A simple technique to eliminate ground loop hum
in a terrestrially based hydrophone monitoring system.
Jack Kassewitz* - SpeakDolphin.com - Miami, Florida
www.speakdolphin.com
Robert Lingenfelser -Marine Mammal
Conservancy - Key Largo, Florida
www.marinemammalconservancy.org
Mark Weber-Sound
Specialist, Inc.-Miami, Florida *Author for
correspondence (e-mail J. Kassewitz:
info @
speakdolphin.com)
June 2007
Hydrophones have been used for the recording and
documentation of ocean events for many years. However
there is an inherent problem whether you are recording
on a boat or monitoring from a land based research site
- ground hum. This paper will outline a unique technique
that eliminates ground hum on land and propose a similar
technique for recording while on a boat.
|
|
Photo by Bob Care |
I. INTRODUCTION
Hydrophones are used worldwide to monitor underwater
sounds. Hydrophones, underwater microphones, are a
sound-to-electricity transducer for use in water or
other liquids, analogous to an ear for listening to
underwater sound. Such a transducer can sometimes also
serve as a projector (emitter), but not all have this
capability, and may be destroyed if used in such a
manner.
There is an inherent problem whether you are
recording on a boat or monitoring from a land-based
research site - ground hum. We were faced with a project
that needed to record, and periodically playback,
pristine extended frequency sounds for research
analysis, 24 hours a day for 7 months.
Our subject was Castaway, a pregnant female Atlantic
bottlenose dolphin (Tursiops truncatus), stranded on the
morning of November 11th, 2006 at Castaway Cove in
Florida. After that stranding Castaway was eventually
quarantined to the Marine Mammal Conservancy's (MMC)
rehabilitation facility in Key Largo, Florida, where it
was determined that the dolphin was hearing impaired.
In isolation, Castaway and her calf would not have
the normal "chatter" of a dolphin pod to assist the calf
in its language development - both in utero and also
after birth. The mother's hearing impairment raised
questions as to her ability to teach the calf on her
own. Therefore, an environmental enrichment plan was
implemented to play pre-recorded dolphin sounds to the
unborn calf through underwater speakers. We were
simultaneously monitoring and recording any responses
from the dolphin with high quality hydrophones.
During her quarantine at MMC, Castaway demonstrated
predatory stunning and killing abilities with sound
bursts. We were fortunate to not only have eye witnesses
during an event in which the dolphin stunned and killed
a mullet, but also to able to record this crucial
acoustical data without hum corruption because of the
grounding technique described in this paper.
Our biggest audio challenge was to record clean
samples for measurement and analysis. These recordings
of Castaway's vocalizations, echolocations and predator
stunning are of significant value to the study of
dolphin communication. It was a rare opportunity to be
able to record while Castaway was giving birth, the
calf's first breath, vocalizations, and development of a
signature whistle.
II. METHODOLOGY
A self-powered (two 9v batteries)
hydrophone, in less than three feet of water, was placed
adjacent to the birthing pen constructed for Castaway's
impending delivery. The cables were strung on an upright
pole and then passed through PVC conduit openings into a
research shed on land adjacent to the birthing pen.
Hydrophone XLR connectors terminated at the Grace
Lunatec V3 premium microphone preamp and the digital
192k/24 bit signal delivered interconnected via 50 ohm
AES/EBU digital cabling to the Sound Device 722 HD
digital recorder running at 192k/24 bit.
Immediately
upon connecting this XLR hydrophone cable hum was quite
prevalent and loud which would make any recording
useless. In fact, the hum was so loud that could not
monitor any of the sounds Castaway made or potentially
any of her expected calf's sound production. (See
picture 1a)
To remedy this problem we dropped a 1 meter
copper grounding rod into the salt water at the pen and
connected it directly to the recording equipment chassis
ground via a GB 10 grounding block. (See figure 1).
A
ground rod is a copper metal shaft used for establishing
earth-ground. These rods are typically 8 feet in length,
5/8 inches in diameter, driven into the earth and
grounding wires are clamped onto the sunken shaft to
establish electrical ground potential with the earth.
Ground rods should be free of paint or any other non
-conductive material and should be not less than 1/2
inches thick. Most applications call for a ground system
of 25 ohms or less. While, as a practical matter, you
may not get to 0 ohms, you certainly can get to 25 ohms
or less if the ground rods are properly installed (See
Picture 2).
We changed this usual configuration to a
water-ground. Although not necessary for non-ferrous
copper, zinc sacrificial anodes can be attached at both
ends to prevent electrolysis from attacking the
recording equipment. The rod was dropped into the water
beside the birthing pen. A10 awg solid copper wire
clamped onto the middle of the rod was run to the
grounding block in the research shed. At the recording
equipment a chassis-ground was established by soldering
an empty BNC connector shell (one with no center
conductor) onto a length of the 10 awg ground wire
leading from the GB10 ground block. This BNC connector
was plugged onto an unused wordclock output on the Sound
Devices 722 recorder to establish chassis-ground with
the electronics. For a recorder without an available BNC
we would have either terminated the ground wire through
pin 1 of an XLR connector and used any available I/O XLR
port or at the very least just twisted the ground cable
onto a connector release tab or alligator clipped to any
exposed chassis potential piece of metal.
The results
were instantaneous and remarkable! (See picture 1c)
There was absolutely no electronic hum even though
numerous other electronic research devices were running
in close proximity as well as the air conditioners and
security video system. (Please note that the
air-conditioner was grounded separately) Studio quality
sound was achieved only after a water- ground was
established.
1a |
|
1a is without any ground.
|
1b |
|
1b is with standard earth-ground.
|
1c |
|
1c
is with a re-configured water-ground. Notice
the hum line is completely gone. |
Note: See how the central
lines in the upper portion of these waveforms differ.
Picture 1a is without grounding protection and 1b is
with earth-ground and 1c is with water-grounding. There
is still residual hum in 1b; 1c is completely free of
hum.
III DISCUSSION
In dolphin research, where sound
quality and analysis are critical, a special signal
ground known as a "technical ground" (or "technical
earth") usually must be installed. This is not the same
thing as an AC power ground and no appliance ground
wires are allowed any connection to it as they may carry
electrical interference (See Figure 2). Great care has
to be taken to prevent any AC-grounded appliances (air
conditioners, etc), from making contact to the technical
ground as a single AC ground connection to the technical
ground will destroy its effectiveness. With the recent
proliferation of digital multi-track capable, high
sampling/bit rate enabled hard disc recorders the
installation of high-quality audio grounding has become
mandatory. Any ground-introduced interference tends to
be cumulative and obscures many of the frequencies being
recorded. This creates noise levels that may have been
satisfactory for older analog systems but become
intolerable and skew the results beyond scientific
recognition for today's exacting standards. We revised
this idea by creating a water-ground which differed
structurally only in the fact that rather than being
driven into the ground it was placed in the water
adjacent to the birthing pen.
The majority of available
hydrophones are wired in an unbalanced electrical
configuration making them very susceptible to
interference due to lack of shielding and the hum
rejection offered with balanced twisted pairs. Water
often has a different electrical potential than the
electrical ground potential of the recording equipment
being used. This relationship causes ground loop hum and
unwanted noise in hydrophones. In order to avoid this
interference from taking place on the recording, the
water must be properly connected to the chassis ground
of the hard disk recorder. In the case of recording from
a boat a copper wire immersed in the water and connected
to the chassis-ground will solve the problem. You may
use the same grounding wire as described below and
connected to a small metal plate for weight so that
submersion over the side of the boat is guaranteed.
Note: This metal plate should be connected with solid
core copper wire and an alligator clip at the equipment
end. Also sacrificial zinc oxide discs should be
attached to this plate if the recording is going to take
place over a long period of time to prevent electrolysis
from eroding the recording equipment.
IV. EQUIPMENT
A Sound Device 722 is a high-definition
two-channel digital audio recorder. The device records
audio to an internal hard drive, CompactFlash cards, or
external FireWire Drives. It writes and reads
uncompressed PCM audio at 16 or 24 bits with sample
rates between 32 kHz to 192 kHz. Additionally,
compressed (MP3) audio playback from 64 kb/s to 320 kb/s
is possible.
The 722 implements a non-compromised audio
path that includes Sound Devices' (next-generation)
microphone preamplifiers. Designed specifically for high
bandwidth, high bit rate digital recording; these
preamps set a new standard for frequency response
linearity, low distortion performance, and low noise.
The hydrophones used were
Model 8178-7 Hydrophones: We
currently are using two matched and calibrated Model
8178-7 hydrophones. They are wideband, omnidirectional
underwater sensors with a frequency response from 20 Hz
to 200 kHz. The spatial response is omnidirectional in
orthogonal planes with minor nulling along the axis
looking up the cable.
Hydrophone Sensors: The sensor is a small lead zirconate titanate (PZT) spherical sensor. It is fairly
thick walled and will withstand several thousand psi of
pressure. It is reverse poled on one hemispherical half
and series wired to increase sensitivity. The element
center is approximately ½" from the end inside the
rounded dark polyurethane end.
Hydrophone Electronics: The electronics include a low
noise, wideband FET input preamplifier, a pass filter at
15 Hz to reduce geophysical noise, a power supply
filter, and power input protective circuits. The
amplifier is included inside a metal EMI filter and is
encapsulated inside the PVC plastic body of the
hydrophone assembly.
Hydrophone Cable: The cable is 45 foot underwater
shielded cable. There are four conductors inside a metalized shield. On the top end, the power lines are
terminated in a small dual 9 volt battery clip set.
Hydrophone Acoustic Parameters: These hydrophones are
provided with calibration that indicates that the
hydrophone is flat down to about 20 Hz at about -168.5
dB re 1 v/uPa. The response from 2 kHz to about 10 kHz
is -168.5 dB, and then the response rolls off gradually
at about -172 dB. From 80 kHz to about 140 kHz, the
response is about -174 dB, then the response climbs
again to about -170 dB by 180 kHz. After about 200 kHz,
the response rolls off gradually to about 250 kHz, then
drops rather sharply at about -12 dB per octave.
Grace Lunatec V3 Signal Processing: The signal from
the hydrophone is routed through a Grace Lunatec V3 mic-preamp.
The Grace Lunatec V3 has an ultra-low distortion 24-bit
A/D converter. It has flexible outputs including SPDIF
(RCA) and two sets of AES3 (XLR). Sample rates are 44.1,
48, 88.2, 96, 176.4 and 192 kHz. The Grace Lunatec V3 is
also equipped with a unique analog noise shaping dither
circuit, or ANSRTM, which can be applied for word-length
reduction when sending signals to a 16-bit digital
recorder. Grounding Rod: A grounding rod is a metal
shaft used for grounding. These solid copper rods are
meant to be driven into the ground. Most codes call for
a ground system of 25 ohms or less. (In our case the rod
was just laid in the water) While, as a practical
matter, you may not get to 0 ohms, you certainly can get
to 25 ohms or less if the ground rods are properly
installed. The resistance of this ground or water is
intended to be 25 Ohms or less, but is rarely tested. To
verify the resistance of ground or water, it should be
tested with instruments using the fall-off-potential
method. (See picture 2)
|
|
Picture 2 |
Grounding Block (Electrical Isolator): Experiments
using hydrophones require that your electrical equipment
not add to the data that you are trying to collect. This
simple and cost effective grounding block will localize
and organize the ground circuit in your system. The
GB-10 comes with all the hardware needed to mount the
block to your isolation table and attach up to 10 single
leads. (See figure 1)
GB 10 |
|
Figure 1 |
V CONCLUSION
It was our goal to create a hydrophone
recording system that was beyond music studio quality
for precise research measurement as well as applications
demanding very broad extended frequency response and
extreme dynamic ranges. The first part of this applied
research had to overcome electronic hum and its damaging
effect to our research data. By using a simple grounding
rod and ground block inter-connected by copper cable and
terminating at water's ground potential we were able to
eliminate ground loops and potential interference.
Our
current spectrographic analysis shows no observable
electronic hum in the sounds we are recording. (See
picture 3.)
|
|
Picture 3 |
This relatively simple answer has allowed us to
record continuously for weeks at a time with no
appreciable sound distortion. The results from these
pristine digital recordings are currently being analyzed
and prepared for publication. We have seen the
remarkable first echolocation from the calf as well as
robust vocalizations only moments after birth. The
results of this research are forthcoming over the next
few months.
For questions contact:
info @ speakdolphin.com |