cell capacitance
0.4 and 1 V is shown in Figs. 3共c兲 and 3共d兲. By numerical 共charge equivalent兲 关using Eq. 共3兲兴. This is compared with
integration the total charge accumulated is calculated. From the charge equivalent capacitance calculated from the imped-
this the capacitor value is estimated. In Tables I and II, the ance spectroscopy data,5 using Eq. 共4兲 at different cell volt-
measured values are compared with the actual capacitor val- ages, for the GaAs/Ge solar cell in Fig. 5 and for the BSFR
ues in network-1 and network-2, respectively. It is observed silicon solar cell in Fig. 6. The cell capacitance in the
that the error is more at low capacitance and at low bias GaAs/Ge solar cell is due mainly to transition capacitance,
voltage since the charge accumulated is very small. whereas in the BSFR silicon solar cell the cell capacitance
up to 0.4 V it is transition and beyond that the diffusion
IV. MEASUREMENTS ON SOLAR CELLS capacitance is dominant and it varies exponentially.1 Thus
The measurements were carried out on 20 mm⫻40 mm C q in BSFR silicon solar cell varies exponentially with cell
GaAs/Ge and 20 mm⫻40 mm BSFR silicon solar cells at voltage, whereas in the GaAs/Ge solar cell it varies inversely

FIG. 4. Transient response of solar cells current.

Rev. Sci. Instrum., Vol. 74, No. 7, July 2003 Time domain capacitance measurement 3519

FIG. 5. GaAs/Ge solar cell charge equivalent capacitance at different cell

voltages. FIG. 6. BSFR silicon solar cell charge equivalent capacitance at different
cell voltages.

nique can be implemented for on-line measurement of solar

with square root of voltage. It is observed that the deviation
cell capacitance in the manufacturing industry.
is within ⫾5% for GaAs/Ge solar cell, whereas for the BSFR
silicon solar cell it is less than ⫾2% around the maximum ACKNOWLEDGMENTS
power point. This is because in the BSFR silicon solar cell The authors are grateful to ISRO Satellite Centre and
around the maximum power point the cell capacitance is Indian Institute of Science, Bangalore for providing neces-
large when compared with the GaAs/Ge solar cell. Hence the sary support to carry out the above investigations.
error in measuring accumulated charge is minimum in the
1
BSFR silicon solar cell. R. Anil Kumar, M. S. Suresh, and J. Nagaraju, Sol. Energy Mater. Sol.
This time domain technique is quite simple for measur- Cells 60, 155 共2000兲.
2
J. Millman and C. C. Halkias, Integrated Electronics 共McGraw–Hill, New
ing the voltage dependent capacitances of any nonlinear de- York, 1972兲.
vices such as the solar cell. Here the postmeasurement data 3
R. Anil Kumar, M. S. Suresh, and J. Nagaraju, IEEE Trans. Electron
process is quite simple and easy to use when compared with Devices 48, 2177 共2001兲.
4
R. Anil Kumar, M. S. Suresh, and J. Nagaraju, Rev. Sci. Instrum. 72, 3422
the impedance spectroscopy technique.4 The measured cell
共2001兲.
capacitance can be used for the design of switching charge 5
R. Anil Kumar, M. S. thesis, Indian Institute of Science, Bangalore, India,
controller suitable for photovoltaic power systems. This tech- 2000.