Autonomous Temperature Sensor Based on a Photovoltaic Energy Harvesting System M. Ferri ∗ , D. Pinna ∗ , E. Dallago ∗ , P. Malcovati ∗ , G. Ricotti † ∗ Department of Electrical Engineering, University of Pavia, Pavia, Italy † STMicroelectronics, Cornaredo (Milano), Italy E-Mail: {massimo.ferri, daria.pinna, enrico.dallago, piero.malcovati}@unipv.it, [email protected]Abstract—In this paper we present an autonomous tempera- ture sensor supplied by a on chip photovoltaic energy harvester. Then system is realized in a BCD SOI technology. The energy harvesting elements consist of a 34 trench-insulated p- n junctions, while the sensing system consists of a bandgap reference circuit, including an integrated high precision temperature sensor, and a high voltage low drop-out voltage regulator (LDO). The entire system operates also at low illumination levels and tolerates a wide variation of the voltage produced by the micro-photovoltaic cell chain. I. Introduction Energy harvesting technologies and systems are emerging as the new challenge in the research and industrial field, growing at rapid pace. A wide range of applications can involve energy harvesting technologies, including distributed wireless sensor nodes [1]–[3] for structural health monitoring, embedded and implanted sensor nodes for medical applica- tions, battery recharging in large systems, monitoring environ- mental parameters, monitoring tire pressure in cars, powering unmanned vehicles, and running security systems in household conditions. Modern ultra-low-power integrated circuits [4]– [7] have reached such a level of integration and processing efficiency that the power consumption of the electronics in many applications is compatible with the amount of wasted energy [8]–[11] available in the environment. Breaking down the barriers of traditional sensors, wireless devices based on energy harvesting eliminate long cable runs as well as battery maintenance. Combining processors with sensors, the wireless nodes can record and transmit data, use energy in an intelligent manner, and automatically change their operating mode as the application may demand. Harvesting energy from the environment in the form of vibrations, strain, or light, these devices use background recharging of a battery or a super-capacitor to maintain an energy reserve. Recent applications include piezoelectric powered damage tracking nodes for helicopters as well as solar powered strain and seismic sensor networks for bridges. Photovoltaic phenomena [12], [13] allows us to retrieve the highest amount of energy with respect to any other type of harvesters, but when the source is the sun, power collecting becomes an intrinsically discontinuous process, forcing the adoption of storage elements in order to supply the system during the dark period. Moreover the light energy source usually features several noise components, such as the 50- 500 mV 1 V 17.5V Bandgap LDO Temperature Sensor 3.5 V 3.3 V V sensor Fig. 1. Block diagram of the proposed system Hz modulation of a light bulb, or, simply, the refraction and absorption by air molecules. In this paper, we present a photovoltaic energy harvesting power source, realized in a 0.35-μm BDC SOI technology [14], which supplies an autonomous temperature sensor. The system, whose block diagram shown in Fig. 1, consists of a series of 34 trench insulated p-n junctions, a bandgap reference circuit, including an integrated high precision temperature sen- sor, and a high voltage low drop-out voltage regulator (LDO). The regulator allows us to deliver a constant 3.3-V power supply voltage to a load also in low environment illumination conditions, as the large number of micro-photovoltaic cells in series ensures that, also considering a degradation of almost 80% of the voltage produced by each cell, the generated voltage is enough to properly operate the system. II. Integrated Micro Solar Cells In order to convert the incident light power into electrical power, we designed a chain of 34 micro-photovoltaic cells. Fig. 2 shows the structure of each photovoltaic cell, imple- mented in a p-well insulated from the common p-substrate by an oxide trench. The geometry of the n-diffusion realized in the p-well consists of a series of short-circuited rows, thus maximizing both the area and the perimeter of the diodes and creating several photovoltaic structures connected in parallel. The BCD SOI technology and the configuration realized allow us to create series structures and provide a voltage higher than 3.3 V, eliminating the parasitic diode between each single cell and the chip substrate. In particular, since the open circuit voltage V oc obtained for each illuminated cell is almost 530 mV, the entire chain can provide up to 17.5 V. This voltage value ensures that the system can operate, ideally, with an open circuit voltage as low as 20% of the nominal value. This voltage reduction can be caused by a condition of The 2010 IEEJ International Workshop on AVLSI, Pavia, Italy 275
4
Embed
Autonomous Temperature Sensor Based on a Photovoltaic ...ims.unipv.it/FIRB2006/pub/Ferri10c.pdf · Autonomous Temperature Sensor Based on a Photovoltaic Energy Harvesting System M.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Autonomous Temperature Sensor Based on a
Photovoltaic Energy Harvesting System
M. Ferri∗, D. Pinna∗, E. Dallago∗, P. Malcovati∗, G. Ricotti†∗Department of Electrical Engineering, University of Pavia, Pavia, Italy
Abstract—In this paper we present an autonomous tempera-ture sensor supplied by a on chip photovoltaic energy harvester.Then system is realized in a BCD SOI technology. The energyharvesting elements consist of a 34 trench-insulated p-n junctions,while the sensing system consists of a bandgap reference circuit,including an integrated high precision temperature sensor, anda high voltage low drop-out voltage regulator (LDO). The entiresystem operates also at low illumination levels and tolerates awide variation of the voltage produced by the micro-photovoltaiccell chain.
I. Introduction
Energy harvesting technologies and systems are emerging
as the new challenge in the research and industrial field,
growing at rapid pace. A wide range of applications can
involve energy harvesting technologies, including distributed
wireless sensor nodes [1]–[3] for structural health monitoring,
embedded and implanted sensor nodes for medical applica-
tions, battery recharging in large systems, monitoring environ-
mental parameters, monitoring tire pressure in cars, powering
unmanned vehicles, and running security systems in household
conditions. Modern ultra-low-power integrated circuits [4]–
[7] have reached such a level of integration and processing
efficiency that the power consumption of the electronics in
many applications is compatible with the amount of wasted
energy [8]–[11] available in the environment.
Breaking down the barriers of traditional sensors, wireless
devices based on energy harvesting eliminate long cable runs
as well as battery maintenance. Combining processors with
sensors, the wireless nodes can record and transmit data, use
energy in an intelligent manner, and automatically change their
operating mode as the application may demand. Harvesting
energy from the environment in the form of vibrations, strain,
or light, these devices use background recharging of a battery
or a super-capacitor to maintain an energy reserve. Recent
applications include piezoelectric powered damage tracking
nodes for helicopters as well as solar powered strain and
seismic sensor networks for bridges.
Photovoltaic phenomena [12], [13] allows us to retrieve the
highest amount of energy with respect to any other type of
harvesters, but when the source is the sun, power collecting
becomes an intrinsically discontinuous process, forcing the
adoption of storage elements in order to supply the system
during the dark period. Moreover the light energy source
usually features several noise components, such as the 50-
500 mV
1 V
17.5V
Bandgap LDO
TemperatureSensor
3.5 V
3.3 V
Vsensor
Fig. 1. Block diagram of the proposed system
Hz modulation of a light bulb, or, simply, the refraction and
absorption by air molecules.
In this paper, we present a photovoltaic energy harvesting
power source, realized in a 0.35-μm BDC SOI technology
[14], which supplies an autonomous temperature sensor. The
system, whose block diagram shown in Fig. 1, consists of a
series of 34 trench insulated p-n junctions, a bandgap reference
circuit, including an integrated high precision temperature sen-
sor, and a high voltage low drop-out voltage regulator (LDO).
The regulator allows us to deliver a constant 3.3-V power
supply voltage to a load also in low environment illumination
conditions, as the large number of micro-photovoltaic cells in
series ensures that, also considering a degradation of almost
80% of the voltage produced by each cell, the generated
voltage is enough to properly operate the system.
II. IntegratedMicro Solar Cells
In order to convert the incident light power into electrical
power, we designed a chain of 34 micro-photovoltaic cells.
Fig. 2 shows the structure of each photovoltaic cell, imple-
mented in a p-well insulated from the common p-substrate
by an oxide trench. The geometry of the n-diffusion realized
in the p-well consists of a series of short-circuited rows, thus
maximizing both the area and the perimeter of the diodes and
creating several photovoltaic structures connected in parallel.
The BCD SOI technology and the configuration realized allow
us to create series structures and provide a voltage higher
than 3.3 V, eliminating the parasitic diode between each
single cell and the chip substrate. In particular, since the open
circuit voltage Voc obtained for each illuminated cell is almost
530 mV, the entire chain can provide up to 17.5 V. This
voltage value ensures that the system can operate, ideally,
with an open circuit voltage as low as 20% of the nominal
value. This voltage reduction can be caused by a condition of
The 2010 IEEJ International Workshop on AVLSI, Pavia, Italy
275
N-Diffusion
P-Substrate
Trench
Single Solar Cell PhotovoltaicString
500 mV
1 V
17.5 V
Fig. 2. Block diagram of the integrated photovoltaic energy harvesting system
low illumination or by a large power request from the load.
Moreover, the series connection of photovoltaic cells allows
us to obtain directly all the reference voltages required for the
entire system. The geometrical dimensions of each cell are
385 μm × 245 μm. The width of the depletion region in the
p-n junction is given by
xdr =
√3εSiΦbi
q
(1
Na+
1
Nd
), (1)
where Na is the p-well doping concentration, Nd the n-
diffusion doping concentration, εSi is the dielectric permittivity
of silicon, q the charge of the electron and Φbi is the built-in
potential, given by
Φbi = VT ln
⎛⎜⎜⎜⎜⎝NaNd
n2i
⎞⎟⎟⎟⎟⎠ , (2)
where VT = kT/q is the thermal voltage. Substituting the
values of each variable in (1), xdr results equal to 3.2 μm.
In order to avoid overlapping between the depletion regions,
the width of the n-diffusion strips and the space among the
strips have been set to 5 μm.
In order to estimate the photogenerated current available
for the design of the system, we realized several micro-
photovoltaic cells in a 0.35-μm standard CMOS technology
on a test chip. The structure that we tested features as an
area of 0.5 mm × 0.5 mm. The power curve that we obtained
with 300 W/m2 of incident light power at 30 ◦C is shown
in Fig. 3 with a short-circuit current of 8.5 μA. Since the
doping concentrations can be assumed as the same in both
technologies, while the area of the BCD SOI photovoltaic
cell is almost 2.8 times smaller than the measured standard
CMOS photovoltaic cell, for the BCD SOI photovoltaic cell
we can estimate a short-circuit current (Isc) of about 2.5 μA.
Measurements on the realized BCD SOI photovoltaic cell are
in good agreement with this estimation. Indeed, Fig. 3 shows
also the power curve of the BCD SOI cell with 300 W/m2 of
incident light power and constant temperature of 30 ◦C.
0 100 200 300 400 500 6000
1
2
3
4
5
6
7
8
9
Photogenerated Voltage [mV]
Pho
toge
nera
ted
Cur
rent
[μA
]
Standard CMOSBCD SOI
Fig. 3. Power curve of photovoltaic cells realized in standard CMOS(0.25 mm2 of area) and BCD SOI (0.09 mm2 of area) technologies with300 W/m2 of incident light power
Vdd
VB,1
VB,2
M6
M4 M5
M7
M3M8
M1 M2
R1
R2
M9
Q1 (x16) Q2 (x2)
Vbandgap
Ban
dgap
Ref
eren
ce C
ircui
t
Ms
Tem
pera
ture
Sen
sor
Vsensor
Fig. 4. Schematic of the bandgap reference circuit
III. Bandgap Reference Circuit
The bandgap circuit is necessary to generate a temperature
independent voltage reference. The circuit operates on the
principle of compensating the negative temperature coefficient
of Vbe with the positive temperature coefficient of the ther-
mal voltage VT . The temperature coefficient of Vbe, at room
temperature, is −2.2 mV/◦C, while the positive coefficient
of the thermal voltage is 0.086 mV/◦C. Therefore a full
compensation, at room temperature, is obtained by combining
the two terms to achieve Vbandgap = Vbe +mVT , where m must
be equal to 25.6. If this condition is satisfied, the resulting
output voltage, approximately equal to 1.2 V, is at first order
temperature independent. Fig. 4 shows the schematic of the
bandgap reference circuit used in the proposed system. The
circuit is particularly critical, since it has to manage the vari-
ations of the supply voltage, due to illumination reduction or
output power changes, providing a constant reference voltage,
equal to 1.2026 V, to the LDO. The current that flows in
transistors M4 and M5 is mirrored in transistor M3, thus biasing
the two external branches with Id,M3= Id,M8
+ Ib,Q1+ Ib,Q2
.
Since Id,M4= Id,M5
, it results that Vgs,M6= Vgs,M7
. The bipolar
transistors, with emitter area ratio equal to 8, drain the same
The 2010 IEEJ International Workshop on AVLSI, Pavia, Italy
276
25 30 35 40
Temperature [°C]
45 50 55 601.2
1.22
1.24
1.26
1.28
1.3
1.32
1.34
Sen
sor
Out
put V
olta
ge [V
]
IdealExperimental
Fig. 5. Temperature sensor output voltage
current, leading to a ΔVbe = VT ln (8). The resulting current
flowing trough the bipolar transistors is
IR1=
VT ln (8)
R1
. (3)
At 27 ◦C IR1is equal to 1 μA. Since the same current is
mirrored in M3 and M8, the total power consumption of the
circuit is
Ptot =(Id,M3
+ Id,M4+ Id,M5
+ Id,M8+ Ic,Q1
+ Ic,Q2
)Vdd. (4)
The power supply voltage of the bandgap reference circuit
(Vdd) corresponds to the photogenerated voltage of the 7th
micro photovoltaic cell of the series chain (about 3.5 V). The
used bandgap reference circuit does not require any opera-
tional amplifier. The output voltage is fixed by the feedback
loop including transistor M8, which compensates any eventual
variation of Vbe,Q1,Q2. The cascade transistors M6 and M7 are
used to increase the gain of the loop. The voltage drop across
resistor Rs (Vsensor) is proportional to the absolute temperature
and it is used as temperature sensor in the proposed system.
Fig. 5 shows the value of Vsensor as a function of temperature.
The variation of Vsensor is closely related to the temperature
dependent current IR1, which is mirrored in the Rs branch. The
sensitivity achieved by the temperature sensor is 3.8 mV/◦C.
IV. LDO Circuit
The proposed LDO circuit is shown in Fig. 6. The circuit
consists of an error amplifier (M6, M7, M4, M5, M10), an
Bandgap voltage (Vbandgap) 1.199 V 1.202 VOutput voltage (VLDO) 3.357 V 3.369 VVdd,high 17 V 17.4 VVdd 3.5V 3.8 VTemperature sensor sensitivity 4.2 mV/◦C 3.8 mV/◦CMaximum load current (IL) — 500 nA
VI. Conclusions
In this paper we presented an autonomous temperature
sensor with a photovoltaic energy harvester and a voltage
regulator. The voltage regulator consists of a bandgap ref-
erence circuit and a high voltage LDO circuit. The realized
chip has been extensively simulated and measured, showing a
good agreement between simulated and experimental results.
Presently, we are designing an integrated solution in order to
obtain a completely autonomous wireless sensor node.
Acknowledgments
This work was supported by the Italian Ministry of Uni-
versity under FIRB project RBAP065425 “Analog and Mixed-
Mode Microelettronics for Advanced Systems”. The BCD SOI
technology has been provided by the R&D Department of
STMicroelectronics, Cornaredo (Milano), Italy..
References
[1] C. H. Zhiyong, L. Y. Pan, Z. Zeng, and M. Q. H. Meng, “A novel FPGA-based wireless vision sensor node,” in Proc. of IEEE InternationalConference on Automation and Logistics, Aug. 2009, pp. 841–846.
[2] K. Yamasaki and T. Ohtsuki, “Design of energy-efficient wireless sensornetworks with censoring, on-off, and censoring and on-off sensorsbased on mutual information,” in Proc. of IEEE Vehicular TechnologyConference, vol. 2, May 2005, pp. 1312–1316 Vol. 2.
[3] R. Zurawski, “Keynote: wireless sensor network in industrial automa-tion,” in Proc. of International Conference on Embedded Software andSystems, May 2009, p. 12.
[4] M. Koo, H. Jung, I. Song, H. S. Jhon, and H. Shin, “A 2.4 GHzCMOS ultra low power low noise amplifier design with 65 nm CMOStechnology,” in Proc. of International Conference on Solid-State andIntegrated-Circuit Technology, Oct. 2008, pp. 1488–1491.
[5] M. Alioto, “Design of nanometer MOS current mode logic: fromvery high-speed down to ultra-low power,” in Proc. of InternationalConference on Microelectronics, Dec. 2009, pp. 12–13.
[6] Y. Lee, M. Seok, S. Hanson, D. Blaauw, and D. Sylvester, “Standbypower reduction techniques for ultra-low power processors,” in Proc. ofEuropean Solid-State Circuits Conference, Sep. 2008, pp. 186–189.
[7] L. Turicchia, S. Mandal, M. Tavakoli, L. Fay, V. Misra, J. Bohorquez,W. Sanchez, and R. Sarpeshkar, “Ultra-low-power electronics for non-invasive medical monitoring,” in Proc. of IEEE Custom IntegratedCircuits Conference, Sep. 2009, pp. 85–92.
[8] R. Amirtharajah, J. Wenck, J. Collier, J. Siebert, and B. Zhou, “Circuitsfor energy harvesting sensor signal processing,” in Proc. of ACM/IEEEDesign Automation Conference, Jul. 2006, pp. 639–644.
[9] A. D. Joseph, “Energy harvesting projects,” IEEE Pervasive Comput.,vol. 4, no. 1, pp. 69–71, Jan. 2005.
[10] J. Elmes, V. Gaydarzhiev, A. Mensah, K. Rustom, J. Shen, andI. Batarseh, “Maximum energy harvesting control for oscillating energyharvesting systems,” in Proc. of IEEE Power Electronics SpecialistsConference, Jun. 2007, pp. 2792–2798.
[11] D. Niyato, E. Hossain, M. M. Rashid, and V. K. Bhargava, “Wirelesssensor networks with energy harvesting technologies: a game-theoreticapproach to optimal energy management,” IEEE Wireless Commun.Mag., vol. 14, no. 4, pp. 90–96, Aug. 2007.
[12] D. Brunelli, C. Moser, L. Thiele, and L. Benini, “Design of a solar-harvesting circuit for batteryless embedded systems,” IEEE Trans.Circuits Syst. I, vol. 56, no. 11, pp. 2519–2528, Nov. 2009.
[13] D. Dondi, A. Bertacchini, D. Brunelli, L. Larcher, and L. Benini,“Modeling and optimization of a solar energy harvester system for self-powered wireless sensor networks,” IEEE Trans. Ind. Electron., vol. 55,no. 7, pp. 2759–2766, Jul. 2008.
[14] M. Ferri, D. Pinna, P. Malcovati, E. Dallago, and G. Ricotti, “Integratedstabilized photovoltaic energy harvester,” in Proc. of IEEE InternationalConference on Electronics, Circuits and Systems, December 2009, pp.229–302.
The 2010 IEEJ International Workshop on AVLSI, Pavia, Italy