Truth and Compasses
Compasses have come a long way since Columbus, and things are still heading in the right direction.
When Columbus set off across the Atlantic, he took along a magnetic compass, but he didn’t entirely trust it. His attitude was understandable: Most seafarers of his time either stayed in sight of the coast or looked to “the star of the Virgin Mary” for guidance. A compass, to them, was nothing more than a backup.
For the first three weeks of the voyage, Columbus kept meticulous notes of how “the needles” compared with the star. But within a month, he had made up his mind: “The star,” he wrote, “moves as other stars, and the needles always point truly.”
Swinging-card compasses have developed quite a bit over the past 500 years but we now know that they do not “always point truly;” they work by sensing the earth’s magnetic field so they are affected by the magnetic fields of the boats that carry them (deviation) and by the fact that the earth’s magnetic field isn’t lined up with its north-south axis (variation). And they’re not very good at sending electronic heading data to other instruments. It is this factor, probably more than anything else, that has led to the rising popularity of flux-gate compasses.
But flux-gate compasses also depend on the earth’s magnetic field, so they are still subject to variation and deviation. To those traditional problems they add a new error—by producing significant “tilt errors” if they are not kept perfectly horizontal.
Fortunately, over the last few years, GPS compasses are appearing in increasing numbers. These are even more sophisticated than the typical GPS receiver, yet their underlying principle is quite simple: If you have two GPS antennas onboard one boat, you can calculate your heading by comparing the position of one with the other. The fact that they are in more or less the same place and working at exactly the same time means that most of the errors that affect GPS as a positioning system become irrelevant; it doesn’t matter, for instance, that a satellite may be a couple of feet out of position or that its signals have been distorted by the earth’s atmosphere because the effect of these errors is exactly the same at both antennas.
The trouble is the GPS C/A code signal (the one for civilian use) is transmitted at a chip rate of 1.023 MHz. What that means in everyday language is that a civilian GPS receiver is measuring its distance from each satellite with an electronic tape measure whose graduations are more than 300 yards apart—nowhere near accurate enough to produce a practical GPS compass.
The solution is a technique called carrier-phase comparison. It’s useless for finding the position of a single antenna but very good at comparing the positions of two antennas. It works by ignoring the coded messages that are being sent by the satellites and looking instead at the radio waves that carry them. The primary GPS frequency is 1,575.42 MHz—more than 1,500 times higher than the C/A code chip rate—so it’s like using a tape measure whose graduations are only eight inches apart instead of 300 yards.
A typical GPS compass is a flat-topped, elongated dome, a few inches wide and a couple of feet long, mounted parallel to the centerline of the boat. Inside are two GPS patch antennas about a foot apart and linked to a receiver and processor in the center. If it is receiving signals from a satellite directly abeam, the radio waves that carry those signals arrive at the two antennas perfectly in step with each other (“in phase”). But suppose the satellite moves or the boat turns so that the satellite is a few degrees aft of directly abeam. Now the forward antenna is slightly farther from the satellite than the aft one, so instead of the received signals being perfectly in phase, they gradually slip farther and farther out of step. When the satellite moves to approximately 20 degrees aft of abeam, the radio waves received by the two antennas are precisely out of phase with each other.
As the satellite continues to move or the boat continues to turn, the waves received by the two antennas gradually come back into phase, until, when the satellite is about 40 degrees aft of the beam, they are back in step with each other. The actual figures vary depending on the geometry of the antenna mount, but there is always a solid mathematical link between the phase difference between the two antennas and the direction from the boat to the satellite.
Nothing connected with GPS technology is really that simple, so it almost goes without saying that there are many other things happening inside those simple white domes. One radio wave is exactly the same as the waves immediately before and after it, so the processor has to use some clever math to decide whether the two antennas are receiving the same wave as each other or not. And any GPS compass has to include a conventional GPS-positioning receiver so that it knows where the boat is and where the satellites are. Without that information, knowing the direction to a moving satellite would be useless as a heading reference. And because it’s working in three dimensions rather than two, it has to be able to calculate and compensate for changes in phase caused by the boat pitching and heaving.
Some GPS compasses—generally those intended for larger or commercial vessels—overcome some of these problems by using three antennas rather than two, but more recent models tend to go for the double-antenna arrangement with the addition of rate gyros (similar to those used in stabilized cameras). So a GPS compass is capable of delivering all the basic GPS data including position, track, and velocity, plus rate of turn, roll, pitch, and heave. Definite numbers as to the amount the boat is rolling, pitching, and heaving doesn’t just add credibility to those barroom bull sessions when you’re describing how nasty it was out there; it can also be used by some sonar systems to smooth out the wavy bottom trace and smeary fish echoes that you get in rough weather.
But the best bit is that gyro-assisted compasses deliver an accurate heading reference that is immune from the variation and deviation afflicting magnetic compasses, without the inertia of ≠swinging-card compasses, and without the high cost, weight, and long start-up times that make gyrocompasses so unattractive.
More than half a millennium after Columbus, we are at last using man-made stars to give us “needles” that really do point truly.
This article originally appeared in the July 2011 issue of Power & Motoryacht magazine.