Marine Heading Sensors and Satellite Compasses

Heading is the least glamorous number on a marine network and the one whose errors do the most damage, because it is an input to almost everything else.

A six-degree heading error is invisible on the display. It is not invisible in its consequences: it rotates the radar overlay, biases every MARPA solution, corrupts true wind and therefore every layline, and gives the autopilot a false reference to steer against. One badly sited sensor degrades four systems at once, and each of them looks broken in its own way.

That is why this guide treats heading as infrastructure rather than as an accessory.

#Heading is not course over ground

These get conflated constantly, and the difference is physical.

Heading is the direction the bow points. It is an orientation, and it exists whether or not the boat is moving.

Course over ground is the direction the vessel actually travels across the earth. It is derived from successive GNSS positions, so it requires movement to exist at all.

They differ whenever something pushes the boat sideways:

• current, which can be substantial on the New South Wales coast where the Chart Datum in Australia 2026">East Australian Current runs
• leeway, the sideways slip a sailing yacht makes under wind pressure
• drift while manoeuvring or holding station

Heading:             040°
Course over ground:  057°

Seventeen degrees of difference is a lot of water going past sideways, and the boat is not misbehaving — it is being set. As the GNSS guide explains, a single-antenna receiver can derive course over ground while moving but cannot tell you where the bow points, and at anchor or near zero speed its derived direction becomes unstable because tiny position changes produce wildly varying directions.

#What a modern sensor actually is

Current heading sensors are not magnetic compasses with a wire attached. They combine three sensor types and fuse the outputs:

• magnetometers, which sense the earth's magnetic field and provide the absolute reference
• gyroscopes, which sense rate of rotation and provide short-term stability through motion
• accelerometers, which sense the gravity vector and provide attitude

Fusing them produces an AHRS — an Attitude and Heading Reference System. The gyroscopes carry heading through the moments when the magnetometers are being thrown around by pitch and roll, and the magnetometers stop the gyroscopes drifting over time. Each covers the other's weakness, which is the same design principle that runs through the whole of marine navigation.

Depending on the model, an AHRS may output heading, rate of turn, pitch and roll. Rate of turn in particular is what lets an autopilot anticipate rather than merely react, and what lets radar processing stabilise a picture during a turn.

#Magnetic interference is the main enemy

A magnetometer measures the magnetic field it sits in. It cannot distinguish the earth's field from the field produced by a loudspeaker magnet forty centimetres away.

Common offenders, roughly in order of how often they cause problems:

• loudspeakers — a large magnet, frequently mounted in exactly the dry accessible locations that look ideal for a sensor
• high-current DC cables, especially windlass and thruster feeds, which produce a field only when energised
• alternators and electric motors
• batteries
• steel tools, spare anchors and chain, which may be stowed somewhere different on every passage
• iron ballast and engine blocks
• anything with a magnet in it — headphones, phone mounts, magnetic catches

The intermittent sources are the worst. A constant field can be modelled out by calibration; a field that appears only when the windlass runs cannot be, because the calibration was performed when it was absent.

#Siting, and testing before you drill

Aim for low, near the centreline, near the vessel's centre of pitch and roll, on a rigid mounting surface, in a dry location, with a known orientation.

Then test the location before committing to it. Walk a handheld compass slowly around the candidate site and watch the card for deflection. Repeat the test with the engine running, the windlass operating, the thruster energised and heavy DC loads switched on. Ten minutes of this will save a great deal of subsequent confusion, and it is the core of the sensor siting guide.

A convenient installation location is very often a poor magnetic location. That sentence is the whole chapter.

#What calibration does and does not do

Automated compass calibration — usually a slow circle or two at low speed — builds a model of the vessel's local deviation, the error the boat's own magnetism induces at that sensor position. It is genuinely effective and should always be performed.

It models deviation. It does not rescue a sensor mounted beside a major magnetic field, and it cannot model a field that comes and goes. If calibration will not converge, or the result varies between attempts, the answer is almost never to try again — it is to move the sensor.

Recalibrate after moving the sensor, after significant steel work, after adding equipment near it, and if the radar overlay starts drifting.

#Satellite compasses

A GNSS compass sidesteps magnetism entirely. It uses two or more antennas separated by a known baseline and derives orientation from the difference in satellite carrier phase between them.

Advantages:

• stable at zero speed, so heading is trustworthy at anchor and while manoeuvring
• immune to local magnetic deviation, so a steel hull is no longer a problem
• fast, high-rate heading, which benefits radar and autopilot
• pitch and roll output on advanced models, which allows radar and sonar stabilisation

The cost: price, two antenna positions with a clear sky view, and a baseline that must be aligned to the vessel's centreline within the manufacturer's tolerance. Get that alignment wrong and you have introduced a constant heading bias — the exact fault you bought the unit to avoid.

#Who genuinely benefits

• steel and aluminium vessels, where magnetic siting may have no good answer
• large catamarans, where the sensor is far from the centre of motion and the motion is quick — see catamaran navigation
• premium radar installations, where overlay and target tracking quality is the point
• vessels needing reliable heading at zero speed, including anything using dynamic positioning or working an anchor in tight quarters
• boats with difficult magnetic environments that testing has already proven

For a conventional fibreglass coastal cruiser, a well-sited solid-state magnetic sensor is enough, and the money is better spent on autopilot drive margin or a transmitting AIS.

#Source selection: the fault nobody looks for

A modern NMEA 2000 boat can contain several devices capable of supplying heading — a dedicated AHRS, a GNSS compass, an autopilot computer with its own sensor, sometimes a legacy instrument.

The network does not necessarily use the best one. Verify which source is selected after:

• a refit or any sensor replacement
• any firmware update, which has been known to reset source selection
• adding a device that also publishes heading

The best sensor aboard is useless if the network quietly prefers a weaker one. This is a standing item in commissioning and sea trials for exactly that reason.

#Radar as a diagnostic instrument

The fastest way to find a heading problem costs nothing.

In clear daylight, near a well-charted coastline, overlay radar on the chart and compare the two coastlines.

• Consistent rotational offset → heading misalignment, calibration error, or the wrong network source
• Offset that grows during a turn and settles afterwards → sensor lag or a rate-of-turn problem
• Radial mismatch, bearings correct → a radar range or timing adjustment, not a heading fault

Correct the heading first. Applying a radar bearing offset to compensate for a heading error hides a fault that is still corrupting MARPA, true wind and the autopilot — and it will reappear the next time the compass is recalibrated. The radar tuning guide puts this check before any radar control.

#Bottom line

Treat heading as infrastructure, not as an instrument. Site the sensor for physics and test the site before drilling, calibrate it and believe the result, verify which source the network actually selected, and use radar overlay as a routine health check. A better heading source improves several systems at once, which makes it one of the highest-leverage upgrades on the boat.