Modelling of Hybrid Energy Harvester with DC-DC Boost Converter

0 downloads 0 Views 2MB Size Report
Boost Converter using Arbitary Input Sources for ... the integration with the Boost Converter. An input ... the gist of Ultra Low Power (ULP) energy harvesting. With.
IEEE-ICSE2014 Proc. 2014, Kuala Lumpur, Malaysia

Modelling of Hybrid Energy Harvester with DC-DC Boost Converter using Arbitary Input Sources for Ultra-Low-Power Micro-devices Michelle S.M.Lim1,3, Sawal H.M.Ali2, S. Jahariah1 and MD.Shabiul Islam1

Michelle S.M.Lim1,3 3

Faculty of Applied Sciences and Computing (FASC), Tunku Abdul Rahman University College (TAR UC) Jalan Genting Kelang, Setapak, 53300 KL, Malaysia Email: [email protected]

1

Institute of Microengineering and Nanoelectonics (IMEN) 2 Department of Electrical, Electronics and Systems, Faculty of Engineering and Built Environment Universiti Kebagnsaan Malaysia (UKM), 43600 Bangi, Selangor, Malaysia Email: [email protected]/[email protected]

the solution. As such, for reliability purposes, multi-input ambient sources have been proposed by several literatures [1][4] to ensure continuous flow of energy when either one or two sources are absent. Past researchers for multiple inputs use duty cycling, pulse-width, switching frequency and time multiplexing [1], [4], selecting the input with maximum power [5] or parallel combinations of harvesters [2]. This paper focuses on the parallel approach alike [2] to increase the combined current, albeit for three ubiquitous ambient sources from piezoelectric (PZT), Photovoltaic (PV) and Thermoelectric (TEG). These three commercially available sources will be arbitrarily modeled both individually and combined, in which Table I distinguishes between the three different harvesters in terms of material, impedance for Optimum Power Transfer and Maximum Output Power. These sources, once combined, will then be fed into a DC-DC Boost converter with considerations of suitable parametric analysis such as parasitic inductive resistance, capacitive resistances, rise and fall times, load resistance, duty cycle and switching frequencies in accordance to the three combined ambient sources.

Abstract— This work involves the modeling of three arbitrary input sources representing Hybrid Energy Harvesters (HEH) using a DC-DC Boost converter. These sources are combined in parallel and targeted at scavenging passive human power, therefore the three suitable ambient sources are motion, thermal and indoor light. Multiple sources mitigate limitations caused by single source harvesters but suffer impedance mismatches which greatly limit the total combined power that could have been harvested. A Boost Converter with suitable parameters has been designed and integrated to the HEH, and PSPICE software has been used for both the modeling of arbitrary sources as well as the integration with the Boost Converter. An input source as low as 18 mV to 907 mV was able to be boosted into a 310 mV-27.9 V output when suitable parametric values were selected for the Ultra Low Power (ULP) HEH. A duty ratio of 0.5, with 10 kΩ load, 22 µH inductor as well as a switching frequency of 25 kHz was selected to be slightly above the audio range as well as being high enough to reduce passive component sizes. While VO/ VS of the boost converter is linear, PO/PIN is a function of third order polynomial. Therefore, at the HEH’s lowest combined configuration of 1 K temperature difference, 0.25 g of vibration and 100 lux of indoor lighting, a combined 14 µW can be harvested. At its maximum of 10 K heat difference, 1 g vibration and 1000 lux of indoor lighting a combined 187 µW can be harvested. At its minimum, this enables possibility of battery-less applications in powering a quartz watch at 5 µW while at its maximum capacity powering a pace maker of ~50 µW as well as micro devices of ~100 µW solely from passive human activity. Once a 33 mF input capacitor is placed between the sources and converter, an output power of between 9.61 µW- 78 mW can be obtained. Keywords—Modelling Arbitary Sources; Harvester(HEH); Ultra Low Power (ULP)

Hybrid

TABLE I.

Parameters Material Size Operating Range Open Circuit Voltage, VOC Optimum Impedance, ZOP Operating Voltage, VMPP Maximum Power,PMPP Maximum Power Extraction

Energy

I. INTRODUCTION Autonomous systems such as wearable devices, biomedical implants and wireless sensor nodes are essentially the gist of Ultra Low Power (ULP) energy harvesting. With considerable motivations in ULP consumer products and the need for lifetime lasting power supply, energy harvesting is

978-1-4799-5760-6/14/$31.00 ©2014 IEEE

COMPARISONS OF ENERGY HARVESTER CHARACTERISTICS

a.

28

Energy Harvester Typesa PV [AM1417]

PZT [V21BL]

TEG [CP60133]

Amorphous 4.865 cm2 50 - 1000 lux

Cantilever 13.09 cm2 0.25 – 1 g

Peltier Module 2.25 cm2 ∆T = 1 - 10 K

1.177 – 2.8 V

1.98 - 6.05 V

53 - 530 mV

31 k – 293 kΩ

64.941 kΩ

0.5 Ω

0.709 - 2 V

1–3V

26.5 – 265 mV

1.71 126.8 µW Requires Tracking

15.01 140.876 µW PMPP at Voc/ 2

1.368 139.8 mW PMPP at Voc/ 2

Note that lower values are obtained during circuit operation due to PMOS diode drop and impedance mismatches.

IEEE-ICSE2014 Proc. 2014, Kuala Lumpur, Malaysia

II.

Firstly, the power that can be obtained from the PV cell is PPV = VPV IPV. A plot of P-V, I-V and P-R characteristic between 50 lux to 1000 lux of fluorescence illumination based on modeling of (1) gives a maximum power, PMPP between 2.4-128.7 µW at an operating voltage of 1.31 - 2.24 V as summarized in Table I. Secondly, a temperature difference of between 1-10 K is simulated leading to 26 - 200 mV of operating voltage and a possibility of 1.368 mW -139.8 mW if impedance were matched at a constant 0.5 Ω as shown in Fig. 1 and summarized in Table I. For the cantilever type MIDE PZT of V21BL, fP = 110 Hz at 0g tip mass and only 1 of the 2 PZT connected, the input amplitude range of 0.25 g to 1 g provides a VMPP of 1-3 V and PMPP of between 15.01 µW 140.876 µW at an optimum impedance of 64.941 kΩ as summarized in Table I and verified by the PSPICE simulation in Fig.1.

Fig. 1. Equivalent electrical model of energy harvesting sources with corresponding P-V curve using PSPICE.

The corresponding P-V characteristics that has been simulated is shown in Fig. 1. This generally shows that PV harvester requires tracking to obtain the Maximum Power Point (MPP) due to its constantly varying internal impedance in regard to varying Illumination levels. Previous Maximum Power Point Tracking (MPPT) approaches include the fractional open circuit and fractional short circuit [6], [14] or more precise techniques that require extra computation such as perturb and observe (P&O) [1], Incremental Conductance Method [14] as well as particle swarm method [17] and fuzzy algorithm [18] to track its MPP. The TEG and PZT harvesters however have its MPP at half of its open circuit voltage with a fixed internal impedance, thus normally only requires one time setting of impedance [1] and do not require any complex tracking algorithm.

All three transducers in this literature has been modeled in PSPICE according to the datasheet parameters [9]-[11] as shown in Fig.1 where the single diode model has been used for the modeling of an amorphous PV cell, similar to [1],[2],[6],[9]. The equation representing the solar model is as given by (1) referring to the PV circuit model in Fig. 1 with a Voltage Controlled Current Source (VCCS) indicated as Gvalue and derivation using the Kirchhoff Current Law and Shockley Equation of ideal diode [2], [6], we have I PV  I L  I O [exp(

VPV  I PV RS V  I PV RS )  1]  ( PV ) nVT RSH 

where VT = kTC / q is the thermal voltage with k = 1.381 × 1023, q = 1.602 × 10-19 C and TC = operating Temperature in Kelvin. The other five parameters in (1) are IL ≈ ISC = light current/ short circuit current, IO = diode reverse saturation current, RS = series resistance, RSH = shunt resistance and n = diode ideality factor. The TEG module however is based on the equation as shown in (2) [2],

VTEG (OC )  S * T  nT *  (TH  TC )

Previously, power conditioning architectures were built to cater for specific harvesters but today the challenge lies in the interactions between these sources which causes impedance mismatches as the TEG harvester has generally very low impedances in the range of Ω while PV and PZT harvesters has impedances in the range of kΩ as can be observed from datasheet values [9]-[11] and the P-R curves of all three harvesters where its optimum impedance for different input source levels, RMPP has been tabulated in Table I. Therefore, it has been stated that no common MPPT technique exists between these three harvesters [12], however, past literatures has already proposed time multiplexing [1] or pulse counting method [4] to adaptively adjust harvester impedance individually before entering any DC-DC converter topology.



which is made up of nT thermocouples connected thermally in parallel but electrically in series [1], [2],[7],[10], where  and S are the Seebeck’s Coefficient of a single thermocouple and TEG respectively [2] in which S is modeled as 52.8 mV/ K based on [10]. Finally, the PZT harvester together with Mide datasheet values is modeled based on the Thevenin Equivalent circuit of a sinusoid with full bridge rectification as suggested in [1], [8]. A PZT represented normally by a sinusoid current source, given as

I PZT  I P sin( 2f P )

MODELLING OF HYBRID ENERGY HARVESTERS

Here, these three harvesters will be connected in parallel, as has been proposed previously to combine two sources by [2], [12]. Each of these sources will have a resized PMOS transistor configured as diode to avoid reverse current flow thus contributing to certain amount of losses. The combination of these three sources has been tested for all ranges both individually and combined to obtain optimum impedances in preparation for Boost Converter’s parametric analysis in PSPICE. As in [2], the combined HEH voltage, VHEH = VPV + VDPV = VTEG + VDTEG = VPZT + VDPZT, while the accumulated

(3)

at a resonant frequency of fP shunt with internal Capacitance, CP. and full bridge rectification with ideal diodes has been modeled here as a voltage source of VPZT(OC) = IPZT/ 2ᴨfPCP with an effective series resistor, RPZT = 1/4*CP *fP just as indicated in [1],[8] and as shown in Fig.1.

29

IEEE-ICSE2014 Proc. 2014, Kuala Lumpur, Malaysia TABLE III.

EFFECTS OF SOURCE VOLTAGE VARIATIONS

Simulation Condition: D= 0.5, tr=1 ns, R IND=0.1 Ω, RC = 0.01 Ω, fSW = 25 kHz, RLOAD=10 kΩ, W/L = 1:1000, T=27 °C Active Energy Optimum Boost Output Values Vs Harvesters Sources Vo,mpp Po,mpp (Levels) PV (100 lux) 829 mV 25.530 V 65.180 mW TEG (dT = 1 K) 18 mV 310.003 mV 9.610 µW PZT (0.25 g) 907 mV 27.945 V 78.092 mW PV + TEG 829 mV 25.547 V 65.266 mW PZT + TEG 907 mV 27.945 V 78.092 mW PV + PZT 876 mV 26.926 V 72.501 mW PV + PZT + TEG 876 mV 26.921 V 72.474 mW

III. HEH INTEGRATION WITH DC-DC BOOST CONVERTER Energy harvesters from ambient environment need to be conditioned to an appropriate voltage level either for a fixed load requirements or variable load situation such as battery charging [15] or transmission in a WSN. Therefore, most literature will use either one [4] or multiple DC-DC converters [1], [2], [5] units for these input sources to obtain desired output levels. Here, a single conventional DC-DC Boost converter is considered for the three parallel sources in this literature. Several parameters are analyzed to obtain an optimum output value from a variety of input voltages scavenged from the ambient environment. The parameters to be considered for this Boost converter are inductive resistance, capacitive resistance, rise and fall times, switching frequency, W/L ratios, switch voltage drops and load changes. Based on Table II, the worst case scenario is when only TEG is present with a mere 18 mV with a maximum power transfer when an optimum impedance of 284 kΩ is present. We will proceed to use this voltage condition for the following analysis and finally integrating the sources to the optimized Boost converter is shown in Fig. 2. Several parametric effects contributing to the converter’s energy losses will be investigated in future.

Fig. 2. Equivalent circuit model of boost converter with HEH sources configured in parallel

current, IHEH = IPV + IPZT + ITEG. The power, PHEH is given by (3),

PHEH  PPV  PTEG  PPZT V  I PV RS  VHEH * [I SC  I O exp( PV )  1] nVT VTEG(OC)  VHEH VPZT (OC)  VHEH  [VHEH * ( )]  [VHEH * ( )]) RTEG RPZT



with all its parameters defined previously. Fig. 2, on the other hand, shows the combined electrical model of the proposed HEH in PSPICE with its corresponding P-V, P-R and transient analysis values in Table II when individual or all three sources are present at 100 lux of illumination, 0.25 g amplitude and 1K thermal difference. It should be noted that the low voltage levels and impedance of TEG limits its domination over the other two sources. The RMPP values are obtained during a DC sweep of parametric R values so that optimum impedances are used during transient analysis for Maximum Power Transfer (MPT). It can be observed that due to the additional PMOS diode and parallel combinations for accumulated current values from all sources, the RMPP values will no longer correspond to the originally simulated values as shown in Table I previously. Finally, these combined harvesters will be fed into the input of the Boost converter to be explained in the next section. TABLE II.

Table III summarizes the optimum values of the Boost converter when different configurations of harvesting sources are available at T = 27 °C. Source Voltages between 18 mV to 907 mV can be boosted to 310 mV to 27.9 V with power output ranging from 9.6 µW to 78 mW when an input capacitor of 33 mF and a 22 µH inductor is used with 25 kHz of switching frequency while Fig. 3 shows the boosted voltage and output power curve when all input sources are active at various room temperature conditions, with 26.921 V and 72.474 mW at 27 °C as tabulated in Table III.

OPTIMUM POWER, VOLTAGE AND IMPEDANCE VALUES FOR VARIOUS HARVESTERS COMBINATIONS

Active Energy Harvesters (Levels) PV (100 lux) TEG (dT = 1K) PZT (0.25g) PV + TEG PZT + TEG PV + PZT PV + PZT + TEG

Optimum Operation Values @ 27°C PMPP

RMPP

VMPP

3.707 µW 1.0899 nW 10.407 µW 3.707 µW 10.407 µW 14.085 µW 14.085 µW

185 kΩ 284 kΩ 79 kΩ 185 kΩ 79 kΩ 54 kΩ 54 kΩ

829 mV 18 mV 907 mV 829 mV 907 mV 876 mV 876 mV

Fig. 3. Ouput power (Left) and boosted voltage curves (Right) of HEH during dT=1 K + 100 lux + 0.25 g vibration at 22 °C, 27 °C and 32 °C

30

IEEE-ICSE2014 Proc. 2014, Kuala Lumpur, Malaysia

voltage values from 18 mV-907 mV to output voltage of 0.3 – 27.9 V. A total output power of between 9.61 µW-78 mW can be achieved when a 33 mF input capacitor is inserted before the converter. Also, a plot of VO/VS shows output voltage as a linear function of input source voltages while output power is a function of third order polynomial. With this amount power, a possibility if powering Quartz watches, cardiac pace makers, WSN and hearing aids are possible. REFERENCES [1]

[2]

[3]

[4]

[5] Fig. 4. Plot of VO/VS (left) and PO/PIN (Right)

In its entirety, in order to integrate these three sources optimally, the combined and chosen parameters for an optimum boost converter design for the combined harvesters with an input of 18 mV to 907 mV requires a sampling frequency of 25 kHz, D = 0.5, ESR = 0.01 Ω, tR = 1 ns, RL = 0.1 Ω, PMOS W/L = 1000:1 when CIN = 33 mF, CLOAD = 0.2 µF and L = 22 µH in order to obtain an output voltage of ~20 times the source voltages and an output power range of 9.61 µW to 78 mW. All of which are simulated at a room temperature of T = 27 °C. A plot of VO/VS is represented by a linear regression trend line while a plot of the integrated PO/PIN for this boost converter is represented by a third order trend line with corresponding equations shown in Fig. 4.

[6] [7]

[8]

[9] [10] [11] [12] [13]

IV. CONCLUSION Arbitrary source modeling of PV, PZT and TEG harvesters which is commercially available has been investigated and its generic characteristics achieved from as low as ∆T = 1 K from the TEG to a combined ∆T = 1 K, 100 lux of illumination and 0.25 g of vibration amplitude. It has been verified via PSPICE simulation, mathematical model and datasheet parameters that all harvesters bear wide range of impedance from low Ω values of TEG to kΩ values of the PV and PZT harvesters. The mismatching of these impedances causes loss in overall maximum harvested power, therefore it is pivotal to match these harvesters impedance to its load, in this case a Boost converter. These harvesters are fed into a boost converter to investigate optimum boost converter design for these three harvesters which includes a switching frequency of 25 kHz, duty cycle of, D = 0.5, load of 10 kΩ, inductive resistance of 0.1 Ω, rise times of 1ns, load capacitance of 0.2 µF, L=22 µH, CIN = 33 mF and a ripple below 2.09 % . This boost topology with three harvesting sources achieves almost 20 times input

[14]

[15]

[16]

[17]

[18]

31

S. Bandyopadhyay, and A. P. Chandrakasan, “Platform architecture for solar, thermal, and vibration energy combining with MPPT and single inductor,” Solid-State Circuits, IEEE Journal of, vol. 47, no. 9, pp. 2199-2215, 2012. Y. K. Tan, and S. K. Panda, “Energy harvesting from hybrid indoor ambient light and thermal energy sources for enhanced performance of wireless sensor nodes,” Industrial Electronics, IEEE Transactions on, vol. 58, no. 9, pp. 4424-4435, 2011. S. Mi, M. Lim, M. Ali, S. Hamid, and M. Shabiul Islam, “A Novel Architecture of maximum power point tracking for ultra-low-power based hybrid energy harvester in ubiquitous devices: a review,” American Journal of Applied Sciences, vol. 10, no. 10, 2013. C. Shi, B. Miller, K. Mayaram, and T. Fiez, “A multiple-input boost converter for low-power energy harvesting,” Circuits and Systems II: Express Briefs, IEEE Transactions on, vol. 58, no.12, pp.827-831, 2011. H. Lhermet, C. Condemine, M. Plissonnier, R. Salot, P. Audebert, and M. Rosset, “Efficient power management circuit: From thermal energy harvesting to above-IC microbattery energy storage,” IEEE Journal of Solid-State Circuits, vol. 43, no. 1, pp. 246-255, 2008. L. Castaner, and S. Silvestre, Modelling photovoltaic systems using PSpice: John Wiley and Sons, 2002. Y. K. Ramadass, and A. P. Chandrakasan, "A batteryless thermoelectric energy-harvesting interface circuit with 35mV startup voltage." IEEE Journal of Solid-State Circuits, vol. 46, no. 1, pp. 333-341, 2011. Y. K. Ramadass, and A. P. Chandrakasan, “An efficient piezoelectric energy harvesting interface circuit using a bias-flip rectifier and shared inductor,” Solid-State Circuits, IEEE Journal of, vol.45, no.1, pp.189204, 2010. “SANYO Amorphous Solar Cell datasheet: Amorton AM-1417”. 2008. “CUI Inc. CP60 series Peltier Module”. 2012. “MIDE piezoelectric Energy Harvesters”. 2013. Y. K. Tan, Energy Harvesting Autonomous Sensor Systems: Design, Analysis, and Practical Implementation: CRC Press, 2013. N. A. C. Mustapha, A. Alam, S. Khan, and A. W. Azman, "Parametric analysis for designing low voltage and low frequency energy harvester booster." 2013 IEEE Regional Symposium on Micro and Nanoelectronics (RSM). pp. 122-125. Sept, 2013. N.Femia, G. Petrone, G.Spagnuolo, M.Vitelli, Power Electronics and Control Techniques for Maximum Energy Harvesting in Photovoltaic Systems. CRC Press, 2013. F. Ongaro, S. Saggini, and P. Mattavelli, “Li-ion battery-supercapacitor hybrid storage system for a long lifetime, photovoltaic-based wireless sensor network,” Power Electronics, IEEE Transactions on, vol. 27, no. 9, pp. 3944-3952, 2012. A. Richelli, S. Comensoli, and Z. M. Kovacs-Vajna, “A DC/DC boosting technique and power management for ultralow-voltage energy harvesting applications,” Industrial Electronics, IEEE Transactions on, vol. 59, no. 6, pp. 2701-2708, 2012. K.Lian, J.Jhang, and I.Tian, “A Maximum Power Point Tracking Method Based on Perturb-and-Observe Combined With Particle Swarm Optimization,” IEEE Journal of Photovoltaics,vol.4,no.2, pp.626-633, 2014. B. N. Alajmi, K. H. Ahmed, S. J. Finney, and B. W. Williams, “Fuzzylogic-control approach of a modified hill-climbing method for maximum power point in microgrid standalone photovoltaic system,” Power Electronics, IEEE Transactions on, vol. 26, no. 4, pp. 1022-1030, 2011.