Every boat going to sea should have at least two compasses. One MUST be an analog, or mechanical, unit -- one that requires no batteries or electrical connections, most often this is a steering compass, mounted within easy view of the helm. The second compass MUST be a hand-bearing compass of some kind, Either a mechanical or electronic compass will do. Given that this is a marine electronics column, I thought we'd start with a review of how electronic, or fluxgate, compasses work.
For many of us, the Autohelm Personal Compass was our introduction to electronic compasses. It was (and mine is, if I can remember where in the chart table I put it) a great, easy-to-use, ergonomically pleasant machine with one great drawback -- it only worked if you held it absolutely flat. On a rolling deck, you became a human gimbal. If you didn't keep the thing exactly level, your bearing would be off, perhaps even way off. Alas, there was no way you could tell this unless you took multiple bearings of the same target. All fluxgate compasses suffer from this problem, though few as severely as the Autohelm.
The reasons for this have to do with how fluxgate compasses are constructed.
According to Dr. Bill Lee, of AlphaLab Inc., the core of every fluxgate magnetometer is a loop of iron nickel foil. Approximately 10,000 times per second, this loop is magnetized, demagnetized and then magnetized in the opposite polarity by an excitation coil wrapped around the loop. As the AC current in the coil increases, the loop of foil becomes increasingly magnetized. At some point (fixed by the system's construction), the loop's magnetic field saturates and fails to keep pace with the excitation coil. The loop's field remains saturated even as the current in the excitation coil begins to decrease. Eventually, the loop's field decreases, goes to zero and then strengthens in the opposite polarity as the excitation coil goes through the remainder of its AC cycle. Each episode of saturation or desaturation of the loop produces a brief pulse of current in a pickup coil surrounding the foil loop and excitation coil. If the earth had no magnetic field, the foil loop would always saturate at exactly the same point in the excitation coil's AC cycle, and the pulses would be recorded at exactly the same phase of the AC cycle.
But earth does have a magnetic field, about half as strong as the one produced in the foil loop. Earth's field penetrates the loop inducing a magnetic field in it even before the excitation coil starts its cycle. Once the AC cycle starts, the foil loop saturates at an earlier phase in the AC cycle than it should, an event indicated by the emission of saturation pulses, duly picked up by the pickup coils. On the second half of the AC cycle, earth's magnetic field acts to delay the onset of saturation Combining the two readings gives the strength of the magnetic field in one direction.
To get direction, we need to measure the field strength in at least two perpendicular directions. Modern units do this by using two different pickup coils perpendicular to one another. Comparison of these two-axis measurements provides the orientation of earth's magnetic field and hence the magnetic bearing. Typical units run at 10,000 cycles per second; approximately 1,000 cycles are averaged to give 10 readings per second. Some manufacturers use software to select an update rate, but rest assured that any fluxgate compass you have is going through at least 1,000 measurements every time it shows you a single reading.
The problem with the Autohelm was that the foil loop was fixed in the compass' housing. Tilting the compass changed the apparent strength of the magnetic field hitting the loop and hence the apparent direction. The sensitivity at North American (magnetic) latitudes is pretty high -- as much as 3 degree bearing error per degree of tilt! Newer fluxgate compasses solve this in a number of ways. For example, according to Chris Watson at KVH Industries, their Datascope uses a foil loop that floats in a light oil bath. To first order, the ring is self-leveling. The Datascope can be tilted up to 20 degrees from horizontal and still work within the 0.5 degree accuracy limits.
Beyond this angle, however, beware of any readings you take. One might want to measure the horizontal angle between two marks by holding the Datascope sideways, and that's great -- just don't take a bearing at the same time!
For those of you with the interest, you can easily see this behavior yourselves. I know this technique works with Autohelms, and Chris Watson thought it might work with the Datascope. Take the compass to the least magnetically disturbed area you can find. Find magnetic north, and then slowly tilt the compass down while observing the bearing. Eventually you will find a point where the inaccuracy of the reading is maximized. The compass is now pointing exactly parallel to earth's magnetic field. The angle of the compass from the horizontal is the magnetic inclination, or magnetic latitude. It was sometimes used in days of yore as a line of position. I'd appreciate hearing from any readers who try this. Send your name, lat/long, the inclination and the type of compass you have to email@example.com. For those of you interested in comparing your readings to the real thing, see http://geomag.usgs.gov/usimages.html#us_i.
Thanks again for all your emails -- I haven't come close to answering all of them yet, but keep them coming. Practical Sailor recently rated various hand-held bearing compasses, find them at www.practical-sailor.com.
-- Larry McKenna