SSC99-XII-7
Department of Electrical Engineering
Iran University of Science & Technology
Narmak, Tehran, 16645, Iran.
Advance Electronic Research Center
P. O. Box: 19585 - 836
Tehran, Iran.
E-mail: Dehbonei@mavara.com
: Satellite, Power System, Maximum, Tracker.
It is shown that at maximum power, the Photovoltaic (PV) voltage varies nonlinearily with temperature and isolation level, but is directly proportional to the PV cell open circuit voltage. The proportionality voltage-factor is fixed for a given PV generator regardless of temperature, isolation and panel configuration, but depends on cell material and manufacturing. This remarkable property is used to achieve temperature and insolation independent maximum power point tracking of satellite’s solar cells with a simple and reliable technique. The open circuit voltage is continuously measured by a microcontroller and is used to estimate the maximum power operating point of the system.
The Voltage-based Maximum Power Point Tracker (VMPPT) is demonstrated by construction and testing of a solar battery charger (using silicon solar cells, Ni-Cd batteries and a buck mode VMPPT), a solar water pump
(using silicon solar cells, a PM DC Motor and a boost mode VMPPT) and resistive loads supplied by solar cells.
Measured results are satisfactory and confirm the proposed technique. The advantage of this method as compared to the current-based MPPT are “simplicity” and “higher efficiency”.
The satellite Electric Power Subsystem (EPS) has the responsibility of providing continuous, regulated and conditioned power to payloads and other subsystems during all mission phases. Undesired space and mission features such as eclipse and shadowing interruptions, temperature and solar angle variations, and load fluctuations as well as solar array degradation (over time) must be considered by EPS designers. In today’s small spacecraft, these problems are exacerbated due to extreme mass, volume and cost constrains. Accurate sun- tracking solar arrays, generous cooling devices and high capacity batteries are denied the designer and therefore, new generation EPS are usually running with a negative power budget and every drop of available electric energy is captured and used as effectively as possible. These constrains along with the required high reliability performance and the fact that spacecraft is not mostly within the communication range, have made EPS designing exciting and challenging.
There are two general approaches for satellite power system design [1-5]. In Direct Energy Transfer
(DET) method, the payload, subsystems and batteries are directly connected to the solar arrays and the extra power (at low temperatures and/or Beginning
Of Life (BOL)) are absorbed by shunt regulators
(Fig.1(a)). This is a dissipative and simple, but weighty and low efficient power system. A more
1
Dr. M.A.S. Masoum 13th Annual AIAA/USU Conference on Small Satellites

SSC99-XII-7
(a)
(b)
Figure 1. Block diagram of satellite Electric Power Subsystem (EPS); (a ) Power Point Tracking (PPT) method., (b) Direct Energy Transfer
(DET) method common technique for designing the EPS, especially for small satellites, is the Power Point Tracking
(PPT) method (Fig.1(b)): the amount of generated solar energy is continuously controlled and matched with the satellite’s energy demand. The system is naturally running at a low temperature level and there will be no need for large dissipative devices.
But the price is paid by the addition of a PPT device.
The primary source of power on most present satellites result from photovoltaic solar cells mounted on surfaces and illuminated by the sun’s rays. In order to increase the available electrical energy and limit the required solar arrays, the
″ maximum power point
″
of PV cells must be actively tracked and used by the EPS system. In the literature, there are different proposed methods for Maximum Power
Point Tracking (MPPT) of solar cells [6-14] and some have been applied in actual satellite power systems [15]. The current-based MPPT of [9] uses reference (dummy) cells for the measurement and estimation of
″ maximum power point
″
of the solar panel. In this paper, a Voltage-based MPPT is proposed for the optimal tracking of satellite solar cells. The main advantage of this method (which uses a microcontroller circuit for the online measurement of “cell open circuit voltage”) as compare to the current-based MPPT method is: the elimination of the reference cells which results in a
“simpler” and “more efficient” system. The effect of
VMPPT is especially noticed at high power demand phases and during the EOL where the cell degradation is appreciable.
The proposed method is demonstrated by construction and testing of three laboratory PV systems: a solar battery charger (consisting of four
Ni-Cd batteries with a buck type VMPPT circuit), a solar water pump (consisting of a permanent- magnet
DC motor water pump and a boost type VMPPT circuit) and resistive loads supplied by solar cells .
Using the equivalent circuit of photovoltaic cells
(Fig.2), the expression for the nonlinear V-I characteristics of M parallel strings (N series cells per string) is
2
Dr. M.A.S. Masoum 13th Annual AIAA/USU Conference on Small Satellites

SSC99-XII-7
Ρ
V
=
Ν
In



I sc
− i pv
+
M I
0
M I
0


 −
N
M
R i pv
(1) where Io is the reverse saturation current, I sc
is the cell short circuit current, resistance and
R is the series cell s
λ
is a constant coefficient proportional to cell material.
computed V-I characteristics (based on equation 2) are plotted in Fig.3 for different insolation levels.
The laboratory measurement of V-I and P-I characteristics for one OFFC panel are shown in Fig.
4. Comparison of these two figures indicates a good agreement between the computed and measured characteristics.
Figure 2. Photovoltaic cell equivalent circuit.
For the silicon solar panel (M=1, N=36) manufactured by the Iranian optical Fiber
Fabrication Co. (OFFC), the V-I characteristics of
Eq.1 can be written as
υ pv
=
1 767 ln



I sc
.
i .
00005
00005


 − i pv
(2)
Cell specification are given in Table 1 and the
There is also a nonlinear relationship between the V-
I characteristics and operating temperature
[13,14,16]. According to reference [16], this dependency can be modeled as: i
∆
T
=
T cell
−
T ref
∆
I
= α
.
∆
T
∆ υ = − β
.
∆
T
−
R s
.
∆
I
υ pv
= υ pv
− ∆ υ pv
= i pv
− ∆
I
(3)
Taking into account the impact of temperature variations (Eq.3), the new forms of Eq.2 can be written as :
pv
=
1 69 ln

.
− i pv
+
.
00024
00024
 i pv
(4a)
pv
=
182 ln



−
.
i pv
+
.
00001
00001


 − i pv
(4b)
Equations (4a) and (4b) are evaluated for T=70
° c and
T=-20
° c, respectively, and are plotted in Fig.5 for one OFFC panel.
Figure 3. Computed nonlinear V-I characteristics of one OFFC silicon solar panel for different isolation levels (Eq.2)
Figure 4. Measured and computed (Eq.2) nonlinear V-I and P-I characteristics of one OFFC silicon solar panel (table 1)
Figure 5. Temperature impact (Eqs.3,4) on V-I characteristics of
OFFC silicon solar cells (table 1)
3
Dr. M.A.S. Masoum 13th Annual AIAA/USU Conference on Small Satellites

SSC99-XII-7
In order to determine the operating points corresponding to maximum power for different isolation levels, the partial derivative of power is computed as follows d di pv
P
= d di pv



1 767 ln

I sc i
.
.
00005
00005
 i pv  i pv


=
0
(5)
The solution of Eq.5 represents the currents corresponding to maximum power as a function of short circuit currents of PV generator,
I mp
=
( )
. This is the main idea for the current-based MPPT technique [9]. For the OFFC silicon cells, this function is computed as
−
35340 I mp
 3 ( I sc
−
I
.
ln
(
× 3 mp
)
+
1 
( I sc
−
I mp
)
)
I mp
=
0
(6)
Evaluating Eq.1 at maximum power (e.g., and open circuit condition (e.g., i pv
I sc
= I mp
=
0 ) gives:
)
V mp
= f
1
( I sc
) and V oc
= f
2
( I sc
) , respectively.
Therefore, at a given temperature and isolation level, the “voltage corresponding to maximum power” could be expressed as a function of “cell open circuit voltage”; namely
V mp
=
( oc
)
(7)
This is the base for the VMPPT technique. Applying numerical methods (e.g., MATHCAD Software),
Eq.7 is shown by + signs in Fig.6 for the OFFC silicon solar cells.
Using curve fitting techniques, we can numerically find this relationship as:
V mp
= f nonlinear
( V oc
)
≅ f liner
( V oc
)
=
V oc
−
0 328
≅
.
V
O C
= m v
V oc
(8) this equation is compared with the actual characteristics in Fig.6.
Therefore, the voltages corresponding to maximum powers are directly proportional to open circuit voltages at different isolation levels and temperatures. The proportionality voltage-factor
( m v
) is fixed for a given panel regardless of cell configuration, isolation and temperature variations,
Table 1. Specifications of the silicon solar cells used
(manufactured by OFFC)
Current Temperature
α =
0 002086
Coefficient
.
Voltage Temperature
β =
0 0779
Coefficient
[Amp/
° c]
[Volt/
° c]
Reverse Saturation Current
Short Circuit Cell Current
Cell Resistance
I
0
=
.
−
4
[A]
Iph
=
2 926 [A]
R
S
=
0 0277 [
Ω
]
Cell Material Coefficient λ =
0 049
[
V
−
1
]
+
Vmp
=
( )
-
Vmp
= ⋅
Voc
+
Figure 6. Voltages corresponding to maximum power versus open circuit cell voltages (Eq.7) and the linear function (Eq.8) approximation (T=25 and varying isolation level).
but depends on cell materials and manufacturing techniques. According to Eq.8 and Fig.6, for the
OFFC silicon cells m v
=
. , which will also be confirmed by experimental result in the following section.
For the tracking and estimation of the maximum power point of solar panels, the open circuit voltage must be accurately measured under all operating conditions (temperature, isolation and degradation levels). To do this, a microsystem is used with the following responsibilities:
1.
Continuous control and switching of the DC/DC converter (buck or boost mode).
2.
Continuous measurement of the panel open circuit voltage.
3.
Continuous estimation of panel maximum power operating point (Eq.8).
4.
Continuous matching of satellite operating point with the estimated solar panel maximum power point (by changing the duty cycle of DC/DC converter).
5.
If required, continuous matching of generation and demand levels by adjusting the system operating point on the V-I characteristics.
The microsystem driver circuitry consists of: a
80C51 microcontroller, D/A and A/D converters, the series-switch driver (boost), power supplies, display and keyboard as shown in Fig.7.
4
Dr. M.A.S. Masoum 13th Annual AIAA/USU Conference on Small Satellites

SSC99-XII-7
OUT 8
OUT 7
OUT 6
OUT 5
OUT 4
OUT 3
OUT 2
OUT 1
LATCH
19
16
15
12
9
6
5
2
Q7
Q6
Q5
Q4
Q3
Q2
Q1
Q0
74373
LE
OE
D7
D6
D5
D4
D3
D2
D1
D0
11
1
18
17
14
13
8
7
4
3
OUT 16
OUT 15
OUT 14
OUT 13
OUT 12
OUT 11
OUT 10
OUT 9
FROM OUT 2
FROM OUT 1
220
LATCH
LE
OE
19
16
15
12
9
6
5
2
Q7
Q6
Q5
Q4
Q3
Q2
Q1
Q0
74373
470
D7
D6
D5
D4
D3
D2
D1
D0
+
RED
11
1
18
17
14
13
8
7
4
3
GND
BUZZER
LED
19
16
15
12
9
6
5
2
Q3
Q2
Q1
Q0
Q7
Q6
Q5
Q4
74373
LATCH
D3
D2
D1
D0
D7
D6
D5
D4
LE
OE
GND
11
1
18
17
14
13
8
7
4
3
LATCH
19
16
15
12
9
6
5
2
Q7
Q6
Q5
Q4
Q3
Q2
Q1
Q0
74373
LE
OE
D7
D6
D5
D4
D3
D2
D1
D0
11
1
18
17
14
13
8
7
4
3
TO BUFFER OP-AMP 1
VCC
2
4.7K
4.7K
4.7K
4.7K
4.7K
GND
VCC
GND
4
2
Iout
Iout
15
14
Vrf(-)
1
16
Vrf(+)
Vlc
33NF
COMP
33NF msbB1
B2
B3
B4
B5
B6
B7 lsbB8
GND
8
9
10
11
12
5
6
7
DAC0800
33NF
GND
33NF
10
11
8
9
12
5
6
7
GND msbB1
B2
B3
B4
B5
B6
B7 lsbB8
DAC0800
GND
33NF
TO BUFFER OP-AMP 2
VCC
POSITINE VOLTAGE ( +8 V )
4.7K
4.7K
Iout
Iout
Vrf(-)
Vrf(+)
Vlc
COMP
4
2
15
14
1
16
4.7K
4.7K
4.7K
33NF
GND
GND
VCC
NEGATIVE SUPPLY ( -8 V )
LCD (2 LINE & 16 COLUMN))
LCD
VCC
A
K
GND
VCC
SWITCH
2 ohm (0 .5 watt)
( FOR BACK-LIGHT)
TO SUPPLY VOLTAGE (VCC)
+ 100UF
2
LM7805CK
+5V Vin
1
GND
100
7.5 V
570
+8 VOLT FOR CIRCUIT
2
+ 100UF
+Vout
LM317T
GND
Vin
3
D
D
GND
JP?
1
3
2
4
HEADER 2X2
JP?
1
3
2
4
HEADER 2X2
GND
FROM BATTERY(+)
FROM BATTERY(GND)
FROM PANNEL (+)
FROM PANNEL (GND)
GND
1 2
13
12
7
8
5
6
3
4
1
2
MICRO CONTROLLER
P10
P11
P12
P13
P14
P15
P16
P17
P00
P01
P02
P03
P04
P05
P06
P07
INT1
INT0
15
14
31
19
18
T1
T0
EA/VP
X1
X2
P20
P21
P22
P23
P24
P25
P26
P27
9
RESET
17
16
RD
WR
8051AH
RXD
TXD
ALE/P
PSEN
10
11
30
29
24
25
26
27
28
21
22
23
35
34
33
32
39
38
37
36
30p 12Meg
GND
30p
100
10K
GND
VCC
+ 10U
8
74LS04
74LS04
16 15
VCC
10K
74LS04
74LS04
12
A
12
7427
A
12
7427
A
12
7427
A
7427
13
2
1
13
2
1
13
2
1
13
2
1
1
2
13
GND
A
7427
7404
7404
7404
7404
8
13
14
17
18
3
4
7
LATCH
D4
D5
D6
D7
D0
D1
D2
D3
Q4
Q5
Q6
Q7
Q0
Q1
Q2
Q3
1
11
OE
LE
74LS373
9
12
15
16
19
2
5
6
12
GND
VCC
4
5
6
1
2
3
DECODER (3-8)
A
B
C
E1
E2
E3
74138
Y0
Y1
Y2
Y3
Y4
Y5
Y6
Y7
9
7
11
10
15
14
13
12
(FOR RAM)
(FOR LCD)
(FOR A/D
(UNDEFINED)
(FOR LATCH EN)
FOR KEY_RD
(UNDEFINED)
(UNDEFINED)
TO INTRRUPT ( INT 1 )
12
7427
A
13
2
1
GND
A0
A1
A2
A3
A0
VCC
11
12
13
14
15
16
17
18
U?
Q8
Q7
Q6
Q5
Q4
Q3
Q2
Q1
74541
E2
E1
2
7404
A
1 12
7427
A
13
2
1
D8
D7
D6
D5
D4
D3
D2
D1
22
27
26
20
10
9
8
5
4
3
7
6
25
24
21
23
2
RAM (8K)
A7
A8
A9
A10
A11
A12
A0
A1
A2
A3
A4
A5
A6
OE
WE
CS2
CS1
6264
19
1
9
8
7
6
5
4
3
2
A2
A1
A0
4
3
6
5
10
9
8
7
25
24
21
23
2
EPROM (8K)
A7
A8
A9
A10
A11
A12
A0
A1
A2
A3
A4
A5
A6
D0
D1
D2
D3
D4
D5
D6
D7
20
22
27
1
CE
OE
PGM
VPP
2764
VCC
16
17
18
19
11
12
13
15
D0
D1
D2
D3
D4
D5
D6
D7
15
16
17
18
19
11
12
13
TO PIN 11 FROM 4024
VCC
17
14
15
8
18
19
20
21
23
24
25
10
6
9
ANALOG TO DIGITAL 1
CLOCK
START
ENABLE ref(+) ref(-)
22
ALE
IN-7
ADD-C
ADD-B
ADD-A
IN-6
7
EOC
IN-5
IN-4 lsb2-8
2-7
2-6
2-5
2-4
2-3
2-2 msb2-1
ADC0809
IN-3
IN-2
IN-1
IN-0
12
16
1
28
27
26
3
2
5
4
GND
VCC
1
2
13
7427
A
12
GND
FROM KEYBOARD ( C1 )
FROM KEYBOARD ( C2 )
FROM KEYBOARD ( C3 )
GND
10 K
10 K
100
1
2
3
4
7
8
5
6
SW?
SW-DIP8
16
15
14
13
12
11
10
9
CONNECTOR
6
7
4
5
8
1
2
3
8 HEADER
1
4
7
*
2
5
8
0
3
6
9
#
FROM MCU ( ALE )
GND
4024
GND
4024
FROM MCU ( KEY_RD )
RD (FROM MCU)
FROM A/D ( EOC )
FROM C1 (KEYBOARD)
FROM C2 (KEYBOARD)
FROM C3 (KEYBOARD)
FROM R1 (KEYBOARD)
FROM R2 (KEYBOARD)
FROM R3 (KEYBOARD)
FROM R4 (KEYBOARD)
FROM MCU ( OUT 3 )
FROM POSITIVE SUPPLY ( +8 V )
TO A/D CLOCK
1
2
13
A
7410
VCC
10K
6
12
10K
VCC
3
2
LM741
100n 10K
GND
10K
GND
10K
2.6mH
BC107
10
10 W
1N4007
470 UF
+
IRF540
10k
470
R
HIGH VOLTAGE (20V-24V)
1
A
7404
2
10K
VCC GND
POSITIVE SUPPLY ( +8 V )
FROM DAC 1 (PIN 2)
5
4
3
2
7
6
9
8
19
1
E2
E1
D4
D3
D2
D1
D8
D7
D6
D5
74LS541
Q4
Q3
Q2
Q1
Q8
Q7
Q6
Q5
11
12
13
14
15
16
17
18
POSITIVE SUPPLY ( +8 V )
TO DATA BUS (D0_D7)
NEGATIVE SUPPLY ( -8 V )
+8 VOLT SUPPLY
10K
3
2
OP-AMP 1
6
741
N2 / N1 =1.5
1K 1N4007
10K
TRANS1
GND
IRF540
BC107
1N4007
10K
10K
GND
3
2
OP-AMP 3
6
741
OUT DAC 1
+
47 UF
3
LM337
Vin
570
-Vout
2
100
+
10 U
FROM DAC 2 (PIN 2 )
PWM OUTPUT 1 ( - 7 V TO + 7 V P-P)
NEGATIVE SUPPLY ( -8 V )
1
A
7404
2
PWM OUTPUT 2 ( TTL OUT )
3
2
OP-AMP 2
6
741
8 10K
VCC
1
U?
MC1403
IN OUT
2 3
2
OP-AMP 1
6
741
2.5 VOLT (FOR REFERENCE)
10K 10k
2.2 NF
220K
GND
GND
NEGATIVE SUPPLY ( -8 V )
POSITIVE SUPPLY ( +8 V )
8038
2.2 K
POT1 10 K
47 K
NEGATIVE SUPPLY ( -8 V )
Figure 7. The microsystem driver circuitry used in Figs.8 and 12 for measurements.
5
Dr. M.A.S. Masoum 13th Annual AIAA/USU Conference on Small Satellites

SSC99-XII-7
In order to experimentally investigate the performance of the proposed VMPPT tracker, the solar battery charger of Fig.8 was constructed using one OFFC silicon solar module (table 1), four Ni-Cd batteries (1.2V, 800mAh), a VMMPT buck type tracker (MOSFET power switches, ferrite core inductance) and a microcontroller (Fig.7). The measured voltage, current and power waveforms at the output of the solar panel as well as the input of the nickel-cadmium batteries are shown in figure 9, with and without the VMPPT unit. These results indicate an increase of about 150% in panel output power in the presence of VMPPT unit during the charging phases. The corresponding measured V-I and P-I characteristic for the operating condition
(insolation and temperature levels) are shown in
Fig.11 which confirms a maximum power of 25 watts. The two zero-power instances on panel waveform, indicate disconnection of panel by the microsystem and online measurement of its open circuit voltage
Experiments and measurements at different isolation, temperature and load levels using various voltage-factors (Fig.11), confirmed that maximum output power is obtained for a unique value of
S.A
TO MCU ( ADC 1 )
VB (24 V)
FROM MCU ( OUT 7 )
10 K
10K
IRF540
10 V
BC107
D3
2200 UF +
GND
TO MCU ( ADC 2 )
VB (24 V)
FROM MCU (TTL PWM 1)
10K
10K
BC107
GND
BC107
FROM MCU (OUT 8)
10K
0.15 OHM
10V
D4
IRF540
D5 6 A
10 MH
GND
GND GND
TO MCU ( ADC 3 )
TO LOAD( + )
+
2200 UF
0.15 OHM
FROM LOAD( -- )
GND
Figure 8. The constructed microcontroller-based solar battery charger (VMPPT in buck mode) used for measurements
Fig 9.
The waveforms of current, voltage and power at the output of the solar panel and the input of the batteries with (top figures) and without (bottom figures) the VMPPT unit. Horizontal axis indicates sampling times.
6
Dr. M.A.S. Masoum 13th Annual AIAA/USU Conference on Small Satellites

SSC99-XII-7
Figure 10. The measured V-I and P-I characteristics corresponding to
Fig.9.
m v
=
0 7
Figure 11. Comparison of measured panel output power (of solar battery charger) for different values of voltage-factors. The operating conditions (insolation level) of Fig.10 & 11 are different.
The solar water pump of Fig.12 was constructed using one OFFC silicon solar module (Table 1), a small permanent magnet DC motor (24 V, 45W), a
VMPPT boost type tracker and a microcontroller
(Fig.7). Different measured results for various values
S.A
GND
TO MCU ( ADC 1 )
6A
D2
VB (24 V)
FROM MCU ( OUT 7 )
10 K
10K
0.15 OHM
TO MCU ( ADC 2 )
IRF540
10 V
D3
GND
IRF540
2200 UF +
U?
VB (24 V)
2.2 K
L?
D?
DOUBLE STATE SW
BC107 BC107
IRF540
D?
XX
470K
+
2200 UF
D?
TO MCU ( ADC 3 )
TO LOAD ( + )
TO LOAD ( - )
0.15 OHM
GND GND GND GND
FROM MCU (TTL PWM 1)
FROM MCU (OUT 8)
10K
10K of voltage-factor (Fig.13) confirm the unique value of m v
=
.
for maximum power operation of the system under different conditions.
Figure 13. Comparison of measured panel output power (of solar water pump) for different values of voltage-factors. The operating conditions (insolation level) of Fig.13 and 14 are different.
7
Dr. M.A.S. Masoum 13th Annual AIAA/USU Conference on Small Satellites

SSC99-XII-7
Fig 15.
The waveforms of current, voltage and power at the output of the solar panel and the input of solar pump with (top figures) and without (bottom figures) the VMPPT unit.
Figures 14 and 15 demonstrate the V-I and P-I characteristics as well as voltage, current, and power waveforms for a chosen operating condition. The introduction of VMPPT changes the panel output power from 22 watts to 34 watts (55% increase).
different isolation, temperature and load conditions.
As an example, figure 16 shows the measured panel average output power with and without the VMPPT tracker for a chosen operating condition. As for the other types of loads (Figs.11 and 13), the computed value of voltage-factor ( m v
=
. ) results in
maximum output power.
Further experimental investigations of VMPPT technique was done by supplying a set of resistors with one OFFC silicon solar panel (Table 1).
Different measurements were performed to investigate the uniqueness of voltage-factor for
Figures 17-19 show the V-I and P-I characteristics and the corresponding waveforms for two resistive loads. According to Fig.17, the maximum output solar power (for the chosen operating condition) is
32W. Furthermore, the panel voltage corresponding to maximum power ( Vmp
=
18 V ) is about 70% of
Figure 16. Comparison of measured panel output (of resistive loads) for different values of voltage-factor.
Figure 14. The measured V-I and P-I characteristic corresponding to
Fig.15.
8
Dr. M.A.S. Masoum 13th Annual AIAA/USU Conference on Small Satellites

the panel open circuit voltage ( V oc
=
12 6 V ) which confirms the unique computed value of the voltagefactor ( m v
=
. ) for these loads.
The impact of VMPPT is 125% and 28% increase in panel output power for the heavy (Fig.18) and light
(Fig.19) resistive loads , respectively.
SSC99-XII-7
9
Dr. M.A.S. Masoum 13th Annual AIAA/USU Conference on Small Satellites

SSC99-XII-7
Fig 18.
The waveforms of current, voltage and power at the output of the solar panel and the input of the resistive load with (top figures) and without (bottom figures) the VMPPT unit (heavy loads).
Fig 19. The waveforms of current, voltage and power at the output of the solar panel and the input of the resistive load with (top figures) and without (bottom figures) the VMPPT unit (light loads).
Figure 17. The measured V-I and P-I characteristics corresponding to
Figs.18 and 19.
10
Dr. M.A.S. Masoum 13th Annual AIAA/USU Conference on Small Satellites

SSC99-XII-7
A Voltage-base Maximum Power Point Tracker
(VMPPT) is proposed for optimal Power Point
Tracking (PPT) of Satellite Electric Power
Subsystem (EPS) and is demonstrated by construction and testing of a solar battery charger, a solar water pump and resistive loads. It is shown
(numerically and experimentally) that at maximum power, the cell voltage is linearly proportional to the cell open circuit voltage.
A microsystem is used for the on-line measurement of the solar cells open circuit voltage, estimation of panel maximum power point, control of system operating point and matching it with panel maximum power point on the V-I characteristic curves. The proportionality factor is fixed (e.g., m
V
=
.
for OFFC silicon cells) regardless of operating conditions. The following conclusions can be stated:
1.
Introduction of VMPPT increases the output power of PV generators. The percentage of power increase depends on operating conditions
(isolation, temperature, degradation and load levels).
2.
The proposed VMPPT technique is very simple since it requires only the measured open circuit voltage.
3.
The cell open circuit voltage can be measured using a microsystem.
4.
The same microsystem can be used for switching of DC/DC converters (buck or boost mode) and controlling the operating point on the panel V-I characteristics.
5.
The effect of VMPPT is especially appreciable when generated electrical energy is less than the demand level (e.g., at EOL, high power demand phases, high temperature conditions).
6.
The main advantage of the proposed microprocessor-based VMPPT method as compared to the current-based MPPT [9] is the elimination of the reference cells which results in a “simpler” and “more efficient” system.
[1] Chetty, P.R.K., “Satellite Technology and its
Applications”, TAB BOOKS Inc., 1988.
[2] Sullivan, R.M., “The Small Astronomy Satellite
(SAS) Power system”, 4th Annual AIAA/USU
Conference on Small Satellite, 1990.
[3] Wong, H.S., Blewett, M.J. “The UoSAT-2
Spacecraft Power system”, Journal of the
Institution of Electronic and Radio Engineering,
Vol. 57, No.5, Sep./Oct. 1987.
[4] Sakoda, D., Nobel, M., Paluszek, S. “Preliminary design of PANSAT Electric power and
Communication Subsystems”, 4th Annual
AIAA/USU Conference on Small Satellite, 1990.
[5] Dunbar, R., Hardman, G. “A Flexible
Autonomous Power Management System for
Small Spacecraft”, 4th Annual AIAA/USU
Conference on Small Satellite, 1990.
[6] Enslin, J.H.R., Snyman, D.B., ”Combined Low-
Cost, High-Efficient Inverter, Peak Power
Tracker and Regulator for PV Application”,
IEEE, T-PE, VOL.6, NO.1, January 1991.
[7] Saied, M.M., Hanafy, A.A., EL_Gabaly, M.A.,
Safa, Y.A., Jaboori, M.G., Yamin, KH.A.,
Sharaf, A.M., ”Optimal Design Parameter for a
PV Array Coupled to a DC Motor via a DC-DC
Transformer”, IEEE, 91WM 146-1EC.
[8] Jaboori, M.G., Saied, M.M., Hanafy, A.A.R., ”A
Contribution to the Simulation and Design
Optimization of Photovoltaic Systems”, IEEE,
91WM 149-5EC.
[9] Alghuwainem, S.M., ”Matching of DC Motor to a Photovoltaic Generator Using a STEP-UP
Converter with a Current-Locked Loop”, IEEE,
T-EC, VOL.9, NO.1, March 1994.
[10] Hiyama, T., Kouzoma, Sh., Imakubo, T.,
”Identification of Optimal Operating Point of
PV Modules Using Neural Network for Real
Time Maximum Power Tracking Control”,
IEEE, T-EC, VOL.10, NO.2, June 1995.
[11] Hiyama, T., Kouzoma, Sh., Imakubo, T.,
”Evaluation of Neural Network Based Real
Time Maximum Power Tracking Controller for
PV System”, IEEE, T-EC, VOL.10, NO.3
September 1995.
[12] Altas, I.H., Sharaf, A.M., ”A Novel On-line
MPP Search Algorithm for PV Arrays”, IEEE,
T-EC, VOL.11, NO.4, December 1996.
[13] Dehbonei, H., "Solar Water Pump, Using a
Maximum Power Point Tracker (MPPT) for
Optimal Driving a DC Motor Pump": IUST,
Electrical Engineering Department, M.S Thesis,
September 1997, (Persian).
[14] Masoum, M.A.S., Dehbonei, H., “Optimal
Power Point Tracking of Photovoltaic Systems under all Operating Conditions”, I7th Congress of the World Energy Council, Houston, USA,
Sep 12-18, 1998.
[15] Beukes, H.J., Enslin, J.H.R., ”Analysis of a New
Compound Converter and Bus Regulator for
Satellite Power Systems”, IEEE, PESC, 1993.
11
Dr. M.A.S. Masoum 13th Annual AIAA/USU Conference on Small Satellites

[16] Salameh, Z.M., Borowy, B.S., Amin, A.R.A.,
”Photovoltaic Modules-Site Matching Based on the Capacity Factors”, IEEE, T-EC, VOL.10,
NO.2, June 1995.
SSC99-XII-7
12
Dr. M.A.S. Masoum 13th Annual AIAA/USU Conference on Small Satellites