Das Projekt "Untersuchung des Rauschverhaltens von optischen Partikelzaehlern" wird vom Umweltbundesamt gefördert und von Universität Duisburg, Fachbereich 9 Elektrotechnik, Fachgebiet Prozess- und Aerosolmesstechnik durchgeführt. In many production systems and in research the measurement of particles gains more and more importance. With increasing technological advances smaller particles will gain importance. The areas which require reduced particle concentration are for example the semiconductor industry, the food production and the medicine technique. For particle measurements optical particle counters are often used. To fulfil the requirements these particle counters have to be improved. Goals are obtaining an enhancement of the accuracy of size analysis, a higher security of counting and a reduction of the detection limit with regards to the particle diameter. These demands can be met with the improvement of the signal processing. The signal processing of optical particle counters must detect the particle signal in the composite signal. The composite signal consits of the noise and the particle signal. The limit of the particle detection is set by the particle concentration and by the noise. The noise in the signal affects the signal processing in different ways. The amplitudes of the particle signals are stochastically falsified and it results in a wider particle frequency distribution. If the amplitude of the noise is greater than the threshold then the noise is counted as a particle. This effect occurs near the lower detection limit regarding to the particle diameter. To improve the signal detection a digital signal processing has to be developed which uses the features of the composite signal to minimise the signal to noise ratio. The essential condition for this development is to characterise the composite signal. The magnitudes are described with their moments of the distribution. The noise consists of - the noise of the electronics - the scattering light from molecules and - the scattered light from surfaces. The noise of the electronics has mostly thermal sources. The scattering light of molecules interacts with the molecules in the measuring cell. The number of molecules are function of the temperature, the gas species and the pressure. The scattered light from surfaces is produces from the lighttrap and the optics. The characteristics in the distributions of the noise and the particle signals, which are the same in all optical particle counters, have to be investigated. To detect the characteristics of the noise in optical particle counters the parameters temperature, gas species, pressure and the surface must be varied. With the investigation of the noise behaviour a digital signal processing system which improves the optical particle counter can be developed.