YSI SonTek ADVField, Technical Documentation Manual

Page: 210 / 228

SonTek/YSI
Acoustic Doppler Velocimeter Principles of Operation (September 1, 2001) 4
The vertical extent of the sampling volume is defined by the convolution of the acoustic pulse
length with the receive window over which the return signal is sampled. Both of these are pre-
cisely controlled by the ADV software (within the limits of the transducer bandwidth). The total
height of the sampling volume for the 16-MHz and 10-MHz ADVs is 9 mm; for the 5-MHz
ADVOcean probe, it is 18 mm. The vertical edges of the sampling volume can be considered de-
fined to
±
0.5 mm for the 16/10-MHz ADV probes, and to
±
1.0 mm for the 5-MHz ADVOcean
probe. Since the pulse length and receive window are controlled by software, the height of the
sampling volume can be reduced with changes in the data acquisition software (see §7.4).
It is important to note that for all measurements given by the ADV, the sampling volume location
is specified as the vertical center of the sampling volume. For example, if the 10-MHz ADV
shows the sampling volume to be located 2.0 cm from the boundary, the leading edge of the sam-
pling volume will be [2.0 cm – (0.5 * 9 mm)]
≈ 1.5 cm from the boundary.
5. Pulse-Coherent Processing
The description of ADV operation given in Section 2 is a simplification of the way in which ve-
locity is actually measured. Section 2 describes incoherent Doppler processing: the transducer
sends a single pulse of sound and measures the frequency change of the return signal. In reality,
the ADV uses a technique called pulse-coherent (or pure-coherent) processing. In this technique,
the instrument sends two pulses of sound separated by a time lag; it then measures the phase of
the return signal from each pulse. The change in phase divided by the time between pulses is di-
rectly proportional to the velocity of the particles in the water. Pulse-coherent processing is used
because it provides the best possible spatial and temporal resolution.
This section does not attempt to provide a detailed description of pulse-coherent processing. It
presents a general overview with a focus on how this affects ADV operation. SonTek can pro-
vide additional references on pulse-coherent processing upon request.
There are several aspects of pulse-coherent processing that affect ADV operation. First is the in-
herent limitation on the maximum velocity that can be measured. Pulse coherent processing
measures the phase of return signals; phase measurements are limited to a range of [-
π
,
π
]. If
phase exceeds these limits, it will “wrap around” (i.e. if phase increases to just above
π
, the ADV
measures a phase of -
π
). This is known as an ambiguity jump, where (for example) the ADV will
measure a negative velocity rather than the true, larger positive velocity.
The maximum unambiguous velocity is a function of the time lag between the two pulses. The
ADV offers the user a choice of a number of pre-set velocity ranges, each of which corresponds
to a particular pulse lag. All operational changes required for the different velocity ranges are
handled automatically by the ADV.
As is discussed in §6.1.1 and §6.1.4, the instrument noise level scales directly with the velocity
range setting (higher velocity ranges have a higher noise for each sample). Thus, you should al-
ways select the lowest velocity range that meets the requirements of the particular experiment.
Pulse-coherent processing affects ADV operation in two other situations. When making near-
boundary measurements, there is a potential that the reflection of one pulse from the boundary
could interfere with the other pulse; this is discussed in §7.3. Additionally, the ability to adjust
the time lag between pulses gives the ADV excellent performance for applications with low flow
velocities; this is discussed in §7.5.