Jory Bell designed MITESS according to specifications outlined by Ed Boyle. The ultra-low-power electronics were designed by Joe Betts. MITESS was refined according to results obtained on the Bermuda Testbed Mooring and revised over several years by Jory Bell, Joe Betts, Ed Boyle, Rick Kayser, and Barry Grant.
The mechanical and electronic design of the sampling units focused on the following goals and criteria:
* Trace-metal clean materials; easy to clean
* Sample bottle is well-flushed before sealing
* Simple mechanical operation
* Withstands stresses of extended (6-12 mo.) deployment at any depth
* Easy to join in a variety of configurations
* Low cost, simple manufacture
* Reliably deployed on standard moorings at no risk to mooring
After a series of three laboratory-scale prototypes, we have taken this instrument through one field test and more than eight mooring deployments. The first instrument was field-tested in Bermuda in October 1993. This test was successful, but the practical experience gained through the construction and field test taught us many things about how the instrument could be made simpler to construct and more reliable in the field. Accordingly, a second generation design was first deployed on the ALTAMOOR surface mooring in the Spring of 1995; a new sample acidification system were introduced in August 1995, and a new electronic board design deployed in winter, 1996. The basic design is now stable, and only minor modifications are anticipated in the future.
Basic operating principle: The instrument consists of a colony of independent sampling modules mounted onto a durable mooring unit. Each sample module is designed to function independently during deployment, so that failure of any one module does not prevent any other unit from functioning. For maximum flexibility, however, inter-module communication and external event-driven triggering is possible. A timer-controlled DC motor opens dilute (1N) high-purity-acid-filled polyethylene bottles by rotary motion. The low-density dilute acid is replaced by denser seawater through passive density-driven flow . Mixing during this process ensures that the bottle is effectively flushed with several volumes of seawater, with additional flushing from water motion relative to the sampler. Tests show that complete replacement occurs within a few minutes. The timer then reverses the motor, closing and sealing the bottle until recovery.
Sample preservation: Samples are preserved by diffusion of dilute acid retained in the diffusion chamber within the bottle. The diffusion chamber is constructed of high-purity teflon (exhaustively hot-acid-leached before use). When the bottle is opened, the dilute 1N acid in the bottle rapidly exchanges with the external seawater. The acid in the preservative chamber, however, cannot exchange rapidly during the several-minute sampling interval because of restricted flow. But over a longer period of several days, the acid diffuses into the sample bottle and preserves the sample. As a bonus, the acid which was contained within the bottle before sample collection accomplishes long-term leaching of the bottle right up to the moment of deployment, allowing maximum freedom from bottle contamination. (Although a portion of any contamination leached into the bottle will remain in the preservative chamber as well, 98% of any residual bottle contamination leached during deployment will be lost during sample collection. Because the sample bottles are always cleaned so as to be free of contamination before deployment, the additional 98% precaution is icing on the cake).
Mechanical Design: For trace metal cleanliness, virtually the entire exterior of the unit is constructed from ultra-high molecular weight polyethylene. Computer controlled machine tools are used to construct each module. The sample modules snap together in two dimensions; the most compact grouping places six units around a central support rod, and stacks these in two levels. Other wider geometries are possible as well. Each sample module has a sealed pressure-compensated liquid-filled (Fluorinert, a non-conducting fluid) interior chamber which houses the motor, electronics, and batteries. Because the electronics are viable at high pressures and operate in this nearly incompressible fluid, there is no need for a pressure case. The 500 ml polyethylene sample bottles within each module are opened by simple rotation. The bottles are replaceable by screw-cap bottles of other materials of similar geometry - e.g. teflon. The motor rotates a splined (hexagonal) shaft inside a bottle holder with the same screw pitch as the bottle cap (this is the same system used in mechanical pencils to advance the lead by rotating the cap). This simple motion is more reliable than complicated valves or linkages. The triangular footprint of each sampling module allows units to be arranged in a variety of space-saving configurations. The small size of each unit and its components allows the entire exterior portion of unit (or its component parts) to be cleaned and acid-leached to ensure trace metal cleanliness. Finally, the unit's component parts are small and easily machined with equipment found in a basic machine shop. This characteristic contributes to the low cost of constructing the sampler, and makes it possible for researchers to build or modify their own units.
The mechanical aspects of the sampler have gone through 3 prototypes and 4 deployments, with the most recent revision being quite minor.
Electronics Design: As part of the modular redundant design, each sampling unit has its own motor, electronics, and batteries. The electronics now consist of a single 6-cm diameter circuit board containing a microcontroller, real-time clock chip, and 25 other components. Before being deployed, each unit is programmed for the time it is to sample. The programming is done on the fully assembled unit by wireless communication via infrared transmission through the UHMW polyethylene, allowing for inter-unit communication and external reprogramming from an IR-linked laptop computer. The programming can either be done beforehand in the lab, or on the deck of a ship, because there is no need to open or otherwise handle the sampling units. The electronics are ultra-low power consumption (60 microamps continuous draw by electronics, 50 milliamp draw by motor during opening and closing) which should function more than a year . Six C-cell alkaline batteries (7000 ma hrs) are spot-welded in series and soldered to the circuit board for reliability. Motor rotation is detected with a magnetic sensor (Hall Effect sensor), allowing the controller to monitor bottle opening and closing and verify that the sample has been taken and sealed.
The electronics have gone through three major revisions. The first generation was used for the October 1993 deployment and worked as designed. However, we found that the layout of the board and the number of parts made these boards difficult to construct. Hence a radical revision of the design was undertaken by Betts, who made use of his experience at IBM to achieve a drastic reduction in component count (150 before, 27 now), size (one 6 cm diameter board now, two previously), and difficulty of construction (the boards are assembed now on an automated production facility by a contractor; previously they were hand-assembled by ourselves under a microscope). Finally, a final revision of the board design eliminated some potential shorting paths and improved its mechanical strength.
Moored Trace Element Sampler: statistics
Size:
diameter: just under 22 inches
height: a bit more than 6 feet
weight: just under 200 lbs in air, slightly buoyant in seawater
Mooring pin specifications
the 1" strength member is welded along a 2" cutout in plate
material = 316SS
dimensions = 5.5" x 3.5" x .5"
Publication:
Bell, J., J. Betts, and E. Boyle (2002) MITESS: A Moored In-situ Trace Element Serial Sampler for Deep-Sea Moorings, Deep-Sea Research I: 49:2103-2118
This web page was last revised on March 26, 2003.
.