Navigation & Communication Equipment

Every radar set on the bridge is, underneath the display, the same chain of stages doing the same job: generate a short burst of RF energy, send it out, listen for the echo, and turn the timing of that echo into a picture. When a set misbehaves, the practical question is never "what's wrong with the radar" — it's which stage in that chain has stopped doing its job, because that tells you where to put your meter and your screwdriver.

The modulator charges up and then dumps its energy into the transmitter — a magnetron on older or simpler sets, a solid-state transmitter (typically a chain of transistor amplifiers) on modern ones. The magnetron produces a short, very high-power pulse directly; it ages as it runs, its output falls off, and it eventually needs replacing as a unit. A solid-state transmitter produces a longer, lower-power pulse and compresses it electronically to get equivalent range performance, but it runs at lower stress and lasts far longer without the abrupt end-of-life a magnetron has. Either way, the duplexer's job is to protect the sensitive receiver from that transmit pulse and then switch the antenna over to receive within microseconds, so a duplexer fault often looks like "transmits fine, sees nothing" or, worse, a burned-out receiver front end.

Two numbers set radar performance directly from the transmitter timing, and they explain complaints you'll actually hear on watch: a set that "can't see far enough" and a set that "can't tell two contacts apart." The pulse repetition frequency (PRF) — how often the transmitter fires — sets the maximum range before an echo would arrive after the next pulse has already gone out and become ambiguous. The pulse length sets both the minimum range (you can't listen while you're still transmitting) and how close together two targets on the same bearing can sit and still show as two separate echoes rather than one merged blob.

• No transmission — suspect the modulator or transmitter; a magnetron nearing end of life under-performs before it fails outright.
• No reception — suspect the duplexer, receiver front end, or a stuck TR limiter; look for a picture with the sweep present but no returns at all, including no sea or rain clutter.
• No display, chain otherwise sound — suspect the processor or the video path to the monitor; the antenna may still be turning and the transmitter still firing.
• Working aloft on the scanner — treat it as a permit-to-work job: transmitter inhibited at the display, isolated if opening the turning unit, and confirmed dead before hands go near the waveguide or antenna.

A gyrocompass finds north by a completely different mechanism from a magnetic compass, and that mechanism is worth understanding properly rather than just accepting the output, because it explains both why the gyro is more reliable and where its particular errors come from. A fast-spinning rotor resists any twist to its spin axis; applied a small, continuous torque — supplied by a pendulous weight or an equivalent electrical control loop sensing that the rotor is not level — it doesn't tip over, it precesses, moving at right angles to the applied force. Engineered correctly, that precession drives the spin axis to settle on the meridian: the plane of true north, sought the way a compass needle seeks magnetic north, but by a torque-and-precession mechanism rather than a magnetic field. Because it is not sensing the earth's magnetic field at all, it has no variation and no deviation — the two errors that dominate magnetic compass work simply do not apply to it.

What it does have is settling time: after start-up, or after a course or speed change disturbs it, the gyro takes a period to precess back onto the meridian and stop oscillating around it. It also has speed and course-dependent error: a gyro on a moving ship is not quite responding to the earth's rotation alone, because the ship's own velocity over the surface adds a component that shifts the apparent settling point off true north by a small amount depending on the ship's speed, course and latitude. And it has latitude error and, transiently, ballistic deflection — a temporary swing immediately after a rapid speed or course change, before the gyro settles back down. None of this makes the gyro unreliable; it makes it a compass with known, predictable errors rather than an unpredictably deviated one.

Because those errors are predictable but not zero, the gyro heading is checked against a true bearing — typically an azimuth of the sun or a transit — once a watch, a deck duty the ETO supports by keeping the gyro and its repeaters correctly synchronised. The follow-up repeaters around the bridge and on the wings are themselves mechanical or electrical slaves to the master, and they drift out of synchronisation independently of the master unit; a gyro that is "maintenance-free" is a myth that costs marks and, at sea, costs a genuinely reliable heading reference.

• Settling time — the gyro's recovery period after start-up or a significant course/speed change; don't trust the heading until it has settled.
• Speed/course error — the ship's own velocity shifts the settling point slightly off true north; magnitude depends on speed, course and latitude.
• Latitude error — the precessional mechanism becomes less effective approaching the poles, where the horizontal component of earth rate falls away.
• Ballistic deflection — a transient swing following a rapid manoeuvre, settling out rather than leaving a permanent offset.

An autopilot is, at its core, a control loop: it compares the heading the gyro (or a magnetic compass repeater) reports against the heading you've set, and it commands the rudder to close that difference. What makes a good autopilot setup is not the electronics but the tuning of that loop to the ship's actual handling characteristics at the time, because a setting that works well for one loading condition and sea state can be actively wrong for another.

Three settings do most of the work. Rudder (or gain) sets how much helm the autopilot applies for a given heading error — too little and the ship wanders and is slow to correct; too much and it oversteers, wasting rudder movement and fuel and inducing a rolling, snaking track. Counter-rudder anticipates the ship's swing and takes off helm before the heading actually reaches the set course, so the vessel arrives on course without overshooting past it — a ship with a lot of momentum needs more counter-rudder applied earlier than a small, quick-turning one. Weather adjustment biases the rudder to compensate for a steady beam wind or sea pushing the bow off, so the autopilot isn't fighting a constant disturbance stroke by stroke.

None of these three are "set once." A ship that is light and high out of the water handles very differently from the same ship deep-loaded; a following sea that starts to push the stern around calls for different rudder and counter-rudder behaviour than head seas do. Part of keeping the steering system serviceable is making sure these settings actually get adjusted for the conditions rather than left on a default from the last passage, and that the off-course alarm and the changeover to manual or hand steering are tested and known to work before they're needed in an emergency, not discovered to be faulty in one.

An echo sounder measures depth the same way a radar measures range, just with sound instead of radio: it transmits a short pulse downward from a transducer in the hull, times how long the echo takes to return, and converts that time into distance using an assumed speed of sound in seawater — commonly taken as about 1500 metres per second, though the true figure varies a little with temperature, salinity and pressure. That reading is the depth below the transducer, and it is worth saying plainly: not below the keel, not below the waterline, and not the charted depth — it is the depth below wherever the transducer happens to be mounted in the hull, and every other figure has to be derived from that by applying known offsets.

Two corrections turn the raw reading into something useful. First, the instrument's own index (or zero) error — a small, fixed offset built into the electronics or introduced by the transducer's exact mounting — is applied to the raw reading before anything else, and it's found and logged by comparison with a lead-line or a known depth, not assumed to be zero. Second, the transducer's own position in the hull — how far above the keel, or below the waterline, it is actually fitted — is applied to convert "depth below transducer" into whichever reference the bridge team actually needs: depth under the keel for grounding risk, or depth below the waterline to compare against a chart.

The speed log works on an entirely different principle but suffers from the same "measured relative to what, exactly" problem. An electromagnetic (EM) log measures speed through the water by the voltage a moving conductor — the ship's hull, in effect — induces as it moves through the earth's magnetic field, picked up by electrodes on a small sensor unit; it always reads speed through the water, never speed over the ground. A Doppler log sends an acoustic beam down at an angle and measures the frequency shift of the return: it can bottom-track, giving true speed over the ground, when the water is shallow enough for the bottom echo to be usable, and falls back to water-track — measuring against a scattering layer in the water column — in deep water. Neither is wrong; they're answering different questions, and confusing "speed through the water" with "speed over the ground" is precisely the kind of error that costs marks and, at sea, costs an inaccurate ETA or a misjudged set and drift.

• Index/zero error — a fixed offset in the sounder found by comparison against a known depth, applied to every reading.
• Transducer offset — the fixed distance from the transducer to the keel or the waterline, applied to convert the reading to the reference actually needed.
• EM log — speed through the water, from induced voltage; unaffected by depth but never gives speed over the ground.
• Doppler log — bottom-track (over the ground) in shallow water, water-track (through the water) in deep water — know which mode it's in.

Strip away the chart display and ECDIS is a purpose-built computer, and it needs to be treated with the same two-sided discipline as any other piece of safety-critical IT: the hardware and software have to be of an approved, known type, and the content it's displaying has to be kept current. Type approval covers the unit itself — the manufacturer's hardware and software combination has been tested against the performance standard and found to meet it, which is why you can't simply run chart software on an arbitrary PC and call it an ECDIS. Software version compliance is the part that keeps changing after the unit is installed: as the international chart presentation and content standards are periodically revised, the ECDIS software has to be updated to remain compliant with the current version, and a unit running software that has fallen behind the current standard is exactly the kind of thing a port state control inspector checks — an out-of-date ECDIS software version has genuinely been used as grounds for detention.

Chart updates are the content side of the same discipline: electronic navigational charts (ENCs) are corrected regularly, and the ship needs a working arrangement — automatic where the connectivity supports it, manual update loading where it doesn't — to keep the displayed charts current, with the update history auditable. An ECDIS showing a beautiful, clear picture built from charts that are months out of date is not meeting the requirement just because the screen looks fine.

Finally, ECDIS carriage comes with a backup arrangement, because a vessel navigating primarily on ECDIS cannot simply go dark if the primary unit fails: that backup is either a second, independent ECDIS or an appropriate folio of paper charts, and it needs to be genuinely usable — kept current, and with someone aboard who can actually navigate on it — not a box ticked once at delivery and never revisited. From the ETO's side, the practical work is keeping the unit's software and the vessel's ENC permit and update subscription current, confirming the backup arrangement is real, and knowing what evidence a surveyor will actually want to see for each of those.

• Type approval — the hardware/software combination tested against the performance standard; not any PC running chart software.
• Software version compliance — kept current with the presentation/content standard as it's periodically revised; a known detainable deficiency when it lags.
• Chart updates — ENCs corrected on a working schedule, automatic or manual, with an auditable update history.
• Backup arrangement — an independent second ECDIS or a current paper folio, genuinely usable, not just present.

GMDSS equipment earns its keep only if it works the one time it's actually needed, which is why the whole installation is built around a testing schedule rather than an assumption that hardware left switched on stays serviceable. The daily routine is an internal self-test of each set — VHF, MF/HF, NAVTEX, and the rest — checking that the equipment powers up correctly and reports no fault, without transmitting to another station. Weekly, a DSC test call is made to confirm the digital selective calling function — the part that actually triggers an automated distress alert — is genuinely working end to end rather than just the voice channel. Monthly, the reserve batteries are checked and the EPIRB and SART are inspected, including their own battery expiry and hydrostatic release where fitted.

None of this matters to a surveyor if it isn't written down: every one of those tests is logged, with the date, the result, and who carried it out, and a radio survey spends as much time reading that log as it does looking at the equipment itself. A perfectly functioning radio room with a gap in the test log looks, to an inspector, exactly like equipment nobody has actually checked — because from the paperwork, that's indistinguishable from the truth. Keeping the log complete and contemporaneous, not reconstructed from memory the night before a survey, is as much a part of GMDSS maintenance as replacing a battery.

Underneath the testing schedule sits the reserve source of energy: a battery, or batteries, sized and maintained to keep the required GMDSS installation running on its own if the ship's main and emergency supplies are both lost, for a duration set by the vessel's equipment fit and area of operation. Sizing and checking that reserve — working out what it actually has to power, for how long, and whether the battery as it currently stands can still deliver that — is a genuine calculation, not a guess, and it's one the ETO should be able to do from first principles rather than from memory of a number.

• Daily — internal self-test of each set, no transmission to another station required.
• Weekly — a DSC test call, confirming the distress-alerting function itself, not just voice.
• Monthly — reserve batteries, EPIRB and SART checked, including battery/hydrostatic release expiry.
• The log — inspected as closely as the equipment; an unlogged test is, to a surveyor, an untested set.

The three problems below work through the numbers behind a radar timing question, an echo-sounder depth correction, and a GMDSS reserve-battery check — the kind of layered calculation an exam question builds from a single formula into a decision.