The master’s thesis deals with the noise diagnostic in the solar cells. Describes the main kinds of noises. The samples were quality and reliability screened using noise reliability indicators. The samples were surveyed by measuring the I-V characteristics, the noise spectral density as a function of forward voltage and frequency. It was calculated the noise spectral density as a function of forward current.

The doctoral thesis deals with diagnostics of local defects in PN junctions and brings new information about microplasma noise behaviour and its usage for the temperature changes detection inside PN junctions. Defects in PN junctions are the source of microplasma noise. There were deviations observed in microplasma noise from the common known rectangle shape pulses during the measurements. These deviations were correlated with the temperature change directly in the defect area and in the defect area surroundings. Generation and recombination coefficients are commonly thought to be constant. However, these coefficients were observed to be not stable with time and this effect is explained in this work. The doctoral thesis then focuses on the PN junction parameters determination in the case when it is not possible to define unambiguously whether it is abrupt or linearly graded PN junction. The most significant parameters which are to be determined are barrier capacity, diffusion voltage and depleted area width in dependence on the voltage. The correlation between local avalanche discharge in PN junction and negative differential resistance appearance on VA characteristics of reverse-biased diode was qualitatively verified. The last important point in the work is computer modelling of temperature behaviour in the defect area and its surroundings during local avalanche breakdown. Thus the method of real diodes heating area parameters determination was introduced.