[ Home Page ]    [ History ]    [ The Project ]

Project Log:  Friday, April 25, 2014

I'd ordered a little push button switch that I planned to use for the refrigerator compartment fan, which I'd be installing on the same wiring circuit as the refer but wanted switched independently.  Unfortunately, I discovered too late that the switch was momentary only, and unsuitable for the intended task, which would have been a setback since I really wanted to finish up the work in this compartment and move on.

Fortunately, I found a rocker switch in my inventory that I could use.  Normally I hate rocker switches because the rectangular holes are unnecessarily irritating to cut in tight places, and the fit of the switch itself is always fussy, and the switches are to bulky, but here it was a matter of getting the job done.  The switch would be within the cabinet anyway and not visible.

I installed the switch in the upper corner of the refrigerator cabinet where it'd be easily accessible, but out of sight.

Inside the narrow utility space next to the locker, I installed a 4" computer fan to draw additional air through the space and hopefully promote ventilation, and wired it as needed to the switch and the nearby terminal block for the refrigerator circuit.


I'd not yet installed catches on the door to the refer locker for whatever reason, so I took care of that next.  This was one of those projects that took 500% longer than it should have, thanks to an endless need to go up and down the latter to retrieve this tool or that fastener or some such. 

To provide a landing zone for the door catches, I needed a filler strip within the cabinet to bring the surface flush with the face frame, and the first piece I cut here didn't work, since I'd not paid attention to the depth of the latch body itself, so my first strip was too narrow, and I had to cut a wider one.  More trips.  But at long last, the task was done.


To secure the refrigerator in the cabinet, I wouldn't rely on the door, but needed some other means.  It was well restrained fore and aft, and against the hull, but could slide out through the opening.  There were probably any number of ways I could have done this, but in the end I chose a simple lashing, run between a pair of eye straps that I secured to the two bulkheads deep in the cabinet.



By now, midway through the day, the weather had warmed up, and with inclement weather forecast for over the weekend I knew I should get to work outside on the spars, where plenty of critical work remained.

I started with the new spreader bases, which I'd had fabricated earlier in the project, along with new airfoil spreaders.  The original tubular spreaders and bases had been sort of flimsy, and were just plain ugly and droopy. I hate droopy spreaders.

Installation was cheerfully straightforward.  I secured the two bases with a bolt through the spar (incorporating a spacer cylinder within the spar), which held them in position while I drilled and tapped for five machine screws in each base.  Then, I removed the bolt, cleaned out drill spoils, and reinstalled the bases with a light coat of Tef-Gel on the bonding surface of the base, plus on all the screw threads.



I'd erred in determining the length of the through bolt required.  The original bolt had been galled, and during removal I'd twisted the end off with the nut.  My new bolt was not long enough to allow the lower shroud tangs to be installed, and in fact didn't even have enough exposed threads for the nut that I used, so I ordered a longer one, and would complete that installation once it arrived.

For now, I mocked up the new spreaders to check them out, then removed them for storage for the moment.


I was thinking that before I got too far into the mast reassembly, I'd get the wire runs out of the way.  But as I was starting to look into that project, I noticed that the wind, which had been brisk all morning, had died, and with the relative calm I thought it would be a perfect opportunity to figure out where the new booms and goosenecks needed to go:  for this, I'd need to rig the sail temporarily, something I'd not wish to do on a windy day.

I "hoisted" the sail in its groove, and discovered an issue that I'd need to deal with:  the regular slot intended for slide installation was too short for the long slides at the forward ends of the full batten pockets (Harken Battcars).  There was another slot further down, where the original sliding gooseneck car would have been installed onto the track, and that worked to get these cars in the slot for now, but once I installed the fixed gooseneck, this would no longer be possible.  I'd need to enlarge the upper slot to allow these cars to fit.

Before continuing, I slid the tabernacle mockup into position and marked its top edge on the spar for reference, to ensure that my gooseneck remained above this point.


I secured the head of the sail near the masthead, leaving ample room for the headboard shackle, splice, and/or knot, then stretched the sail at the tack with another line.  Then, I secured the tack to the gooseneck pin, and let the gooseneck assembly fall where it may from there, making reference marks with tape on the mast.



After putting away the sail, I prepared the gooseneck for installation.  The fitting itself was a "universal" gooseneck, with a hinged design that allowed it to fit many spar sections.  I laid it out as needed and drilled and tapped for machine screws, before coating the back of the gooseneck in Tef-Gel and installing it permanently with ten of my favorite screw type, 1/4" cheese head (fillister) machine screws.



I repeated the process with the mizzen mast, which used the same gooseneck fitting.



By now it was growing late, but I wanted to at least get my snake run through the mast for the wiring.  My earlier, albeit brief, foray into this task had shown that it would be a challenge.  The mast incorporated a plastic wire conduit into the inside of the extruded groove on the aft side of the mast, a nice system because it would protect the wires and prohibit wire slapping. 

The problem was that the conduit ended short of the top of the mast, and with the interior welded sheave support structures at the top, plus a welded top cap, there was no direct access to the wire conduit:  just a pair of holes in the top cap.  I found I could peer into one hole and feed the snake through the other, but I wasn't able to get the snake into the conduit from the top.

From the bottom, it was easy to get the snake started, and up to the masthead, but then I had to find a a way to get it to come out the small hole at the cap--not an easy or natural thing.  So with some effort, and two flexible grabby tools, I managed to first direct the end of the snake towards the proper hole (with one flexible tool inserted from the side through a small hole intended for a sheave pin), then, with the end of the snake near the hole, use the second grabby tool to grab it from the top of the mast cap, and eventually--and with great ceremony--pull the snake through the top.


I left things as is for now, not wanting to attempt to fuss with wires this late in the day.  I had a feeling that getting the wires down and into the chase successfully from this small top hole was going to be its own challenge, best left for another time.  I needed one wire pair and a VHF cable at the masthead, and another wire pair to the midpoint of the mast (where a hole already opened up into the conduit). 

Total Time Today:  6 hours

< Previous | Next >

The Motorsailer Project
Site design and content ©2010-2015 by Timothy C. Lackey.  All rights reserved.

Please notify me of broken or missing links or other site issues.
You can always find every day's project log links on The Project page.

Questions and comments | Home Page
V1.0 went live on 8/26/10