Marine Electronics Refit Guide

A marine electronics refit is a systems-engineering job, not a shopping list.

The biggest mistake is replacing equipment before understanding what the existing equipment does. It is a very easy mistake to make, because the old gear looks obsolete and the new gear looks exciting, and the dependencies between them are invisible.

Some real examples of what that invisibility costs:

• An old display is the only device feeding position to the VHF. Remove it and the DSC distress alert silently loses its position.
• A legacy instrument is the only interface to a perfectly good depth transducer. Remove it and you have bought a hull penetration you did not need.
• An autopilot control head looks twenty years old while the expensive mechanical drive behind it remains entirely serviceable — and the drive is most of the cost.
• A "redundant" cable turns out to power the instrument bus

Document first. Buy later.

#Step 1: inventory every device

Record, for everything:

• manufacturer
• model
• approximate age
• serial number
• power circuit it sits on
• network connection type
• sensor inputs
• outputs to other devices

Include hidden black boxes. Autopilot computers, sonar modules, AIS units, gateways and network switches are frequently behind panels, under bunks or in lazarettes, and they are frequently the expensive part.

#Step 2: draw the existing network

Identify which technologies are actually present:

• NMEA 0183 — serial, common on older gear
• NMEA 2000 — the modern sensor bus
• proprietary Ethernet — radar, sonar, video
• SeaTalk variants — older Raymarine
• analogue transducers — depth, speed, wind on legacy systems
• engine interfaces, often proprietary

Do not cut an unknown cable because it appears redundant. Trace it. The half-hour spent following a cable is cheaper than the day spent working out why the wind instrument stopped.

#Step 3: identify dependencies

This is the step that saves the most money, and it is a set of specific questions:

• Where does the VHF get position from?
• Where does the autopilot get heading from?
• Which unit supplies depth, and to what?
• Which device powers the old instrument bus?
• Does the AIS have independent GNSS, or does it rely on the network?
• What is providing the NMEA 2000 network power, and where is it injected?
• Which device is the terminator at each end?

Answering these often reveals that a device you were about to remove is load-bearing, and that a device you were about to keep is doing nothing.

#Step 4: inspect DC power

Check:

• battery voltage, at rest and under load
• breakers and fuses, and whether they are the right rating
• cable size against the run length
• corrosion, particularly at ring terminals and in bilge-adjacent runs
• grounds and bonding
• spare capacity for what you are adding

New electronics connected to poor wiring are still unreliable electronics. A display that resets occasionally is very often a voltage-drop problem, not a faulty display — and it will be diagnosed as a faulty display, replaced, and continue resetting. See marine DC power.

#Step 5: decide what can stay

Good retention candidates:

• a quality through-hull transducer in a sound fairing block — expensive to replace, often still excellent
• a sound autopilot drive with serviceable ram, quadrant, mounts and linkage
• a recent AIS transceiver
• compatible NMEA 2000 sensors
• a good VHF antenna with sound coax

Replace equipment because it fails the mission or the compatibility plan, not merely because it is old. Age is not a fault. A 2015 through-hull transducer in clean flow will outperform a new one badly sited.

#Step 6: choose the primary ecosystem

Only now choose between Garmin, B&G, Raymarine, Simrad, Furuno and Lowrance.

Base the choice on:

• radar family and form factor
• sonar and transducer compatibility
• autopilot architecture the steering system needs
• sailing features, if the boat sails
• camera and thermal requirements
• local installer support from your home port

The full comparison is in marine electronics brands. The reason it comes at step six rather than step one is that steps one to five usually eliminate most of the candidates for you.

#Step 7: design the new NMEA 2000 backbone

Plan:

• the route through the boat
• termination at both ends, and exactly two
• power injection point, sized for total load and run length
• T-piece positions at each device
• cable lengths, including drop lengths kept short
• future expansion — spare T-pieces cost almost nothing now and a great deal later

A significant refit should finish with a network diagram. See marine networks.

#Step 8: design Ethernet

Radar, sonar and cameras require high-bandwidth networking. Plan:

• switch locations, and how many ports
• cable routes, respecting bend radius
• spare ports for later additions
• service access to the switch

#Step 9: inspect the autopilot mechanics

Inspect, physically:

• the ram or drive unit
• the quadrant
• the tiller arm
• hydraulic hoses and fittings for weeping
• mounts, and whether they flex under load
• the steering linkage throughout
• the clutch or engagement mechanism

Do not assess the pilot only from the control head. The head is the cheapest part and the least likely to be the problem. As autopilots argues, the drive is the muscle, and a new computer on a tired drive is money spent in the wrong place.

#Step 10: check transducer compatibility

Before replacing the sounder, identify:

• transducer model
• frequency and bandwidth
• connector type
• mounting arrangement and fairing angle
• hull material and construction

Some existing transducers are worth preserving, sometimes via an adapter or a black-box sonar module. See depth sounders and transducers.

#Step 11: plan sensor locations

For each of:

• heading sensor — low, central, away from magnetic influence, and compass-tested before drilling
• GNSS antenna — clear sky view, clear of the radar beam
• radar — height versus serviceability, level in normal trim
• wind — masthead, aligned, with the offset calibrated
• VHF antenna — height, and quality coax sized for the run
• AIS — dedicated antenna or approved splitter, decided deliberately

Sensor position directly affects performance, and it is the one thing a firmware update can never fix. See sensor siting and installation.

#Step 12: preserve service access

You should be able to reach, without dismantling joinery:

• fuses and breakers
• network terminators and T-pieces
• connector backs
• junction boxes
• the network power injection point

A beautiful helm that cannot be serviced is a poor installation. This is the step most often sacrificed to finish quality, and the one most regretted at 0300 three seasons later.

#A staged refit

#Stage 1

• a new NMEA 2000 backbone, designed once with spare capacity
• a new primary MFD
• retain usable sensors

#Stage 2

• AIS
• radar
• heading upgrade

#Stage 3

• autopilot computer or drive, as the inspection in step nine determined

#Stage 4

• instruments
• cameras
• advanced sonar

This spreads cost without creating a fragmented system — provided the backbone and Ethernet plan are done properly at stage one. A staged refit becomes a mess when each stage improvises network around the last one. See the cost guide.

#Commissioning

A refit is not complete when the display turns on. Test:

• GNSS source selection
• heading, and its source
• radar alignment against a charted coastline
• AIS transmit and receive, with correct static data
• VHF position input
• depth offset, measured and labelled
• wind calibration, including the alignment offset
• autopilot, including in a following sea
• engine data and correct device instances
• alarms, and their audibility from the sleeping cabin
• backup navigation, actually opened

The full sequence is in commissioning and sea trials.

#Documentation pack

Leave aboard, physically and in a cloud folder:

• a power distribution diagram
• an NMEA 2000 diagram with terminators and injection point
• an Ethernet diagram with switch locations and spare ports
• a fuse and breaker list
• software versions for every device
• the MMSI as programmed
• calibration values and offsets, including the depth offset convention
• instance settings
• manuals, or links to them

This is what makes the boat troubleshootable by someone who did not install it — including you, in three years, at night, somewhere with no dealer.

#Bottom line

A good refit makes the boat simpler, more reliable and easier to troubleshoot. A bad refit installs new boxes on top of old weaknesses, and costs more because the weaknesses are now harder to reach.