Galley & refrigerator

Rotten from the start

On Day 3 I cleaned the rotten plywood out of the galley wall cabinet and kept the frame as a template; that's in the Beginning story. In September 2022 the galley countertop came out. Nothing to cry about: it was rotten trash, and I knew that from the first moment.

In February 2023 the faucet and the sink went back in, so at least we could use water there.

In March I took off the plywood around the galley window and found the window frame completely rotted out. With the plywood gone there was a lot of extra space, for the electric panel and for getting into the ice chest.

An ice chest becomes a refrigerator

My plan: take the guts from a small household refrigerator and build them into the ice chest. Here's a picture for my notes of how the wires connect. I drilled a hole into the ice chest for the evaporator lines and built a shelf for the compressor. The red and blue pipes are for water. The shelf is glassed from below and to the hull on three sides, and I re-glassed the joint between the ice chest and the hull.

Then the compressor shelf was finished (final sanding still needed), and I glassed in rails for the condenser coil on the back of the LPG locker.

In April the condenser went on the back of the LPG locker, and the system was assembled and brazed, I hope well. It was really crazy to braze copper on a fiberglass boat with a MAPP torch and no oxygen. Wet rags everywhere, a couple of stainless plates against fire, and an extinguisher right next to me.

Then I charged it... and nothing got cold. The compressor ran, and the pressure dropped almost to zero. Something was blocked. I cut the thin capillary tube and saw pressure pushing oil back from the evaporator, but nothing coming from the high side. When I opened the filter-dryer, a nice fountain of liquid refrigerant came out. Looking closer, I found that I had brazed the capillary tube shut. Engineer at work.

I brazed the filter and everything else again, and now it works! The evaporator gets really cold, icing up. But the lowest I could get in the box was +8.2 °C. Not great, but the lid isn't sealed and the box insulation needs repair; that was just a dry run. I think I need a fan to move the air inside the box.

I screwed the compressor to its shelf. Only three mounting points can be reached, but I don't see a big problem with that. And I glassed in some blocks to hold the tubing.

A new galley base

In May 2023 I decided to redo the front of the galley, with a new drawer frame and shelves on the left. Everything had been mounted very loosely. I took the walls out and put the three back panels in again, glued, with new screws. The plan: new front panels and new drawers. The original drawers were tiny and not practical at all.

Here it is before the demolition, and then with the walls out. Prototyping the new front; a laser helps a lot. Cables organized. I turned one wall to make room for the drawers.

Then the front panels were fitted, and next came the new drawer module. I measured the countertop pieces: 58 and 68 inches. In 3D, the new drawer set: two 4-inch and two 6-inch drawers.

Then the front panel and the new drawer box went in, dry fit first, then glued. The new drawers use the whole cabinet; the original ones used only half. And I made "exact" countertop templates.

A door went in on the left side, with storage underneath.

The countertop

June brought the countertop dry fit, adjusting, then the final edge cut and trim.

The prep was done at home: the sink and storage openings routed on the CNC. Exact size for the sink, 12x12 inches for storage, and 12x16 for the refrigerator lid. The lids will need glued frames around them so they slide in snug. Then both pieces went to the boat, and strangely, every opening was in the right place! My biggest worry was the sink hitting the new drawers. Nope, all good.

I routed grooves so the wall cabinets sit tight on the countertop without trim, although I'll probably need some trim around the door and where it meets the next panel anyway. I glassed over some exposed foam in the bottom of the aft wall cabinet, in case of a leak. Yes, the epoxy is red: my fast hardener sat in its can too long. West System says it doesn't hurt the cured strength, so it'll be red until I use it up. And I cut out the gimbaled stove opening and made templates for joining the front and back countertop pieces.

Then the countertop got mounted, the sink installed, and all the inserts fit as planned. I made it so the plank pattern runs through without a break. Looks nice, doesn't it?

In late June the countertops and the upper cabinets went in for good, with a new faucet. A removable shelf over the pump compartment makes a nice spot for the trash bin and cleaning stuff.

In July: countertop inserts around the stove, everything sanded, and hinges on the storage lid.

Propane

The old propane line was several pieces joined together. I don't think that's a good idea for gas, so I put in one new line.

And the LPG locker drain, one more strange Irwin thing. The fiberglass tube is fine, but at the locker it was barely glassed, maybe just stuck on with 5200. Drop anything heavy in the locker and that joint breaks, and water, or propane if there's a leak, drains into the bilge. A Darwin Award candidate.

In the starboard cockpit locker I built a protective wall over the high-pressure refrigerator tubing, the condenser and the compressor, glassed on both sides to the structure and with one layer (if you can call it that) of 1708 to the hull. I painted the locker with TotalBoat bilge paint.

Then the LPG drain connection got epoxied and glassed with 4–5 layers, and the line was tied to the locker bulkhead so it can't move and break again. And a new LPG valve went in, mounted on a stainless bracket.

In June 2025 I installed an LPG alarm, and it alarmed all the time. I'm sure there's no propane on board. I had bought the sensor years ago, and it sat on the boat the whole time, probably "poisoned" by paint thinner fumes. I ordered a new one.

A fan, more foam, and 9 °C

Back in June 2023, I filled the gap around the ice chest with foam as much as I could and added a small fan inside the box. 7 °C in 3 hours, with a lowest of 5 °C. But the condenser was really hot, at least 60 °C. Probably it needs another fan too, for better efficiency.

Then I built the evaporator mount and shelf, mounted the air fan, added one more layer of insulation on top and foamed everything airtight.

In July 2023 the thermostat was mounted and wired: the small screen over the main switch panel. The refrigerator could reach and hold 9 °C with 38 °C outside.

Painting the cabinets

After the first coat of primer I saw the problem: plywood REALLY needs to be sealed before you paint it. So I sanded off that first coat and put on three coats of wood sealer, so the grain almost doesn't show. A lot of time, but after three coats it looks very impressive.

I felt like a squirrel in a wheel of priming and sanding. Five coats of primer in total. Sounds crazy? It is. First, Pre-Kote (the lesson: seal plywood first if you're going to paint it), and I sanded almost all of that off. Second and third, Interlux wood sealer, with sanding between. Sand again. Fourth and fifth, Pre-Kote again, sanding between. At 6 hours minimum drying per coat, that's a lot of time. And all of it in a Texas summer, 40 °C outside. The AC ran 100% of the time and kept 26–28 °C inside: nice for relaxing, a bit sweaty for working.

Now it's nice and smooth, with no wood grain. I'm still practicing roll-and-tip; I think my brush isn't soft enough. The final paint will be in October.

In August the headliner panels and a light went in over the galley. Not a perfect match, but there'll be trim anyway.

In November the last galley cabinet door went in, after two failed attempts.

A drawer on roller-blade wheels

In December 2023 I epoxied two slats along the long sides of the fridge opening to hold the lid.

Every Irwin Citation 34 owner knows this place: a strange, hard-to-use cavity under the starboard cockpit seat, opened from the galley. And there's nothing to stop things from rolling out of it right onto the stove. So I made a drawer. It came out pretty big, about 20 by 16 inches.

There are no straight walls to mount rails on, so I had to invent something. The drawer rides on roller-blade wheels that run on the countertop, and full-extension stainless rails stop it from moving sideways. It needs to come out about 1.5 times its length to really reach inside, so I joined two rails together to keep it straight.

A gimbaled microwave

My gimbaled stove came without an oven, so there was no counterweight to keep it level. I don't really need an oven, at least for now, so I put a microwave under the burners. I built a box the size of the stove top for a compact microwave I had. The gimbal swings about 30° one way and 35° the other. I hope that's enough. The burners are off in the photos because the test was done on a level boat, and they just fell out. One more 110 V outlet went in behind the stove, so no wires show.

A hood for the condenser

Since the first day I'd been thinking about making the refrigerator more efficient. The evaporator is in the ice box, the compressor under the LPG locker, and the condenser between the LPG locker and the starboard hull. A pretty "cozy" space with not much airflow, especially when the system is working hard to cool 20–30 lb of stuff.

First I hung a temporary fan between the exhaust hose and the locker wall. A week later I made a hood to guide the air and mounted a fan to pull air through the condenser. The fan runs whenever the compressor runs and dumps the heat into the hull, the LPG locker and the cockpit locker. After that, the condenser output came out barely warm. Before, it was 70–80 °C.

In March 2026 I measured the galley drawers.

In April 2026 I put a very hot coffee maker on the galley countertop and burned a ring into it. Using tung oil on that countertop was a great idea: with some elbow grease I sanded and re-oiled it almost without a trace. With varnish it would have been a disaster.

The household compressor dies at sea

Three years before, I had put the guts of a household refrigerator into my icebox. It worked surprisingly well through three years of lake sailing. A few days in Galveston Bay finished it off.

A household compressor has the motor and pump hanging inside the case on springs. At sea, especially in "washing machine" conditions (2-foot waves at a 3-second period), it got thrown around so much that one hot wire rubbed through its insulation and shorted to the case. The next time it got wet, the compressor died.

I always knew it wouldn't last forever on a boat. It was a 120 V compressor too, and the long-term plan had always been a proper 12 V marine/RV compressor. Well, the time had come.

Brazing on a fiberglass boat isn't the safest job, so this time I took lots of precautions: an extinguisher next to me, my son standing by with a water hose, a carbon felt welding blanket over the area, and a small flame deflector that keeps the torch heat on the tube instead of everywhere. The first time I had no deflector and only wet rags. Looking back, I was playing close to the edge. This time the brazing took about five minutes, calm and much more professional.

Then a 20-minute vacuum and a charge of exactly 1.8 oz of R-134a. Everything went as planned.

The box was loaded with about 20 bottles of beer and cider, so the first cool-down took a while. After about four hours of hard work, the compressor started cycling, holding 5 °C inside. All that with the compressor limited to 60 W. It's a BLDC motor and can run much faster, but I'm not sure there's any real benefit. Next time I'll let everything settle and measure the average power.

In August 2026 the compressor stopped switching on. After three years, the little relay contacts inside the cheap thermostat had failed, probably because it drives the coil of a 30 A relay that has no flyback diode. I replaced the thermostat. I still need to add the diode.

So I built my own thermostat

I went through three of those cheap Chinese thermostats in three years. I think the problem is the sensor and its circuit: moisture and vibration make them unreliable. The last one, after a few days, showed +26 °C when the box was cold and +2 °C when it was warm. Maybe it was interference from the AC wiring, but I got tired of this battle.

Welcome to my life. I designed my own: a 1-Wire digital temperature sensor, remote control, a link to the boat computer, a food safety alarm if the box stays above 7 °C for too long, and extra compressor protection. It's built around an ESP-12F module, a great fit for this, on a board made for a small IP68 box. Full user manuals are on the blog, and schematics, BOM and Gerbers are available on request.

What it does: it holds the box temperature, runs the fan in the box if there is one, and can drive a variable-speed DC compressor to boost cooling when needed. With extra sensors it can watch a freezer, the compressor, the condenser and the evaporator, and tell you when something's wrong. Any alarm can be turned off.

The main interface is as simple as it gets: one button with an RGB light in it, and a buzzer. On and off, mute alarms, and read the alarm code when there's a problem. Most of the time the light will just be blue. It has day and night modes, so it doesn't blind you at night.

Everything else is on a web page over Wi-Fi: set the temperature, assign sensors, set alarm temperatures, turn alarms on or off, compressor settings, and firmware updates, in Basic and Advanced modes. There's MQTT for OpenPlotter, and a serial port for a "nice" display, though I think that's overkill. The alarm codes:

• Minor, keeps cooling (red flash every 5 s): 11 door open, 22 cooler below 0 °C, 24 freezer below −25 °C, 25 extra sensor lost, 26 cooler above 10 °C, 27 freezer above −5 °C, 31 compressor hot, 32 condenser too hot, 35 evaporator below −30 °C.
• Food safety, keeps cooling (amber): 21 cooler warm for 30 minutes, 23 freezer warm for 30 minutes. Check the food, then double-click.
• Major, cooling stopped (steady red): 51 not cooling, 52 cooler sensor lost, 53 compressor overheated. Fix it, then power cycle.

A real lid

The refrigerator lid had zero insulation. It was just a plug for the hole in the countertop, and for three years after the conversion it was a piece of foam. I wanted better insulation, and a lid that closes airtight, so humid air doesn't build up ice on the evaporator.

I modeled two parts in 3D. An outer frame, glued into the opening, sealed around the insulation and epoxied to the walls, so the top insulation can't get wet. And an inner shell, nested inside the frame, that becomes the lid: filled with foam and glassed over, so the foam can't soak up water. A gasket goes between them.

I printed the parts only 0.8 mm thick, because they're just shapes to glass over. The frame needed trimming in place, because the boat is far from square and parallel. The lid shell got one layer of CSM on the bottom and the inside walls, and the frame got glassed on the outside.

Between boat days I painted the frame and the lid with topside paint, to make them white and seal the printed parts. Then I sealed the fridge foam with epoxy and installed the frame, and the lid was assembled.

November 2026: the lid is finished. The refrigerator project is officially complete. From an ice chest with a hole in it to a fridge with its own Wi-Fi.