Dynamic Noise Reduction in the System Measuring Efficiency of Light Emitting Diodes
Algorithm and method of signal waveform reconstruction in the problem of light emitting diodes efficiency measuring system was developed. Introduced method uses blind sequential signal extraction technique based on generalized skewness and generalized kur
- PDF / 686,019 Bytes
- 8 Pages / 439.37 x 666.142 pts Page_size
- 44 Downloads / 164 Views
Abstract. Algorithm and method of signal waveform reconstruction in the problem of light emitting diodes efficiency measuring system was developed. Introduced method uses blind sequential signal extraction technique based on generalized skewness and generalized kurtosis allows to solve problem of noise reduction. Keywords: Light emitting diodes Generalized kurtosis
Neural network Generalized skewness
1 Introduction The problem of a noise suppressing in the channels of information - measuring systems (IMS) cannot be solved always via classical methods of linear low pass filtering or linear adaptive filtering. An example of such system is IMS for measuring parameters of light emitting diodes (LED), which comprises several channels to measure the efficiency of the LEDs and photodiodes. Measurement is performed by comparing two cooling curves, characterizing crystal LED, obtained after exposure of short current pulses of positive and negative polarity Iforv ; Irev as shown in Fig. 1. Problem of measurement LED efficiency is especially critical for LEDs used in medicine instruments [1, 2]. The scientific publications [3, 4] consider metrological problems of LEDs. This research is devoted to the problem of measurement of efficiency of LEDs. Loss associated with heating of crystal dominates among all reasons of the losses of LED capacity. LED is affected by forward pulses of direct current I þ , backward pulses of direct current I and small forward bias current to turn-on the LED. The first part U þ ðt) of curve characterizes cooling process after effecting current I þ . At this part crystal radiates light and heats simultaneously. The second part U ðtÞ characterizes cooling process after effecting current I . At this part crystal heats only. Difference between power on the first and the second part of curve allows to calculate power, expended to radiation. The normalized difference of power characterizes LED‘s efficiency. Choosing the amplitude of the reverse pulse Urev subject to a constant amplitude of direct pulse Uforv we compare two cooling curves U þ ðtÞ and U ðtÞ. When two LED
© Springer International Publishing Switzerland 2016 L. Cheng et al. (Eds.): ISNN 2016, LNCS 9719, pp. 302–309, 2016. DOI: 10.1007/978-3-319-40663-3_35
Dynamic Noise Reduction in the System Measuring Efficiency
303
0.3
Uforv
0.2
cooling the crystal
0.1 0
t1
t2
Voltage (V)
t0
t4
t3
U+(t)
t5
U-(t)
heating the crystal
Urev
-110
0
2
4
6
8
10
12
14
16
18
Time
Fig. 1. Curve of heating and cooling the LED crystal, shown without following the scale. Dotted line shows variation of Urev during adjustment of two cooling curves.
cooling curves differ less than a small value e the compensation is obtained. Condition of compensation is of the form kU þ ðtÞ U ðtÞk\e. When compensation is reached the value of LED efficiency g may be calculated by the following formula: Rt2 g ¼ t1
Iforv Uforv dt
Rt4
Irev Urev dt
t3
Rt2
ð1Þ
Iforv Uforv dt
t1
where Iforv ; Irev - forward and reverse current of the diode. Condition
Data Loading...
