OCR Text |
Show source of noise for our measurements was scattering by high concentrations of small particles. This background noise limited the low end of the size range to that shown in the histograms of Figures 5a-5d. Since the concentration of particles was the highest in the acid spray area, this is also the area in which the background noise was the highest. The spent acid contained a small but sufficient amount of iron to account for a concentration of 0.5 pm ash particles which was greater than 10 cm . One method of reducing the noise due to small particle scattering is to decrease the size of the measurement volume by reducing the aperture size. This could have been easily done for our measurements, but also would have reduced the counting rate of the measurement. The concentration of larger particles in our system, the particles which were of greatest interest for the purposes of the measurement, was small. In the acid spray area, the concentration of particles at the measurement volume location which had a diameter greater than 8 urn was approximately 0.1 cm . As will be discussed later in this section, the small number of counts in the signal strength histogram is one of the major limitations on the accuracy of the size histograms which are reported here. Thus, while the use of a smaller measurement volume would reduce the background noise due to scattering from small particles, it would also lower the count rate. Ideally, two measurement volumes should have been used for each measurement - a large measurement volume to enable the measurement of a high number of large particles, and a small measurement volume to reduce the background noise in the measurement of small particles. On the other hand, in measurement locations 2 and 3 where the background noise was very low, particles with diameters as small as ~1 um were measured. Reducing the background noise even further to measure particles sizes of 1 um or less is not warranted because of a problem which is discussed in the following paragraph. The measurement of particles in the vicinity of 1 um is confronted with a problem beyond that of background noise. While single particles on the order of 1 \im or less can be detected by the laser backscatter technique if the background noise is sufficiently low, the accuracy or resolution to which their size can be determined is limited by the highly resonant or "noisy" character of the differential scattering 24-17 |