Entity

Time filter

Source Type


Vukic V.,Nikola Tesla Institute of Electrical Engineering | Osmokrovic P.V.,University of Belgrade
Nuclear Technology and Radiation Protection | Year: 2010

Low-drop out volt age regulators with various geometries and technological realisations of serial pnp power transistors were exposed to ionizing radiation. Although de vices with vertical emitters were considered much less susceptible to the influence of radiation on for ward emitter current gain than circuits with round emitters, the experiment showed a similar degradation of current gain in both cases. The main reason of high radiation susceptibility of the examined vertical serial pnp transistor is the implementation of an interdigitated emitter, with high perimeter-to-area ratio, causing the great in crease of serial transistor's base cur rent, but a mi nor influence on the maximum out put current. Transistors with round emitters with small perimeter-to-area ratio expressed a moderate current gain degradation, but a rapid fall of the emitter injection efficiency, causing a significant de crease of the maximum out put current. Regard less of the similar for ward emitter current gain degradation, reliability and operational character is tics of two types of low-drop out volt age regulators were completely different. Source


Jankovic M.V.,Nikola Tesla Institute of Electrical Engineering | Rubens N.,University of Tokyo
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2012

Recently, elements of probabilistic model that are suitable for modeling of learning algorithms in biologically plausible artificial neural networks framework, have been introduced. Model was based on two of the main concepts in quantum physics - a density matrix and the Born rule. In this paper we will show that proposed probabilistic interpretation is suitable for modeling of on-line learning algorithms for Independent Component Analysis (ICA), which could be realized on parallel hardware based on very simple computational units. Proposed concept (model) can be used in the context of improving algorithm convergence speed, learning factor choice, input signal scale robustness, and can be easily deployed on parallel hardware. © 2012 Springer-Verlag. Source


Jaksic O.,Nikola Tesla Institute of Electrical Engineering | Jaksic Z.,University of Belgrade | Matovic J.,Institute of Sensors and Actuating Systems
Microsystem Technologies | Year: 2010

We investigated intrinsic noise in plasmonic sensors caused by adsorption and desorption of gaseous analytes on the sensor surface. We analyzed a general situation when there is a larger number of different analyte species. We applied our model to calculate various analyte mixtures, including some environmental pollutants, toxic and dangerous substances. The spectral density of mean square refractive index fluctuations follows a dependence similar to that of generation-recombination noise in photodetectors, flat at lower frequencies and sharply decreasing at higher. Some of the calculated noise levels are well within the detection range of conventional surface plasmon resonance sensors. An AD noise peak is observed in temperature dependence of mean square refractive index fluctuations, thus sensor operating temperature may be optimized to obtain larger signal to noise ratio. A significant property of AD noise is its rise with the decreasing plasmon sensor area, which means that it will be even more pronounced in modern nanoplasmonic devices. Our consideration is valid both for conventional surface plasmon resonance devices and for general nanoplasmonic devices. © 2010 Springer-Verlag. Source


Kostic M.,Nikola Tesla Institute of Electrical Engineering
Elektrotehniski Vestnik/Electrotechnical Review | Year: 2014

Following results of an investigation the author comes to the conclusion: that the Potier reactance values depend almost solely on the turbogenerator(s) reactive currents (iaQ = ian·sinφ) for a given voltage value. This new rule is after being proven by a general qualitative analysis, and his thesis is then checked for rated values of the active and reactive power, for an example of a GTHW 360 (360 MW) turbogenerator. Potier reactance xp is determined by the open-circuit saturation characteristics and data from the zero-power factor at the rated voltage ( ia = ia,90·cosφ=0, U=Un). The procedures used to determine the Potier reactance (xP or xP,n) are performed by a computer with the open-circuit saturation curves if0 (e0) expressed analytically. The author proposes the IEC 34-4 standard in the part on the determination of the Potier reactance. It is proposed that it is determined for the excitation current corresponding to the rated voltage and armature current value ia,90 = ian·sinφn, at the zero-power factor (overexcitation), and for some characteristic values is needed for the construction of the turbogenerator capability curve (P-Q curve). Source


Kostic M.,Nikola Tesla Institute of Electrical Engineering
Elektrotehniski Vestnik/Electrotechnical Review | Year: 2010

In this paper, a simplified procedure is presented to be used in determination of magnetic reluctance and inductance of short cylindrical coils of a rectangular cross- section. Based on calculations and analysis, it is shown that the coil with a greater ratio of longitudinal/transversal dimensions of the cross-section should rather be approximated with the rectangular cross-section than the circular cross-section, as it is usually done. It is proven that the magnetic conductance to the flux outside the inner part of the coil is directly proportional to the perimeter of the contour of the cross-section given. By using obtained results, the impact of the cross-section shape on the values of the magnetic reluctance and inductance of short cylindrical coils can be analyzed. Source

Discover hidden collaborations