Campbell CSAT3, Instruction Manual

Page: 22 / 72

CSAT3 Three Dimensional Sonic Anemometer
14
signal processing techniques result in more accurate, lower noise
measurements.
Unlike CSAT3s running embedded code version 3, embedded code version 4
does not use a library of calibration templates and pattern recognition to
determine the ultrasonic time of flight. Embedded code versions 4 and 3 have
the same wind resolution. The resolution of the speed of sound measurement
using embedded code version 4 is 15 times worse than embedded code version
3 (0.002
°
C to 0.025
°
C RMS); however, the resolution of the speed of sound
measurement is still a factor of two better compared to any other commercially
available sonic anemometer.
Field comparisons of measurements made with a CSAT3 anemometer
containing embedded code version 4 and one containing embedded code
version 3 show that the new algorithm significantly improves the
anemometer’s ability to obtain measurements under conditions of low signal to
noise ratio, e.g. during rain events. Additionally, the water wicks installed on
the CSAT3 transducers contribute to improved performance during rain events
by preventing large drops from accumulating on the transducer faces.
7.2 Embedded Code Version 3
The CSAT3 measures wind speeds and the speed of sound by determining the
time of flight of sound between pairs of transducers, as described in Appendix
C. However, unlike many other commercial anemometers, it does not use
simple threshold detection to determine the ultrasonic times of flight. Instead,
it uses advanced digital signal processing techniques and pattern recognition
algorithms which, in comparison to other systems, results in more accurate,
lower noise measurements. The pattern recognition algorithms allow the
system to quickly detect and indicate a degradation of signal quality and
system performance. As a consequence, a CSAT3 running embedded code
version 3 was highly sensitive to rain droplets on the transducers faces.
Placing wicks on a CSAT3 running embedded code
version 3 will not improve the CSAT3s performance in
the rain. The wicks prevent large drops of water that
completely obscure the sonic signals from
accumulating on the transducer faces and are part one
of a two part solution in operating in rainy conditions.
The second part is in the digital signal processing in
embedded code version 4.
The CSAT3 signal processing techniques result in slight differences in
operation when compared to simpler sonic anemometers. These differences
include a start-up delay before the system is fully functional plus a requirement
for the correct triggering of the sensor to ensure a constant flow of
synchronized data. The general principles of operation are described in the
following section.
After the CSAT3 transmits an ultrasonic signal from one of its transducers, it
records the received signal from the opposing transducer. To optimize system
performance, the CSAT3 only looks for the received signal during a limited
window in time. The exact start and end times of the window is determined not
WARNING