Holding the catch at −1.5 °C, without the alarm.

§ 01 · What we wire and what we don't

What we wire: the control panel, the field wiring from controller to compressor and from controller to evaporator, the temperature and pressure transducers, the safety interlocks, the alarm horn, the bilge-pump cross-link if you have one, and the data logger.

What we don't wire: the refrigerant circuit, the compressor itself, the evaporator coil, or anything mechanical inside the system. That is the refrigeration shop's responsibility, and we work alongside them, not over them. The shops we co-deploy with most often are Fish Ex Marine Refrigeration in New Bedford and Thermo King Marine Coastal at the South Terminal.

§ 02 · Safety interlocks (the part that matters)

If the system uses ammonia, the wiring inside the engine room is intrinsically-safe per ASHRAE Standard 15-2022, with explosion-proof fittings on every junction box, and the alarm-horn circuit is on a separate fused supply that survives a main-board fault.^[1] The leak detector (typically a Calectro or MSA Galaxy) trips a relay that kills the compressor and opens the engine-room ventilation fan; the wiring to that relay is the most carefully-thought-out part of the boat.

If the system uses R-507A or R-404A, ASHRAE 15 still governs but the Class is different and the explosion-proof requirement does not apply. The relay logic is similar. We always wire the alarm horn loud and we always wire the bridge-light annunciator to the same circuit.

§ 03 · The four ways an RSW board fails

We have written a bulletin on this and the four most common failures, in our experience, are:

1. The temperature probe drifts. The Pt100 RTD probe in the hold ages and reads 2–4 °C low. The hold is actually warmer than the controller thinks. We replace the probe and re-calibrate against an Onset HOBO reference.
2. The low-pressure cutoff trips. Often because the suction-line strainer has clogged with slush. Mechanical, not electrical, but the symptom looks electrical.
3. The contactor pits. A 50-amp contactor on a 25-amp duty cycle fails open after about 11,000 cycles. We replace with a properly-sized definite-purpose contactor.
4. The 24V control transformer fails. Almost always because someone fed it 240 V into the 208 V tap. We re-fuse the primary and re-strap.

§ 04 · Data & logging

For boats serving the premium scallop market — auction quality, day-boat product — the buyer increasingly wants a hold-temperature log for the entire trip. We install Onset HOBO MX2202 loggers with a sensor in the hold, accuracy ±0.2 °C, and we configure them to push to the captain's phone over Bluetooth at the dock. For larger boats with NMEA 2000, we install a Maretron TMP100 sensor on the same backbone we wire under Plate P-06; the data goes to the chart-plotter and is logged to an SD card.

§ 05 · Cost & lead time

A new Carel-based RSW board, including field wiring and commissioning: $14k–$22k, three to five working days. A retrofit of an existing failed Honeywell-based board to Carel: $9k–$14k, two to three working days. A simple controller replacement (same-for-same): $1.6k–$3.2k plus the controller, half a working day.

Hourly rates and parts markup are on the rates page. The single-page estimate is sharp.

Cross-references: P-04 (switchboards) when the RSW load triggers a board upgrade, P-06 (NMEA 2000) for hold-temp logging on the bridge, and the April 2026 bulletin for the long-form on these failure modes.