An UHF wireless power harvesting system--analysis and design.
Today a huge number of electromagnetic energy sources emit to the open space considerable amounts of energy. TV professional and amateur radio stations, 2G or 3G mobile phone base transceiver stations are everywhere. Much of the energy this stations emit is unused (mainly because they are broadcast stations), and hence arises the concept of Energy Harvesting / Energy Gathering. Recovering this energy, it is possible to make self-sustained different kinds of sensors , or power RFID circuits ,. The model and implementation of the model proposed in this paper allows harvesting energy from a large frequency spectrum, thus making possible the recovery of a superior quantity of energy, when comparing to a system that gathers energy from one specific source. This RF energy is harvested using a commercial log-periodic antenna and it is next converted to a DC voltage, using a voltage multiplier and a voltage regulator. OrCAD PSPICE simulations and experimental results are presented, to validate the proper functioning of the different components contained in the system.
As can be seen in Fig. 1, the system is composed by three main blocks. The main goal of the system is to transform the collected RF energy in a DC voltage that can be used in several applications.
[FIGURE 1 OMITTED]
The antenna used in these tests is a commercial log-periodic antenna. Since the system is supposed to work in a large frequency spectrum, a power matching network, if used, would have a very low quality factor. Because of this, it won't be used any power matching network. Another reason for not using a power matching network is the fact of having to use microstrips, since this is the correct way to construct the matching network at the working frequencies in this project. At the working frequencies the corresponding wavelength is around 0.6m (considering 500MHz). With this value of wavelength it is possible to build the system in such a way that the reflection of the waves is not a problem.
The voltage multiplier is the most important component of the harvesting system, and therefore the one that requires more careful implementation, being necessary to analyze and choose carefully each component of the multiplier, to ensure a proper functioning. After the voltage multiplier, a voltage regulator assures that the output voltage of the system is regulated to the required value. Although most of these RF to DC converters are built in CMOS, in this case it will be used, in a first phase, prototype development boards with copper tracks (SRBP strip board). This choice is due to the fact that these boards are much cheaper than using CMOS technology, and have the necessary characteristics to implement one initial prototype.
2 IMPLEMENTATION OF THE SYSTEM
The implementation of the energy harvesting system corresponds to the implementation and testing of the several components present in Fig. 1.
The antenna is responsible for collecting the RF energy from the surrounding environment. The antenna chosen is a commercial log-periodic antenna for TV applications (UHF). With this antenna, which is a large spectrum antenna, it is possible to gather energy from several different frequencies, as can be seen next. Using a narrow band antenna, it is possible to have a bigger gain (of the antenna), which makes the system work more easily. Nonetheless, the energy captured by the antenna would be smaller because of its narrow band. The main characteristics of the antenna can be seen in Table 1.
As can be seen in Table 1, the size of the antenna makes it suitable mainly for outdoor applications. In this case that's not a problem, because the project is supposed to be for outdoor sensors powering. This system also works for indoor applications, changing the used antenna. This makes this kind of projects very flexible because changing the application, the only thing that needs to be changed is the antenna. Also, even if the application requires a larger working frequency, an impedance matching network can be used, without changing the used voltage multiplier and voltage regulator. The experimental tests to obtain the results present in Fig. 2 were done by connecting the antenna directly to a GW Instek GSP827 spectrum analyzer. Both the antenna and the spectrum analyzer are adapted to 75 [ohm].
[FIGURE 2 OMITTED]
Although the antenna is commercialized as an UHF antenna, in Fig. 2 can be seen that it receives VHF signal, mainly from radio stations (~ 100 MHz). The power peak obtained in the 900 MHz ~ 1 GHz frequency range is a GSM signal. The system is supposed to gather power from all this frequencies. However, this paper focuses in scavenging energy from the TV UHF frequencies, showed next in Fig. 3.
[FIGURE 3 OMITTED]
These tests were made 7.54 km away from the nearest TV Transmitting Station, and without fully aligning the antenna. With a more careful alignment, or closer to an emitter station, the energy collected can reach higher values which facilitate the system operation. Nonetheless, the power contained in this frequency spectrum is 9 dBm, which corresponds to 8 mW. Considering 75 [ohm] load resistance (only the antenna resistance), this corresponds to around 750 mV. Since this voltage value is above the required value of at least 1.2 V, a voltage multiplier is needed, in order to raise this value.
2.2 Voltage Multiplier
In order to raise the voltage value obtained by the antenna to a desirable value (over 1.2V DC) it will be used a Cockcroft-Walton voltage multiplier (VM) . Two voltage multipliers were implemented, with different components. These VM's were simulated in OrCAD PSPICE using proper diode models, and implemented using the prototype development boards. The voltage multipliers use Schottky diodes, because these have a better performance when compared to p-n junction diodes . The results obtained are presented next, including a comparison between the two voltage multipliers.
Voltage multiplier A (VMA), in Fig. 4, with the electric circuit of Fig. 5, has 4 levels, which means it multiplies the input voltage by 8 (theoretically). The Schottky diodes used are 1N5819. The capacity of the capacitors used is 100 nF.
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
Figures 6a and 6b contain the results obtained in the several tests made to this voltage multiplier. It is made a comparison between the simulation and the practical results obtained. It is also indicated the ideal value of the voltage that should be achieved with the multiplier. These tests were made using a Wavetek function generator (Model 145) with a sinusoidal output, and connecting the voltage multiplier without any charge to a digital oscilloscope.
[FIGURE 6a OMITTED]
[FIGURE 6b OMITTED]
As shown by the results obtained, this voltage multiplier doesn't work in the desired frequencies. The maximum working frequency achieved is 5 MHz; this malfunctioning is due to the parasitic capacitances of the diodes that increase with the frequency. As the ohmic losses of the diodes are high the voltage multiplier efficiency drops when the voltage input value drops, and below 50mV the voltage multiplier doesn't work for any frequencies. The prototype development board used is designed for low frequencies too (less than 5 MHz) which increases the problems in the functioning of the system.
Since the first voltage multiplier implemented didn't work, a new voltage multiplier was simulated and implemented. The voltage multiplier B (VMB) was implemented using BAT62-03W Schottky diodes. The used capacitors in this case have an 8.2pF capacity and the multiplier also have 4 stages. The used capacitors and diodes are designed for PCB applications so their substrate losses are also low. Figures 7 and 8 show the implemented multiplier and the corresponding electric circuit, respectively. In this case it is used a parallel configuration, because this assembly guarantees a more stable voltage output value . This configuration used in this multiplier is usually called Dickson configuration.
[FIGURE 7 OMITTED]
[FIGURE 8 OMITTED]
For this voltage multiplier were made the same tests that were made for VMA. The results obtained are shown in Figs. 9a and 9b, for the same frequencies as before.
[FIGURE 9a OMITTED]
[FIGURE 9b OMITTED]
As the previous figures shows, this voltage multiplier works much better that the first one. This behavior is due to the fact that the BAT62-03W are diodes prepared to work at much higher frequencies that the 1N5819. Also the BAT62-03W is much smaller than the 1N5819, which reduces substrate losses. However even with this voltage multiplier there are very important losses that must be eliminated in order to maximize the system efficiency. These losses appear because the frequencies required are significant, and with the prototype development board used there are ohmic losses that can't be ignored. It is necessary to minimize the area used by the multiplier, in order to reduce these losses to an acceptable level. Table 2, shown below, contains a comparison between the two voltage multipliers.
The efficiency values calculated in the table above are referred to an ideal voltage multiplier. The values displayed in bold are the ones considered most relevant to this system, since the real input voltage is in the same order of magnitude.
2.3 Voltage Regulator
Since the output voltage of the multiplier depends of the load current and the RF input at each moment, a regulator is necessary, to assure that the device powered by this system has the correct level of voltage to have a proper functioning. The voltage regulator cannot be linear, since linear voltage regulators use Zener diodes which are dissipative. In energy harvesting systems, dissipative losses are not allowed, since the values of harvested power are low and these losses would be very significant, so the voltage regulator must be a switching voltage regulator. Since there are many commercial solutions of voltage regulators created specifically to energy harvesting applications, there is no need to create a new regulator.
3 SYSTEM TESTING
In order to validate the proper functioning of the system, it was connected to a HLPM-Q156 LED (1,6 [V.sub.on] @ 0,5 mA). Only the voltage multiplier B was tested, because has seen in Table 1, the voltage multiplier A won't work at the required frequencies. Connecting the system to the receiving antenna, the LED stayed off. Despite this malfunctioning, if the system is connected to an oscilloscope in order to view the output signal, it is possible to see a voltage value of around 6V. This means that the system has an output voltage value acceptable, but there are losses that need to be eliminated to assure a proper functioning.
4 CONCLUSIONS AND FUTURE WORK
With the results obtained in the voltage multiplier tests, contained in Table 1 and the tests made to the whole system itself, it can be concluded that it is possible to put this system into operation. The voltage multiplier A cannot be used, but that was expectable, since its components are not designed for high frequencies. Voltage multiplier B works at the desired frequencies, but still has some losses that are very significant for the levels of power gathered. The area in which the voltage multiplier is implemented must be reduced, to assure lower losses, and possibly using a CMOS version of the multiplier is a final solution. Although the antenna was not fully aligned with the TV Emitter Station, the level of power collected is sufficient to make the system feasible. Since the power available to harvest is around 8mW, it is easily seen that even with low efficiency the system can work, and harvest a significant quantity of power.
This paper and the research associated were partially funded by:
[1.] Sample. A, Smith. JR, "Experimental results with two wireless power transfer systems", in proc. of Radio and Wireless Symposium, San Diego, CA, 2009:18-20.
[2.] Bergeret. E, Pannier. P, Gaubert. J, "Optimization of UHF voltage multiplier circuit for RFID application", in proc. of the 17th International Conference on Microelectronics, Marseille, France, 2005:6.
[3.] Olgun. U, Chen. CC, Volakis. JL, "Wireless power harvesting with planar rectennas for 2.45 Ghz RFIDs", in proc. of International Symposium on Electromagnetic Theory (EMTS), URSI, Berlin, Germany, 2010:329-331.
[4.] Lamantia. A, Maranesi. P, Radrizzani. L, "Small-signal model of the Cockcroft-Walton voltage multiplier", IEEE Transactions on Power Electronics 1994:18-25.
[5.] Wang. J, Dong. L, Fu. Y, "Modeling of UHF voltage multiplier for radio-triggered wakeup Circuits", Wiley International Journal of Circuit Theory and Applications 2010, DOI 10.1002/cta.692.
[6.] Kim. S, Cho. J, Hong. AS, "A full wave voltage multiplier for RFID transponders", IEICE Transactions 2008:388-391.
Nuno Amaro, Stanimir Valtcheva
Departamento Engenharia Electrotecnica, FCT-UNL, Quinta da Torre, 2829-516 Caparica, Portugal
Table 1. Characteristics of the used antenna Frequency 470 ~ 862 MHz Gain 11 ~ 12 dB Longitudinal Size 1,269 m Number of Elements 47 Impedance 75 [ohm] Table 2. Efficiency of the two implemented voltage multipliers Efficiency (%) Amplitude Frequency A B (mV) (MHz) 1 85 70 2 73 74 5000 5 34 79 10 13 81 20 - 43 1 60 66 2 40 68 500 50 13 69 10 - 65 20 - 64 1 28 31 2 15 27 100 5 3 33 10 - 28 20 - 21 1 14 18 2 4 20 50 5 - 19 10 - 15 20 - 14 1 4 20 2 - 24 25 5 - 25 10 - 23 20 - 19
|Printer friendly Cite/link Email Feedback|
|Author:||Amaro, Nuno; Valtchev, Stanimir|
|Publication:||International Journal of Emerging Sciences|
|Date:||Dec 1, 2011|
|Previous Article:||System for acquisition and analysis of low frequency magnetic fields in three axes.|
|Next Article:||A cloud-based DSS model for driver safety and monitoring on Australian roads.|