An automated system to minimize variability in an aerosol concentratio
n over rime in an animal exposure chamber was developed. This system i
ncludes a microprocessor-controlled stepper motor to regulate the spee
d of a dust generator, a real-time aerosol monitor to instantaneously
measure chamber aerosol concentrations, and a personal computer with a
n analog-to-digital (AID) converter, serial port and relay board for s
ignal input and output. A computer program was written to use these de
vices to produce a chamber concentration near a level chosen by the op
erator. This program was also written to start and stop an exposure af
ter a specified time period and to continuously display the data recei
ved from the,e aerosol monitor on the computer screen during the expos
ure. A combination of operating conditions, which include a moderately
high chamber Now rate (0.32 m(3)/min) and high generator rotation spe
ed while using a small scraper blade, optimized the performance of the
automated system by minimizing the concentration variability (CV) see
n in the aerosol monitor readings. The CV, which indicates the average
difference between consecutive monitor recordings, decreased from 2.7
9 mg/m(3), when using the large scraper to produce a chamber concentra
tion of 25 mg/m(3), to 0.47 mg/m(3) when using the small scraper at 25
mg/m(3). When operating with these optimum conditions, 5 consecutive
trials of 180 min produced concentrations of 10.52, 10.52, 10.57, 10.3
7, and 10.47 mg/m(3). The system proved to be capable of calculating t
he appropriate generator speed needed to reach and maintain a desired
concentration level after trials were performed to determine the amoun
t oi aerosol losses in the chamber. A total of 14 trials performed al
10 mg/m(3) had an average actual-to-expected concentration ratio of 1.
023 (SD = 0.067). The response of the concentration level to changes i
n generator speed after start-up varied in relation to the magnitude o
f the speed change. The system proved to be more accurate when the gen
erator speed was changed to provide a change in chamber concentration
of more than 2 mg/m(3). The accuracy and reliability of the aerosol mo
nitor were investigated by comparing the rime-weighted average of the
monitor recording with the average chamber concentration measured by a
sample filter (MIF ratio). The average difference of the MIF ratio be
tween consecutive trials was 3.36% when using the small scraper near 1
0 mg/m(3) and 9.71% when using the large scraper near 25 mg/m(3).