Presented:MTS Marine Instrumentation '90
San Diego, CA USA
February 1990

Alan J. Fougere

Steve Smith

Woods Hole Oceanographic Institution

Woods Hole, MA USA

ABSTRACT

1.0 INTRODUCTION

Although present water sampling systems, such as CTD (conductivity-temperature-depth) profilers interfaced to rosette samplers and Nisken bottles, satisfy the sampling requirements of most ocean research programs, the WOCE (World Ocean Circulation Experiment) Hydrographic Program has specialized requirements that necessitated the development of a new sea water sampler with improved capabilities. This new system must fulfill the requirements of the WOCE physical oceanographers, who require high-quality CTD profiles acquired at rapid profiling rates; and, the WOCE tracer chemist, who require uncontaminated 7-liter water samples from up to 36 sample levels per station.

Woods Hole Oceanographic Institution in conjunction with Battelle Ocean Sciences Division of Duxbury, MA, is developing a new integrated water sampler to meet the WOCE requirements¹. The proposed water sampler uses an innovative technique to collect oceanographic water samples and is completely described in reference 1. The water sampler uses a sealed multi-laminate bag with an integral inlet valve which is exposed to the sea water to be sampled. The bag is housed in a water tight drawer assembly which is connected via a selection valve to a high speed pump. Upon the request for a sample from the operator at the surface the selection valve moves to the appropriated drawer and the pump is subsequently turned on. The pump removes the "working water" from the draw which creates a differential pressure across the inlet valve of the sealed bag. This pressure differential results in the inlet valve opening allowing the sea water to be sampled to flow into the sealed bag. After an appropriate interval the pump is turned off reducing the differential pressure across the inlet valve; the inlet valve closes containing the sampled sea water. The controller which operates the sampler is configured as a SAIL² protocol device. SAIL protocol is a serial network with each instrument enabled by the transmission it's discrete address. The protocol allows joint operation of the sampler controller with additional instrumentation packages on the same communication link.

The sampler requires an operator to communicate with the underwater unit to select chambers, set exhaust direction and to set sample size (1 - 7 liters). Also the operator may request engineering status data to monitor sampler performance and to ensure correct operation of the device. Therefore, communication over the EM cable must be Bi-directional, allowing the downward transmission of control "commands" and upward transmission of conformation responses or "status" data.

Commercially available CTD systems use either Frequency Shift Keyed (FSK)³ or Manchester encoded data telemetry⁴ schemes to transmit data via the EM cable to a surface receiving unit. Although EM cables often have multiple conductors there use as individual communication channels is not feasible due to the stray capacitance between the conductors. This capacitance results in CTD data signals due to there higher frequency components being superimposed on all conductors after travel over 10 kilometer long EM cables. Also the use of individual conductors to transmit different signals reduces overall system reliability as individual conductor failures are frequent in EM cables⁵. These requirements resulted in the need to develop a telemetry system which would 1) work with standard commercially available CTD systems, 2) provided a Bi-directional full or half duplex command channel to control the sampler, 3) work over 10 kilometer single conductor EM cable, 4) support standard baud and protocol data transfer.

The command/status telemetry system developed is block diagramed in Figure 1. The system is connected between the CTD underwater unit and its corresponding deck unit. The operation of the CTD system is not effected by the use of the sampler command/status communication system. This is achieved by the use of frequency subdivision techniques of the available bandwidth on the EM cable. Shown if Figure 2 is a plot of the passband characteristics of standard UNOLS hydrographic cable. The response is flat with a small 2 dB peak at 3 Kilo hertz, and then drops of rapidly after 20 Kilohertz. Shown in Figure 3 is the spectral content of a Neil Brown Instrument Systems (NBIS) MKIIIB CTD FSK telemetry signal. As is expected the significant energy bands are centered on the two telemetry frequencies of 5 Kilohertz and 10 Kilohertz, however, there is significant energy across the entire sea cable bandwidth radiated from the instrument. From Figures 2 and 3 it is evident that there is little bandwidth available to add the command status channel above the MKIIIB CTD Data signal. Therefore, the command/status communication uses transmission frequencies below the lower CTD carrier frequency.

2.0 COMMAND/STATUS MODEM

The command status telemetry system uses a Di-Bit Phase Shift Key (DPSK) data encoding technique. The use of DPSK allows the telemetry system to operate with low frequency carriers of 1200 and 2400 Hertz while achieving full duplex 1200 baud operation. Listed in Table I are the Di-Bit phase shift representations of the carrier signal. Each of the four (4) relative phase shifts represents two data bits. The carrier is phase modulated at a rate of 600 times per second resulting in a data transfer rate of 1200 bits per second. The DPSK modulator base-band signal output is then filtered to reduce intersymbol interference. Demodulation is the reverse of the modulation process with the incoming analog signal eventually decoded into Di-Bits and converted back into a serial bit stream. The demodulator also recovers the data clock, which was encoded into the signal during modulation. The command/status modem uses a phase locked loop coherent demodulation technique that allows for better performance than other types of di-bit demodulators. Plotted in Figure 4 is bit error rate versus signal to noise ratio for the DPSK demodulator.

The command/status modem is susceptible to interference from signal frequencies transmitted by the CTD. Commercially available CTD systems radiate energy outside of their FSK bands which is a result of internal digital circuitry, DC/DC converters, and signal switching internal to the instrument. These noise sources must be eliminated to allow the command/status modem to operate in an acceptable signal to noise environment. CTD designers also intended their underwater units and deck terminals to be connected directly together providing AC grounding appropriate to signaling frequencies used. At lower frequencies these networks do not provide sufficient AC grounding. Another requirement of CTD operation is that they receive their operating power by DC current supplied on the same conductors as the data telemetry.

3.0 COMMAND/STATUS UNDERWATER UNIT

The CTD issues require the command/status telemetry underwater unit to remove the CTD data from the EM cable. The CTD signal is processed to remove interfering frequency components, then, the command/status data is added to it, and the resultant signal is remodulated back onto the EM cable. Shown in Figure 5 is a block diagram of the underwater unit. The system passes the EM cable through a P1 filter that essentially acts as an AC signal short. The P1-Filter also allows the command/status underwater unit to power its self via a tap taken from the center of the P1-filter. The combined CTD and command/status output signal is amplified and driven onto the sea cable by a Programmable Gain Amplifier (PGA).

A series resistance "Ps" in series with the PGA enables transformer T2 to act as both a transmitter and receiver as it driven from a high impedance source. Winding W2 of T2 receives both the "command" signal transmitted from the surface and also the "echo" of the CTD data and "status" data transmitted from the underwater unit. The magnitude of the receive signal "Vr" across winding W2 is directly proportional to the turns ratio "N" of the transformer, series resistor "Rs", and the resistance of the sea cable "Rc". The relationship is given by the equation:

1) Vr = Vi * N^2 * Rc / ( N^2 Rc + Rs)

Where = Vi = Input Signal from the PGA amplifier.

In the command/status underwater unit receiver a difference circuit enables rejection of the locally transmitted portion of the signal. To insure that the echo suppression circuit is operating properly and that received signals are within the linear range of the Digital Signal Processor Band Bass Filter (DSP BPF) and Automatic Gain Control (AGC) circuit is used. The AGC circuit detects the level of the "locally" transmitted signal as it reflected in the winding W2 of transformer T2 and maintains this magnitude at a constant level. This insures that both the CTD signal and the command/status signal are optimally echo suppressed by the difference circuit. Measured echo suppression was found to be consistently in excess of 30 dB for 0, 5 and 10 kilometer length cables.

4.0 COMMAND/STATUS DECK UNIT

The Command/Status Telemetry Deck unit is similar in operation to the underwater unit. A block diagram of the deck unit is shown in Figure 6. The major difference in the deck unit is the use of a high pass filtering in the CTD signal path to eliminate the effects of the locally transmitted "command" signal from interfering with the CTD Deck Terminal Demodulation process. The unit also uses an AGC loop to control the level of the locally transmitted signal.

5.0 TESTING

At the time of publication the system has been tested over a 0 and 10 kilometer lengths of UNOLS EM Cable (Cable Specifications: 3-Conductor #19 AWG 19/.008" W/ .016" Wall Polypropylene Dielectric, .015" Wall HDPE Belt, 16/.0375" SGXXIPS Inner Armor, 22/.0375 SGXXIPS Outer Armor, R= 9.4 ohm/1000 ft, C= 35 pF/1000 ft). Testing was run to verify non interference of CTD data. EG&G Ocean Products IBM/PC Software was used in conjunction with a MKIIIB CTD and a MKIIIB Data Terminal. Over a period of 5 hours zero (0) frames of bad data were received. The command Status link was simultaneously tested with two (2) IBM/PC computers transmitting a message of 12 ASCII characters. This message was transmitted approximately 70 thousand times during the test. The message was corrupted 2 times resulting in a bit error rate of < 1 ppm. We believe these corruption's were the result of a local radio transmitter that was in close proximity to the test cable.

6.0 PHYSICAL IMPLEMENTATION

The underwater unit resides on two (2) Instrument Bus⁶ cards which mount in a 6" I.D. pressure housing. The underwater unit is self powered from the DC power provided on the sea cable for the CTD. The unit consumes 750mW. The deck unit is configured on a IBM/PC card format and mounts directly in an option slot of any IBM computer. The deck card is powered from the IBM back plane and communicates via a DB 9-Pin serial connector accessible at the rear of the computer.

7.0 CONCLUSION

The authors have developed a Command/Status Telemetry system which works directly with UN-modified CTD systems over standard oceanographic EM cables. This system provides full duplex 1200 baud communication between the surface and the WOCE water sampler, and can be used to control or collect data from any device with serial communication which is used in conjunction with commercially available CTD's.

Table I

DI-BIT PHASE REPRESENTATIONS

DI-BIT VALUES

PHASE CHANGE (DEGREE)

00

01

11

10

+90

0

270

180

Figure 1

TELEMETRY SYSTEM BLOCK DIAGRAM

Figure 2

PASSBAND CHARACTERISTICS OF 10KM UNOLS EM CABLE

Figure 3

SPECTRAL OUTPUT NBIS MKIIIB CTD

Figure 4

DPSK MODEM

BIT ERROR RATE vs. SIGNAL TO NOISE

Figure 5

BLOCK DIAGRAM COMMAND/STATUS TELEMETRY

UNDERWATER UNIT

Figure 6

BLOCK DIAGRAM COMMAND/STATUS TELEMETRY

UNDERWATER UNIT

1. H. Berteaux, S. McDowel, J. Bullister₁ Carl Aibro, and A.J. Fougere, "Conceptual Design: An Integrated Sea Water Sampler for the Ocean Sciences", National Science Foundation, June 1989.
2. IEEE Serial Ascii Communication Protocol, IEEE-997 Standard, 1985.
3. N.L. Brown, "A precision CTD microprofiler", Proc. IEEE Oceans '74 Conference, vol.2 no.2, pp. 270-278, 1974.
4. A.M. Pederson, "A Modular High Resolution CTD System with Computer-Controlled Sample Rate", Proc. MTS 1984 STD Conference and Workshop, San Diego, February 1984.
5. H.O. Berteaux, R.G. Walden, P.R. Clay, and R.E. McDevitt, "Experimental Evaluation of CTD Package Hydrodynamic Behavior and Recommendations for Improved Lowering Techniques", WHOI-83-21, July 1983.
6. E.C. Mellinger, K.E. Prada, R.L. Koehler, K.W. Doherty, "Instrument Bus, An Electronic System Architecture for Oceanographic Instrumentation", WHOI-86-30, August 1986.