BoatUS puts dead batteries at 9% of on-water assistance calls.
Charging problem or parasitic draw? Three tests tell them apart
Test the battery alone. A load test at half the rated cold cranking amps for fifteen seconds should hold above roughly 9.6 volts at normal temperature. A battery that fails is finished, and no charging-system work fixes it — starting batteries deep-cycled by a stereo or trolling motor fail this young.
Test charging, engine running. At 1,000–2,000 RPM a healthy 12-volt system holds roughly 13.5–14.5 volts at the battery terminal. Resting voltage with the engine running means the stator, the rectifier/regulator or the wiring is not delivering. Over about 15 volts means the regulator failed the other way and is boiling the battery. One failed diode in the rectifier cuts output without killing it — the classic "it charges, but never quite keeps up."
Test for a parasitic draw. Everything off, key out, meter in series with the negative cable, reading milliamps. Electronics with memory draw tens of milliamps — normal. Hundreds is a fault and flattens a battery over a long weekend. Isolate a fuse at a time until the number drops. Usual culprits: a bilge float switch cycling on a boat catching rainwater, an amplifier that never sleeps, or an accessory wired to the wrong side of the switch.
| Symptom | Points to | Confirming test |
|---|---|---|
| Dead after two weeks, fine when used | Parasitic draw, or a switch left on | Milliamp draw test, fuse by fuse |
| Cranks slow at the end of each trip | Charging not keeping up | Voltage at the battery, 1,000–2,000 RPM |
| Battery hot, sulfurous smell | Regulator overcharging | Charging voltage — over ~15 V is a fault |
| One engine charges, the other doesn't | Stator, rectifier or a cable | Compare the two engines |
Voltage drop testing, and why resistance testing misses it
The mistake that keeps boats broken is checking connections with an ohmmeter on a dead circuit. A corroded terminal reads near zero ohms at rest and still drops a volt and a half under starter load, because corrosion cut the effective contact area to a fraction of the cable. Drop testing instead runs the circuit live and loaded, probes either side of the connection, measuring what that connection steals. Standard marine wiring practice targets no more than about 3% drop on critical circuits and 10% on non-critical — on 12 volts, 3% is about 0.36 volts. On the starting circuit, more than roughly half a volt across one cable or connection while cranking is a fault worth chasing. The test walks from battery post to load — terminal, switch, bus, ground stud, block — and the reading names the bad joint. The numbers go in the service record, emailed and texted before we leave the dock.
Ground faults and the corroded-connection reality
More marine electrical failures start on the negative side than owners expect, and grounds are the last thing anyone checks. Salt water wicks up untinned copper strand by strand, under the insulation, out of sight — a wire can look perfect at the terminal and be green and powdered eighteen inches back. Butt connectors without adhesive-lined heat shrink fill with salt and become resistors. Standard marine practice is tinned copper, crimps with adhesive heat shrink, and solder never used as the sole mechanical connection: a soldered joint becomes a stress point that cracks under the vibration an outboard bracket delivers all day.
The common faults: a corroded engine-to-negative-bus connection that makes every gauge misbehave, a helm ground stud crusted behind the panel, one overloaded stud shared by accessories added a year apart. Repairs are made at the connection, corroded wire cut back to bright copper — not by stacking another ring terminal on a bad stud.
Gauges not working, and NMEA 2000 bus faults
Every device dies at once. That is the backbone, its power insertion, or a terminator. An NMEA 2000 network is a backbone with a 120-ohm terminating resistor at each end and short drops to the devices. Powered down, the backbone should measure about 60 ohms — two terminators in parallel. Roughly 120 means one is missing; well under 60 means somebody added a third, a common mistake after an electronics install.
One device dies, the rest are fine. That is the device or its drop cable — drops are length-limited by the standard, and one too long or holding water takes out a single node. Intermittents that worsen in a chop are usually water in a connector or an uncapped T on the backbone.
Analog gauges have their own list — a blown accessory fuse, the key switch, a helm ground, a failed sender. Because engine data comes off the ECU, a gauge reading nonsense while the ECU reports the correct value proves the fault is downstream of the engine, saving an entire wrong repair.
Stray current and galvanic corrosion in a saltwater slip
Two different problems, constantly confused.
Galvanic corrosion is dissimilar metals connected electrically in an electrolyte, and saltwater is an excellent one. Slow, normal, managed with anodes replaced past roughly half consumed. On shore power the green safety ground also ties your underwater metal to every other boat on the dock — the job of a galvanic isolator.
Stray current corrosion is DC leaking out of a boat's wiring, or a neighbor's, and returning through the water. Far faster — it can visibly eat a gearcase, trim tab or shaft in weeks. The tells: anodes gone early, corrosion on one boat while its neighbors are fine, new pitting after a wiring change.
Diagnosis is systematic — anode consumption history, hull potential against a reference electrode, then isolating circuits one at a time. The source is sometimes the boat, sometimes the dock; busy facilities like Riviera Beach Marina Village have many boats and much shore power in the same water. Where it is the marina's infrastructure, you get that in writing, with numbers, for the dockmaster.
Questions people actually ask
My battery is new and it still dies. What now? A new battery rules out one cause and leaves two. Next is charging output at the battery terminal at 1,000–2,000 RPM, which should read roughly 13.5–14.5 volts, then a milliamp draw test with everything off. If the boat takes rainwater in a slip, check the bilge float circuit first.
Can you fix corroded wiring, or does the whole harness need replacing? Most is repaired at the connection: wire cut back to bright copper, re-terminated in tinned marine wire with crimps and adhesive heat shrink, then drop-tested to prove it. Full replacement is for harnesses where corrosion reached the connector pins — a call made with a drop reading, not by eye.
My anodes are disappearing in a couple of months. Should I worry? Yes. Anodes lasting a fraction of their normal life is the classic sign of stray current, which can destroy a gearcase in weeks. Test it — the source may be a neighboring vessel or the dock's wiring, not your boat. You get the readings in writing either way.