CD 33 Refit: Embrace the Ω .

Best practices and electrical components have improved a lot since 1980. So for Sanderling, and the upgrades I was interested in, I totally embraced the Ohm. Some boat systems can be jury rigged. But the electrical system isn’t one of them, especially not if you plan to increase demand on the existing system.

Sanderling’s DC panel – typical ’80s

Surveys don’t include any significant analysis of wiring or the electronics those wires serve. So for any new owner, documenting the status of the existing system is always a good first step even if you don’t plan to expand. Sanderling had very basic components sparsely wired for DC circuits (note the 10 AWG wire off the common post serving all the breakers). The main hot bus is a solid copper wire, soldered in through the feed side of all breakers like a backbone. Typical 1980’s. She also had a jury rigged AC system using a residential sub-panel box that had turned to rust, probably because her original AC panel was damaged and removed.

DC panel feeder for Sanderling was 10 AWG – single copper wire loosely soldered in as hot bus

So for Sanderling the long term plan was to upgrade her systems from a 1980’s weekender to a cruiser with a few amenities and room for expansion. For me that did not mean frozen ice cream and HD TV for a week at anchor without running the engine. It meant being able to run pumps, electronics, or even the autopilot for certain periods if I chose not to run the engine (or couldn’t run the engine). I created a spreadsheet to actually anticipate what kind of power I would need for sailing, motoring, or anchoring. That helped me estimate the size of the battery bank I would need, and also the kind of charging I would need to keep the batteries happy.

adding up loads and Amp Hours needed

Sailing and motoring needs differ depending on variables like how often you manually steer vs. using an autopilot, alternator output, etc,. After figuring out what I wanted the boat to be able to do in various scenarios, I estimated the amp-hours each scenario would require over a 24 hr period. Then I subtracted any amp-hours that might theoretically be replaced in that same 24 hr period from charging sources during each scenario (surplus alternator amps, solar panels, wind, etc,.)

the overall design for Sanderling – large house bank, dedicated start bank, high-amp fuses, sub panel

The sailing scenario above (230 Ah) would require the most from the house batteries. If you use Calder’s 50-80 rule (never go below 50% state of charge, and charge to 80% in regular use) then I would need a house bank with total capacity of 650 Ah to actually get a usable 200. But space and cost were factors, so I compromised by installing four golf cart batteries (6 volt) that would produce a bank with total capacity of 460 Ah. To get my 200 usable amp hours for the most demanding scenario, I’d have to either charge to 100% (time consuming), install a solar panel to offset consumption, or ration the 138 Ah I would have if the house bank was only at 80% charge.

DC Bus bars and terminal block (solder gun to remove sketchy old soldered connections)
  • Four 6 volt deep cycle in series/parallel to create a 12v house bank with 460 Ah
  • Dedicated crank battery (750)
  • Pure sine inverter (isolated circuit for laptops, or any sensitive tech)
  • Modified sine inverter for miscellaneous (undiscriminating) devices
  • Rewired and labelled DC panel busses and connections
  • 2nd DC panel for expansion
  • Custom AC panel in new location
  • DC sub panel at Nav station (with alarm panel)
  • DC outlets and USB charge ports at Nav Station & cabins
  • High Amp fuse bus
In my CD 33 I chose to raise the quarter berth 3.5 inches to accommodate the height of the golf cart batteries in custom fiberglass box

For the upgrade every connection and circuit was replaced, except the original circuit for the cabin lights (which would now serve low power LEDs) and the existing galley/cabin GFCI AC outlet branch circuit (which was a still healthy “direct burial” Romex 12-2 with ground). I fished wire in places I did not know existed on a Cape Dory! But the panel design and bus wiring was done through the winter on the bench in a toasty warm workshop.

new teak plywood cabinet face, main and aux DC panel cut-outs

I found a source on eBay for 1/8 inch aluminum plate that had a nut-brown finish on one side that looks good with the CD interiors. So I decided to just make my own panels from scratch (I also could not afford Blue Seas products). It’s not difficult to make a jig out of hardwood so you can drill out all your holes in consistent rows. With a special multi-step drill bit you have all the sizes you need in one bit, and the aluminum is easy to work with.

Carling technologies sells replacement toggles (match toggle, wire, and breaker specs to the load)

Updating your electrical system doesn’t mean replacing it entirely ($$$). These original “aviation” toggles used by Cape Dory are rated for 6A @12 VDC (ok for small loads). But the older ones have 16 AWG pigtails (only rated a mere 10A if the circuit is*under* 6ft length). That doesn’t meet todays standards if the switch follows a breaker rated higher ( 15, or 20 Amps). Cape Dory seems to have just used the same low amp toggle for all circuits regardless of the rating of the breaker, even for AC circuits. And depending on who makes the switch, the AC rating and DC rating may not be identical. I have seen these older toggles heat up considerably on AC panels with loads that did not trip the breaker.

So the toggles are fine to replace as long as you purchase a quality switch rated *at or above* the 12 VDC amp rating (or 125 VAC rating) of each particular circuit’s over current protection. Use the same gauge wire for the entire circuit, and get replacement toggles with ring terminals instead of solder or blade connectors. The same applies to replacing a push-button breaker, you can buy them with ring terminals instead of blade.

Best practice these days is also to fuse every circuit at the source. There is an ABYC exception for “starter circuits”, so you can still forego the fuse at any battery that might be used to start with, but it has become common practice to have a high-amp slow blow fuse at each battery bank.

For making your own busses you can buy tin coated 110 copper bar online. You can lookup the conductivity ratings so you know you have a thick enough bar. For my purposes, a 1/4″ thick by 1 -1/2″ wide bar exceeded the combined amperage loads of the fused circuits it would serve.

The completed high amp fuse block is fed by the common pole of the selector switch. The fuses off this block feed the DC panel hot busses, the starter circuit, and the inverter, with room for expansion.

custom high amp fuse block

If you decide to keep using the toggle switches, you’ll need a terminal block to organize the load side (from toggle to actual circuit). The line sides of the push-button breakers are all connected directly to the hot bus. That is one advantage of replacing all the push button/toggles with magnetic breakers instead (if you have the $). Because in that installation the hot bus is integral to the breakers on the panel, and you could eliminate the intermediary terminal block shown in my installation above.

custom panel for additional DC circuits – including indicators for voltage at each bank

But also consider that a terminal block is still good for organizing, and stand-alone hot busses can be useful for expansion. Here’s the additional DC panel (above right) using magnetic breakers. The pumps and sub-panel main were all put there using magnetic breakers, keeping the boat’s low-amp circuits on the original push-button panel. The two digital voltmeters are wired to show voltage for each bank measured at the battery terminals (not off the selector switch).

circuits all labelled with clear shrink tube
Hybrid solution (added second DC panel with magnetic breakers for high-load circuits, replaced some older taggles, eventually will replace all toggles and probably replace selector switch.

CD 33 Refit: Get some sole

The cockpit sole was soft when I purchased Sanderling, but the side decks were solid. Soft side decks would have been a deal-breaker. But a cockpit sole is not too difficult to rebuild especially if you have already removed your rusty steering bracket to be rebuilt (see the Clunk post). I was certainly not going to install my new steering bracket under this cockpit sole and see it quickly destroyed by a leaking pedestal.

top layer lifted off without any effort, there was nothing holding anything together

Sanderling had been on the hard covered for 7 years. I needed to determine how far the core rot had progressed. My tapping skills proved pretty good. I had estimated that from the aft end to about a foot before the bridge deck the core was punky. I did my cuts and removed the top layer. The plywood and balsa core around the pedestal was just dust, and the balsa towards the bridge deck was rotted and still holding water.

top layer off
underside of removed cockpit sole

I used a simple circular saw and common blade to cut just enough depth to penetrate the top layer of glass over the core. I used a piece of square profile wood trim as a straight edge in between the saw and the locker side walls. Where the blade fell short of completing the cut in the corners (too cramped) I finished the cut with a cutting wheel. I got the top layer off in two pieces.

cockpit revealed
the top layer came right off, nothing holding anything together

There was a seam of epoxy within the core about a foot from the bridge deck. So my tap guesstimate of the rot lined up with that. But water had made it beyond, that seam just made things sound solid while tapping from above. So I made a second cut removing the rest of that area of top layer. The balsa in that second piece was still adhered to the top layer, but as I pulled it up a thin layer of the moisture weakened bottom layer came up with it.

I chose not to re-use the top layer. It was tired and warped by water intrusion, with delamination at the pedestal bolt areas. The new sole would be synthetic core with layers of chopped strand and bi-axial glass.

cross section plan of layers – not to scale

The original top layer of the cockpit was a full 1/4 inch thick. I had probably an 1/8 of an inch sag in the bottom layer from age, and the Divinycell (closed cell) core material was not as thick as the original balsa. So after the core was glued in first onto the bottom layer I’d need to build up almost one half inch of glass in order for the whole thing to be flush with the original sole. That worked out well as I’d rather have a very stiff and solid sole since it’s a high traffic area.

cockpit sole revealed
old balsa core removed, bottom layer remains, styro core cut into strips to act as cored edge flange

There was no way to cut an oversized continuous piece of core that would fit into the 1 1/2 inch edge cavity. And I did not want to leave it as a void. So I cut strips out of foam core with double taper and glued them in first using epoxy with filler.

This would stiffen the edges and give the big piece of core a 45 bevel to sit on and glue to. You can see the step I made with the router. Not necessary to make the edge pretty since after epoxy, fairing compound, sanding and paint layers you won’t be able to tell anything happened.

emerg tiller port
patch job by builders on misaligned tiller port

I also noticed from below that original builders had drilled out the hole for the emergency tiller in the wrong spot. Looked like they moved the hole back an inch and filled in the missing bit with a kind of crescent moon shaped plug. So when I glassed in the new sole top layer I was able to cut a fresh new hole out for the emergency tiller hatch in the correct spot. (tip – leave a bit of wiggle room with the hole and dry fit your emergency tiller hatch & tiller to ensure proper position)

foam core (Divinycell)

Bottom layer of cockpit sole wetted out with epoxy first. Single piece of foam core with reverse 45 bevel, sitting on the previously glued in 45 bevel edge strips. Weights applied to press core down while setting up. (tip-put plastic below to catch drips) The position of holes for the new core were marked from below by tracing the position of the existing bottom layer holes. The two extra holes seen here on either side are recessed cavities in the core for the heads of the bolts that hold the first two sheaves in place below (going through the steering bracket). The third extra hole serves the same purpose for the quadrant stop, also bolted in to the bracket below. The layers of glass just covered them up but when the steering bracket was installed it fit flush because those bolt-heads could tuck up into those recessed areas.

routered out a step in the edge of the cut – so this first piece of bi-axial put down covered the core, the

After routering out a 1/8 inch step in the edge of the top layer, I could put the first layer of bi-axial down and it would cover the core, the seam of the 45 bevel edge core I put in, and at least 1/4 inch of the original top glass layer (see illustration above)

first layers of bi-axial tie everything together at the edges

The larger holes were visible enough through each new layer that I could find their center and use a hole saw. But the pedestal bolt holes were hard to see. I didn’t want to have to drill out the holes from below. So I got some short pieces of 1/2 inch PEX and their actual OD was perfect for leaving room for the bolt and bedding. I just stuck them into the bottom layer pedestal bolt holes and built the glass up around them. This kept the holes exactly where they needed to be.

The bottom layer of the cockpit sole sloped a bit towards the bow, so as I built up layers the aft end was rising faster by an 1/8 of an inch or more. So I used fairing compound layers in between, and in the end had to use something bulkier, so I used up some remaining woven cloth. I built a squeegie out of a 2×4 and a piece of PVC trim that was the width of the cockpit, and used that to check the build-up, and to scrape the final fairing layer even with the edges.

some waves on 2nd layer of fairing – but after sanding each layer it all evened off. I also spot fixed certain low areas with a smaller applicator before doing next layer.
about six layers of fiberglass, layers of fairing in between (top fairing layer pretty thick)
a 2×4 with PVC edge used for squeegee – made sure everything was level with original edges
the new cockpit sole is an inch thick (1/2 in of core, 1/2 in of glass)
primekote two-part primer (Interlux) two layers
Interlux Brightside single part

The color of my non-skids was somewhere in between the neutral look of Grand Banks Beige and the warmer look of Bristol Beige, so I bought a can of each and mixed them.

building up layers of Brightside (first layer sanded well before second layer)
Too coffee colored, added more Bristol Beige to the mix for the next layers
fall evenings getting cooler, heat lamps to help curing. Five layers (single part Interlux needs buildup)
final color 75/25 mix favoring Bristol Beige – a perfect match!
for non-slip I used Intergrip powder, broadcast on wet layers and rolled on with impregnated paint

I actually lightly sanded my layers of broadcast non-skid after drying, because I did not want an abrasive surface that stood out against the worn nonskid on the rest of the boat. Then I switched to impregnating the paint with the nonskid and rolling on with a hard mottle roller, that worked great and helped move and evenly distribute the nonskid (the brush was no longer useful on a rough surface). I did two coats of paint over the two-part primer first, then build up about 3 or 4 more layers of paint and light nonskid. The results are much better than I expected, not having done any nonskid before. It looks good without making the rest of the worn out nonskid in other places look bad.

CD 33 Refit: Clunk!

I was moving the rudder back and forth one day (on the hard) to check the swing, when I heard a CLUNK below the cockpit sole. I looked around and found this little beauty. This is the rudder stop that was welded to the steering bracket. It stops the quadrant from turning too far either side.

rudder stop
rudder stop

Cape Dory builders made use of generic steel in certain internal areas. The steering bracket and the horizontal backings for shroud pad-eyes are two such areas. When surveying a Cape Dory, these are two places to check first for corrosion. My chainplate systems were in good shape when I surveyed Sanderling for my own purchase, and still are. Just a little surface rust in a few spots.

looking up at the hull-deck joint area port side

But the steering bracket was corroded even when I purchase her. So the time to deal with it had finally come with the rudder stop going CLUNK! Tied to this project was a pedestal renovation and cockpit sole rebuild, since the long standing water intrusion also got into the sole core (see the post Get Some Sole).

pedestal leaking for many years – serious corrosion (bronze quadrant and ss wires spared)

The stock metal used in these areas was only covered with a slim coating of what appears to be primer by Cape Dory builders. The leak in this area was due to either poorly bedded pedestal at build (others have suffered this condition), or movement of the pedestal over time, which compromised the caulking. Lots of water came through here. There was most likely galvanic corrosion also.

cable guides
cables coming out of pedestal – not much left of backing plate

The pedestal backing plate on these Cape Dory boats came with the Edson steering solution kit they used, and Edson calls it the cross-wire idler. It was a fairly thin piece of cheap metal in the 1980 version, and was actually a weaker link in this system than the original metal Cape Dory used to fabricate the rest of the overall steering support system under the cockpit sole.

used idler plate
what the original Edson idler plate looks like – (pic from someone selling a used one on the internet)

I do not believe Edson has any of these older idler plates. They pointed me to the new versions when I inquired. They are $459 (not cheap) and made of aluminum. Edson was only interested in selling me an entirely new system. But on my salary there was no way I was going to toss out my bronze sheaves for costly aluminum ones. They were still in great shape.

When I got my rotted steering bracket out it looked like Cape Dory builders had just welded the Edson provided idler plate onto their yard-fabricated bracket that holds three things together ( pedestal idler, the four sheaves, and the rudder stop) . You can see below the original Edson idler plate attached to my CD fabricated steering bracket was almost entirely GONE.

rotted backing plate

The two outboard sheaves were still OK. So I had a local fabricator take the entire thing and cut away all the bad stuff, and he made a replica using stock metal for about $400 (New England area price). Down south you can probably get it done for half that.

new steering bracket

I had him embed two stainless grommets as primary wire-guides that were machined to take retaining rings, so the sheaves would mount on them as before and swivel to adjust, but could be removed if needed. On the original Edson idler these were press-fit ss grommets that allowed the sheaves to swivel but made them permanently fixed to the plate. I also called for a 1/4 inch thick idler plate, and a rudder stop that would be removable for doing maintenance on the quadrant area.

there was enough room to double-up the clip rings on each, so it was snug but could still swivel

I did not get fancy with machining a swing arc for the hold-down bolts like the original Edson plate had. There was no need since it was not going to be critical to continually adjust the positions of these first two sheaves. And it was easy to put the bracket in and out via the port lazarette. So I installed the sheaves on the bracket, aligned all the sheaves and wires, marked the positions of these first two sheaves, then removed the bracket and drilled the thru-bolt holes needed.

bracket near outboard sheaves was OK -so that part of the setup didn’t change

That long adjuster arc for the tie-down bolt is really only necessary if you are selling a system to work on lots of different boats (or building a new boat) and don’t know where exactly your pedestal, sheaves, and quadrant will be in relation to each other. On a boat already built, those are already fixed and won’t change. I found that since I could still fully adjust the second set of sheaves (no change to them) that proved to be more than enough fine-tuning for the wire leads where they finally met the quadrant.

two coats of two-part primer

I did not feel this needed to be stainless steel fabricated either. That would have been a much higher cost and difficult to drill, modify or tweak. I gave it 2 coats of epoxy two-part primer. With proper bedding of the pedestal and leak mitigation it can last indefinitely.

first set of sheaves – clearance to swivel but will be through-bolted to exact position needed.

And obviously the pedestal had to come off to do all this. Forty year old aluminum Edson bolts are pretty much ready to break….especially with years of continuous water intrusion.

this one was easy…..

…but other times you gotta fight with them. If they are aluminum, which is soft, using the correct type and size of bit you can can drill the head off cleanly from above without damaging the pedestal flange. But it will not be fun if those old bolts are stainless steel. You’ll need to go at those from below, and may need to cut the nuts off with a specialized cutter or even a torch.

aluminum Edson bolts – head drills off easily.

The pedestal guard (the cockpit oh-shit handle) has these two feet to hold it in place, which were through-bolted (piercing through the old steering bracket below). I considered them to be a huge potential for leakage onto my nice new bracket, so when I redid the cockpit sole I made it thicker overall, with a synthetic core, and solid glass in those footer bolt areas, so I could get the holding power I needed for these guard feet using lag screws instead of through-bolts.

total boat single part flat finish

The Edson pedestal was in good shape so I serviced it by adding new nylon shims for the control arms (from Edson), new ss bolts, and a new paint job.

Total Boat (Jamestown Dist) white single part with flattening agent – no glossy shine.
Sending out my compass to be rebuilt by vikingoptics ($300 for complete rebuild)

CD 33 Refit: Bottoms Up!

Forty years of new season paint applied over last years scaling paint can make quite a mess, so one of the first projects was to remove the old paint, apply a barrier coat, and start over with fresh bottom paint. It does not make sense to remove the old paint and *not* apply an epoxy barrier coat as prevention. But I know people who have skipped that part. In my opinion it is one of the best sweat equity projects you can do on a fiberglass hull, as it increases the boats value and doesn’t break the bank on materials.

pic of scaling old bottom paint
old paint just falling off (left side CD 24 Trawler, right side CD 33)

On boats where the scaling paint is just falling off in large pieces, like on Sanderling, it is relatively easy to scrape the bottom clean down to the gel-coat with my tool of choice, a scraper. But my ability to do this was dependent upon the flaking-off condition of the paint, and because I did not mind leaving her on the hard drying out for several seasons. I worked for a sailing school and was out on those boats enjoying the bay when I was not scraping.

Scraper
2 sided hardened steel blade (change blades often)

If the bottom paint had been old but still sticking to the hull I would just clean, sand, and apply new paint. But with the cost of bottom paint, I can’t justify applying $120 a gallon paint to something I know it won’t stick to for even a week. I’m just feeding the fish in that case, and risking water intrusion past the gelcoat into the hull laminate.

previous refit of CD 24 Trawler – scraping bottom

For anyone who finds evidence of blistering, bulging, or de-lamination of the fiberglass, those owners may need to forcibly remove even the stubborn layers of paint in order to mitigate or diagnose. Those projects require a peeler, blaster, or solvent stripper solution. Lucky for me both my Cape Dory boats were perfect candidates for the scraping method. With these conditions (dry and scaling), all layers just come right off with one swipe of the scraper, held at just the right angle and pressure. But it is serious aerobic work. I was held up in Annapolis once on a delivery waiting for parts, and a van drove up to a 40 footer in the marina yard, four guys got out, and in two casual days they stripped the bottom clean.

kept this garb and used in when covid hit

The gel coat applied to Cape Dory boats is pretty thick both above and below, as well as the layup of the hull. But I noticed that the condition of the below waterline gelcoat on Sanderling was much poorer than that of my previous 1984 CD 24 Trawler Viola, which had been of similar age. The gel coat on the 1980 CD 33 was brittle with small divots falling out even with gentle scraping, whereas the CD 24 Trawler gel coat was very hard and smooth.

pitted gel coat after scraping – will be sealed with Interprotect 2000 system and then filled with fairing

I attribute that to build variables, or infrequent use of cheap and poorly applied bottom paint on the 33. Neither of my Cape Dory boats revealed voids or blisters once stripped, although the 24 Trawler had a couple of 2 inch round spots where the gel coat had not adhered fully to (or had separated from) the hull – mostly likely due to an air pocket when layup was applied to gel-coat. But it was solid fiberglass behind that and with zero trapped moisture. I’ve never seen a true blister on a Cape Dory hull but exposing the gel coat and fully inspecting/tapping the hull will tell you exactly what is going on (better than a moisture meter in my opinion). The sound of a void, saturated, or delaminated fiberglass is distinctly different.

Progress!

Build standards were not an exact science during the recreation boating boom of the 80’s, and this is hull # 8 so she is an early one. The original owner told me she came complete in 1980 with a small 1 degree list to port which they tried to solve with lead ingots . It never worked so she still had the small list when I purchased her. But that’s nothing a little re-arranging of fresh water storage, provisions, or crew can’t fix!

scraping complete, ready for sanding

There was also a strange layer over the gel coat, under the years of paint, which seemed like a polyester resin layer. It was yellowed and in some places very adhered to the gel coat, and other places came right off. This may have been intentional at Cape Dory, or just remnants of the mold release layer the gel goat was sprayed on to. I did not find that yellow layer on my previous Cape Dory Trawler.

barrier complete
first layer of Interprotect 2000 barrier coat

The Interprotect 2000 two part epoxy system worked great for my 24 Trawler, so I used the same system. I have since begun using the Interlux two part primers for any other job that requires a good primer base (emergency tiller protection, wood priming, etc. ) Interprotect as a barrier coat on the hull can usually be built up to the required manufacturer’s thickness with three layers, so that’s what I budgeted for.

fairing
two part fairing compound over entire hull after second coat of Interprotect

These layers were alternated blue/grey/blue. The fairing goes on after the second coat, before the last epoxy layer. I used a two-part fairing compound (Total Boat from Jamestown Dist). Per Interlux, the layers of barrier coat are cured before the next goes on, and I lightly sand off the tits and wipe the hull down in between layers with mild TSP and then rinse and dry.

crysis mixer
mixing epoxy – for CD 33 it came out to 1 1/3 gal of product per coat (4 gal total)

The final layer of barrier coat and the first layer of bottom paint need to be fused. That’s called hot-coating the first layer of bottom paint. This requires doing the last barrier and first bottom coat both on the same day. While the last coat of epoxy is still tacky to the touch, you start rolling on your bottom paint (you need to work fast if the ambient temp is 75 or above otherwise the last barrier coat dries before you get all the first coat bottom paint on) . I have found ablative bottom paint to work well and not break the bank, so I went with ACT.

finished bottom
ready for an early splash in April 2021

The fairing compound was an extra step but I’m glad I did that, the bottom is very smooth and consistent, will be easier to clean mid-season, and might even give me fraction of a knot more of speed while I dodge the foil boats in Newport Harbor out to break the sound barrier!

minion #1
thanks to my #1 minion (helped me avoid a real life cry-sis)

The Bigger Boat

1980 Cape Dory 33 Sanderling

I purchased Sanderling in 2011 from the original family who had ordered her from Cape Dory in 1980.

Aside from the fact that she was a Cape Dory there were a few other things I liked. Her auxiliary engine had been updated with a Yanmar diesel, which I had experience with and preferred over the original Perkins and Volvos that these boats came with. Otherwise she was very much original, without a lot of additions or modifications. And she had been well cared for.

splash for sea trial and purchase

Since I had experience as a delivery captain, sailing instructor, and merchant mariner, I knew what to expect in a used recreational boat. I also new Cape Dory boats pretty well. I did my own survey and knew where to look and found pretty much what I expected.

Monogrammed Plates

The sellers of the 33 in Oxford (Stan & Mary) were wonderful, it was an emotional closing as Sanderling had been a huge part of their family for decades. They left everything they had ever used for boating onboard, including wonderful monogrammed plastic dinner plates:

I’ve you’ve never been to the Eastern shore of Maryland, it is exquisite. Coming from Rhode Island I was enthralled, since the Chesapeake is very much like Narragansett Bay only much, much BIGGER!

Stan and me, Mary taking pic

Previously I owned and refitted a Cape Dory 24 Trawler (see post The First Real Boat). I had Viola for many years and it was a sad day when I sold her. But I was also pretty stoked to trade up to this bigger boat that could potentially sail anywhere or be used as a modest live-aboard (…it has a shower!). Below is my proud new owner face. Sanderling and I sailed from the Chesapeake to her new home of Narragansett Bay without a single hitch.

pic of smiling new owner
proud new owner of Sanderling CD 33 Sloop

I picked up some crew at Statten Island, NY. My friend Steffen needed to get out of Boston for a bit.

Steffen on the high seas

Though my dreams included blue-water sailing, I was pleasantly surprised at how much I enjoyed sailing my new Cape Dory 33 in light wind. On a fabulous June day in 2011 on my way home from the purchase in Maryland, Steffen & I had 5 kts of SW wind in Long Island Sound, and gently moved along at 3 kts wing and wing (full main and 120 Genoa).

view of sails wing and wing
wing and wing northeast bound long island sound 3 kts

As we entered Block Island Sound the late afternoon sun penetrated the first 10 feet of water with dancing golden rays. We glided through an area where dogfish were swimming upwards toward the surface. The visibility of the surface water was crystal clear. You could see their subtle green and black coloring vividly. It was one of those oceanic moments (they looked like miniature whale sharks from above in that clear water). The afternoon sun also beamed through the companionway and lit up the teak & holly sole.

cabin sole afternoon sun
afternoon sun

I sailed Sanderling on Narragansett Bay for the 2011 season. She weathered Irene in August on a mooring in Wickford, with just a little bit of Easterly exposure which abated after Irene’s rotation tracked to bring the wind South, where Sanderling was fully protected. No damage just a little chaffing on the rub rail.

pic of sanderling on the hard
snug in my yard

She’s been on the hard since 2012 getting what I like to refer to as a Galapagos refit: Slow like evolution. I only had the patience because I worked as a delivery captain, taught sailing, and crewed on commercial boats. Those activities kept me out on the water on other people’s boats most of the time. So small projects turned into bigger projects. But not feeling pressured made it work. I grew to really enjoy losing all track of time and then having Amber walk over to the boat and knock on the hull and say “let’s eat”, or one of my nieces climb up the steps late-night poking her head into the cockpit to say “whatcha doin?”. To which I would inevitably reply, “Hand me that wrench next to you”.

Amber keeping me company on a delivery in 2014

Several other posts detail the main projects of the Sanderling refit. Cockpit sole, Steering bracket, Barrier coat . But ten years and a pandemic is enough, and this boat is going in this season no matter what. If you can’t fix it with duct tape… you’re not using enough duct tape.