Electronics that reboot when the engine starts, drop out in a seaway or lose their heading data at the worst moment are frustrating, and the instinct is to blame the unit. In practice the supply, the connector or the network is at fault far more often than the display.
The pattern of when it happens tells you a great deal before anyone opens anything up.
Each pattern points somewhere different, so record it before you start testing.
Electronics are sensitive to low voltage in a way lighting is not. A plotter fed through a long thin run, sharing a circuit with a windlass, autopilot or thruster, can dip below its working voltage every time the other load bites.
Measuring at the back of the unit while the offending load is switched shows this straight away. The fix is usually a dedicated, correctly sized feed rather than a new plotter.
On NMEA 2000, a missing or duplicated terminator, a backbone powered at the wrong point, or too much current drawn through a drop cable will cause devices to appear and disappear at random. One faulty device can also disturb the whole bus.
Isolating devices one at a time, and checking backbone voltage at the far end, sorts network problems from equipment problems.
Behind a helm pod is a damp place. Corroded pins, unsupported cable weight pulling on a plug and non marine connectors all cause intermittent contact that follows boat movement rather than any electrical logic.
Cranking pulls system voltage down briefly. Well designed installations feed electronics so they ride through that, so a reboot usually points to supply routing or cable size rather than a faulty unit.
Damage is uncommon, but incorrect power feeding and missing terminators cause unreliable data, which on a passage is bad enough.
Generally yes. A dedicated, properly sized and protected feed for navigation electronics removes an entire category of intermittent faults.
We test supply, voltage drop and network integrity, then correct the cause rather than replacing working equipment.