There has been significant progress in techniques for the determinatio
n of the activity size distribution of radon progeny. Despite this pro
gress, many inconsistencies still exist with respect to the activity d
istribution, especially in the smaller particle size range. The aim of
this work is to evaluate a measurement technique for determination of
the activity size distribution of radon progeny, and to investigate t
he effect of different screen configurations and different deconvoluti
on procedures. The measurements were performed in a large experimental
chamber, using three different sets of Wire Screen Diffusion Batterie
s (WSDB) constructed for these studies. Aerosol concentration and size
distribution in the chamber were monitored by a scanning mobility par
ticle sizer. The Twomey and Reconstruct deconvolution methods were use
d to derive the activity particle size distribution from the diffusion
battery data. Investigations of activity size distribution using thre
e WSDB sets of different sensitivity ranges showed that the distributi
ons calculated with the Reconstruct program correlate well for each ba
ttery set. It has also been demonstrated that while application of nar
row spectrum battery sets sensitive to particular narrow size ranges c
ould provide better results for these ranges, results obtained with th
e battery set which covers the whole spectrum are, in fact, close to t
hose for narrow spectrum sets. (C) 1997 American Association for Aeros
ol Research.