Geoscience Reference
In-Depth Information
state that the sampling skew is <10 ps among all of the
channels and has a 200 ps overall sample rate jitter. This
represents a very small amount of timing error causing
all of the measurements to be synchronized to a very small
fraction of a sample period, making the data among all of
the channels temporally highly synchronous. The digitally
multiplexed signals are subsequently serialized into a bit
wide data stream and sent through a fiber-optic cable
(to achieve galvanic isolation) to a USB-based computer
interface. The entire system, including the computer,
was operated onbatteries tominimize conductive coupling
with the electrical power system.
The voltage reference for the system is contained
within the measurement area and was designed into
the data acquisition system to be a part of the common-
mode sense (CMS) and common-mode range control
(DRL) electrodes (see Haas et al., 2013, and reference
therein for further information on the common-mode
control used in the BioSemi system and the effect of elec-
trode impedance on the measured response). The CMS
and DRL electrodes are used in combination to make a
feedback control system that keeps the CMS electrode
as close as possible to the reference voltage at the ADC.
In this system, the CMS electrode becomes a dynamic
reference potential. All of the digitized data are saved in
the raw data form and are referenced to the CMS elec-
trode. The data are recordedwith all of the common-mode
signals, and as a result, any channel can be used as the
reference channel. In fact, the maximum common-mode
rejection in the system only fully occurs when a single
channel of choice is subtracted from all of the other
channels (see
http://www.biosemi.com/
; see Crespy
et al
.
, 2008, and Ikard et al., 2012, for further explana-
tions). In these measurements, we selected the least active
signal in the data (channel 4) as the reference. This type of
measurement system allows us to change the reference
electrode as needed to correct for dynamic voltages occur-
ring at the CMS electrode. Therefore, this type of system is
best suited for dynamic self-potential signals. The flow-
chart used to analyze the raw electrical data is shown in
Figure 5.16.
Any AE that would occur during the experiment were
measured using 6 WS
3
8
4
7
6
9
5
1
2
3
Figure 5.15
Unconfined cement block sensor configuration.
(1) Cement block; (2) 34 Ag
AgCl electrodes (BioSemi); (3)
six acoustic emission sensors (Mistras WSa); (4) plastic plate
with top array of 16 channels of Ag
-
AgCl electrodes (BioSemi);
(5) plastic plate with back array of 16 channels of Ag
-
AgCl
electrodes (BioSemi); (6) Hole #9, high-pressure fluid injection
tube; (7) Hole 10, high-pressure fluid injection tube; (8) Hole 6,
high-pressure fluid injection tube (not used); and (9) other holes
(no injection).
-
manufactured by BioSemi, Inc., designed for EEG research
(
http://www.biosemi.com
/). During the experiments, the
electrical potential measurements were acquired with
32 amplified nonpolarizing silver
-
AgCl) electrodes as described earlier. The electrode poten-
tials were measured using the BioSemi ActiveTwo data
acquisition system that is completely self-contained, bat-
tery powered, galvanically isolated, and digitally multi-
plexed with a single high-sensitivity analog-to-digital
converter (ADC) per measurement channel. Thirty-two
24 bit ADC, employing the sigma-delta architecture, were
used in the ActiveTwo data acquisition system. The Acti-
veTwo system has a typical sampling rate of 2048 Hz with
an overall response of DC to about 400 Hz. This measure-
ment system has a scaled quantization level of 31.25 nV
(LSB) with 0.8
-
silver chloride (Ag
V rms noise at a full bandwidth of 400
Hz and a specified 1/
f
noise of 1
μ
V pk/pk from 0.1 to
10 Hz. The common-mode rejection ratio was higher than
100 dB at 50 Hz, and the amplified nonpolarizing electrode
input impedance was 300 M
μ
sensors manufactured by Physical
Acoustics Corporation (PAC; see position in Figure 5.15).
All six sensors had an operational frequency of 100
α
at 50 Hz (10
12
Ohm //
11 pF). Because there is a single ADC per channel, the
digitization part of the system had very small sampling
synchronization errors. The specifications for the system
Ω
-
900
kHz and a resonant frequency of 125 kHz. PAC
sMicro-
II PCI-2-8 digital AE system chassis was used to run the
AEwin data collection and posttest data analysis software.
'
