Online proceedings - EDA Publishing Association

24-26 September 2008, Rome, Italybatteries load when they became fully discharged. Theresults were evaluated with a dedicated software, that wasalso developed by the authors.The measurements showed that at lower temperatures theLi-Po batteries always operate on lower operation voltage,and the discharging time period becomes shorter. Afterevaluation, the model creation process has been presented.The presented models are ready to be used in sensor networkoptimization applications.VIII.SUMMARYFigure 1. Diagram shows an 1350mAh Li-Po battery’s measured curve(400mA, 0°C) and it’s fitted function which it’s own parametersParameterValueA 219.834B 11896.5C 6.40×10 -9D –1.00161×10 -4E 4.1582Table 1. Parameters of the resultant function (time in seconds,voltage in Volts) presented in Fig. 3.Having fitted all the measured data to the curve, theobtained parameters are collected to use them to create thetemperature dependent, constant current Li-Po batterymodel. Each parameter (A, B, C, D and E) has a differentvalue, depending on the curve calculated by the fittingalgorithm. The parameters then can be considered asfunctions of the input parameters of the model: for exampletemperature (T) and load (L), like this: A(T, L). For eachparameter, these functions for temperature (and load) can becalculated. We have the values for discrete points (where theoriginal measurements were performed), and then usingthese points further approaching functions can be calculatedfor the parameters A, B, C, D and E. The moremeasurements are done and evaluated, the more points weget for the functions, so the fitting functions will be moreand more accurate for the model.After these calculations, A,B,C,D and E functions arefilled with real values, and replacing them into the originalapproach function (1), the model is ready to use.Energy is getting more important nowadays. Mobildevices usually use batteries, but batteries can not considerto a single voltage supply anymore: this device has specialeffects on how the energy is drawn out of and also shows agreat influence on the ambient temperature. A measurementsetup with a proper measurement PCB has been created,which allowed us to perform long-time dischargingmeasurement on batteries placed in regulated temperaturechamber. Results were evaluated, and a mathematical modelwas created, which is ready to use for simulations andprediction algorithm developments for devices supplied fromLi-Po batteries.ACKNOWLEDGMENTThis work was supported by the PATENT IST-2002-507255 Project of the EU and by the OTKA-TS049893 andthe NKFP NAP 736-205/2005 BELAMI projects of theHungarian Government.REFERENCES[1] Kanishka Lahiri, Anand Raghunathan, Sujit Dey: Battery-DrivenSystem Design: A New Frontier in Low Power Design, DesignAutomation Conference, 2002.[2] L. Benini, G. Castelli, A. Macii: “A Discret-Time Battery Model forHigh-Level Power Estimation”, IEEE DATE 2000.[3] L. Benini, G. Castelli, A. Macii: “Extending Lifetime of PortableSystems by Battery Scheduling”, IEEE DATE 2001.[4] D. Szente-Varga, Gy. Horvath, M. Rencz: “Creating temperaturedependent Ni-MH battery models for low power mobile devices”,THERMINIC 2006, Nice, France[5] D. Szente-Varga, Gy. Horvath, M. Rencz: Ni-MH battery modelingfor ambient intelligence applications, Proceedings of DTIP 2007, pp.332-337., Stresa, Lago Maggiore, Italy, 25th-27th Apr 2007VII.CREATING MODEL FROM THE DATAMeasurements with Li-Po rechargeable batteries havebeen completed at different temperatures to examine theirdischarge process. The temperature was stabilized with anair basin type cooling-heating thermostat (thermo chamber),and the measurements were controlled by a specialmeasurement board, which attended to switch off the©EDA Publishing/THERMINIC 2008 131ISBN: 978-2-35500-008-9