Meet Skylark

After my co-owners decided to pull the trigger on their cruising plans, we decided to sell our faithful bay sailer Joyous, a Catalina 30 and I started looking for my next boat. I plan to spend some time cruising in my future too, so I decided to look for a proven blue water design.

After months of searching I found Skylark right here in the SF Bay Area. She is a Morgan 382, hull #64, built in 1978. These boats have an ardent following, an amazing community of owners and had been at the top of my list of candidates.

She is in good condition for her age and the survey didn’t reveal any unexpected issues. However most of her systems are original and now pushing 42 years old (including the standing rigging)! I’ve definitely got my work cut out re-fitting and bringing her back to her full glory.

We had a tense but elated motor from the boatyard over to her new home. We’d found a small coolant leak from the fresh water pump during the survey, but when I went to pick the boat up from getting the bottom painted the drip tray was full of coolant. The amazing staff at the yard found us some spare coolant and after running the engine for a while we concluded it must still be a very slow leak – no dripping noticed over 15 minutes. In the end we made it home uneventfully.

Now on to planning the re-fit. Top of the list right now is fixing the coolant leak and replacing the flaky 1978 battery charger. Next up will be lifelines and standing rigging. I’m currently looking into switching to synthetic Dyneema for both lifelines and the rig.

Toyota 4Runner LED Turn Signals

I had a turn signal out on my 2007 Toyota 4Runner Limited V8 and thought this would be a great time to upgrade to LED turn signals. I swapped in these:

They seem well built and are significantly brighter than the stock incandescent bulbs.

However, due to the lower current draw of the LED lights, the flasher module in the vehicle thinks the bulb is out and does a fast blink or “hyper flash”. A common fix is to simply put a high current resistor in parallel with the LED bulbs to increase the current draw but this didn’t seem like an ideal solution for a few reasons:

  • More wiring & components = More points of failure
  • More janky stuff hanging around under the hood
  • Wastes the additional power as heat (not a big deal really, but inelegant)
  • Probably prevents detection and hyper flashing if the LED bulb blows

I did find that aftermarket flasher modules exist that are designed to work with LEDs but as my 4Runner has Daytime Running Lights (DRL), it uses the “12-pin” flasher module which has no LED version available as a replacement.

Luckily I did find a simple mod to the existing module that fixes the problem. The module contains a current sensing shunt (looks like a wire jumper). By swapping out the shunt with a higher resistance, the target current range that the module identifies as a working bulb can be decreased.

Shunt Calculations

In the stock configuration, the nominal voltage drop on the shunt for one side flashing is 75mV [1] (twice this for hazards, i.e. both sides flashing). With the stock bulbs this corresponds to a stock shunt resistance of about 20mΩ.

My new LED bulbs are rated at 6W for the front and 5.5W for the rear, for 11.5W total. This corresponds to a current draw of 0.83A @ 13.8V. To produce the same 0.075V drop in the shunt, I would need a resistance of 63mΩ.

I decided to try out a 70mΩ resistor as a more common value. To calculate the power dissipated in the shunt itself we can use P = IV = 0.83 * 0.075 = 0.062 W. I got some completely overkill 3W ones just in case:


Flasher Module Location

Its a little fiddly to remove as it locks in place onto the top of the fuse box. It can be released by inserting a small screwdriver into the latch from the same end as the connector to release the tab.

Replacing the Shunt Resistor

The case of the module can be released with 4 tabs around the connector faceplate. Its fiddly but be patient and don’t force it, I used various shims to hold each tab open after releasing them one at a time.

The shunt resistor is the metal loop marked “RS”. Simply remove by desoldering one end at a time. You can see the underside of the pads in the photo near C4. After this, I cleaned up the pads with some solder wick to make installing the new resistor easier.

New resistor installed (0.070R).


After reinstalling I verified that the lights were flashing at the normal rate on both sides, and with the hazards. I then tested to see that hyper-flashing would return if one of the bulbs was disconnected, which it did.

Overall this was a simple fix that is robust, preserves hyper-flashing functionality to detect blown bulbs, and doesn’t waste excessive power.

Started riveting T03-05

Today I managed to squeeze in half an hour after work and started riveting the horizontal stabilizer forward spar assembly.

Mostly smooth sailing until I had to use my ground down angled rivet nose to get the rivets close to the edge. I had another few stems break higher than normal and one was actually proud of the rivet head. They all seemed to seat correctly on the blind side though so I’m not sure if this is a problem. I also noticed that it can sometimes take a couple pulls of the rivet gun to break the stem whereas in the normal configuration it almost always breaks first time.

I’m wondering if the larger diameter hole of this nose piece is a problem and will probably see if I can get some extra nose pieces from Harbor Freight in the correct diameter to modify. The larger hole also seems to mar the rivet heads more although I have found even the unmodified nose presses a slight flat on the rivet head which I haven’t seen on other people’s photos. Still trying to get a feel for what the range of acceptability is here and I haven’t been able to turn up much concrete guidance on the web.

Time: 0.5 hours

First finished assembly

Today I finished the T03-01 horizontal stabilizer main spar assembly. I started by trying to improve my tools for riveting in tight spaces. I ground down the spare (slightly larger) tip of the rivet gun to have an angle and also ground down some of the nose piece. The EAA has some helpful videos on this topic:

At first I ground the angle too steep and found the rivet stem would break off too high, at the point where if bent round into the nose of the gun. With a shallower angle I got more consistent results. There were still a couple rivets where the stem broke above the normal point but still below the top of the head and the shop side looked well set. I’m still at the stage where I’m worrying every little detail is going to kill me so I’m planning on getting some feedback on the rivets and what is acceptable.

IMG_0994  IMG_1004IMG_1005

I also noticed that sometimes the stem wouldn’t be pulled as deep into the rivet. This didn’t seem to correlate with the ones where the stem broke too high though. Again, I plan to get some advice on whether this is acceptable.


I also had time to up-drill and deburr the forward spar assembly (T03-05). This went much faster than the main spar now I had most of the techniques down. I found the blue surface conditioning discs in the die grinder made quick work of the hole deburring. I started with the vixen file to get the machining marks out of the spar but found the Scotchbrite wheel or the maroon pads did as good a job with less effort and less finishing work. I used a needle file and Scotchbrite pads on the detail areas. One note on the Scotchbrite wheel – I was initially trying to avoid the parts cutting a groove in the wheel by running them at an angle. I found the groove is actually very beneficial and lets you finish the face and corners of long straight edges in one pass.

IMG_1012 IMG_1016

Time: 7 hours

First rivets

Today I pulled the my first rivets! The Sonex is built from 6061 alloy which doesn’t need to be primed for corrosion protection, however the blind rivets used are made from stainless steel which raises the question of galvanic corrosion. Sonex say they have never seen a problem with galvanic corrosion but because I live by the coast I decided to add extra protection and use Duralac, a jointing compound that is designed to prevent galvanic corrosion, on each rivet. This product contains Barium Chromate (so wear gloves) and is popular for riveting stainless fittings to aluminium boat masts. Its also been recommended by a number of other Sonex builders.

This builder has posted some tests that show it is effective:


I just put a thin coat on each rivet before inserting. These rivets heads are going to the polished so I tired to avoid getting too much around the head but in the end it was easier to just wipe off the excess with some mineral spirits than try to be too precise applying it.


Some rivets on this part are very close to the edge, too close to get the pneumatic rivet gun in there. I’ve seen that people bent the stem so you can get the gun in at an angle and then used a wedge of material to help the rivet set properly. One half of my simple dimple die seemed like it would do a good job as a wedge but about one in three rivets would set with the head not flush with the part. Next steps are to improve my tools for riveting close to the edge, drill out the bad rivets and finsh riveting the spar.

Time: 2 hours

Surface prep discs

Today my friend Niels came over and we cleco’d together the forward spar. I received some very fine (blue) and super fine (white) surface prep discs from Pan American Tool and we tried blending out the scratches from using the medium (maroon) discs. They were pretty deep and I was worried about removing too much material blending them out so I decided not to polish before riveting and just did a pass with the blue and then the white discs.

In the future I’m going to stick with the blue ones for deburring and only use the maroon ones for deep machining marks. The white ones are overkill for deburring but probably great for fine blending before polishing.

Time: 1 hour

Starting Building


Over the past few months I have been slowly clearing out space in the basement below my office in San Francisco. It looks like it used to be an artist’s studio with multi-colour paint spatters on the floor. I’m really excited to have such a great space to begin my build – the only downside is that the only entrances involve a narrow doorways and steep steps. I’m pretty sure I can get the wings out of there but the fuselage is hopeless. Right now my plan is to build the major fuselage sub-assemblies there and do the final assembly in the hangar.



I’ve slowly been amassing the tools to start my build. I picked up a compressor from Craigslist, its a 30 gallon Sanborn. I’m planning on using air tools so its probably marginal but about as good as you can do on a 110V outlet. I picked up an air rivet gun, an air drill and a die grinder from Harbor Freight and so far I’m quite impressed with the value. I’ll probably upgrade the chuck on the drill at some point but the quality is better than expected. I got one of their carts to try and slow the inevitable diffusion of tools across all surfaces in the workshop.

Tool cart

My local Home Depot had good quality Wiss aviation snips for a good price!

I ordered a Cleco kit, drill bits and surface conditioning discs from Pan American Tool. For the items they stock they have the best prices I’ve seen anywhere online.

I also picked up a box each of the two different grades of 1″ deburring wheels from Nebraska Surplus for a fantastic price.

On Saturday I took another trip to Harbor Freight to pick up some tools – files, vice grip style C-clamps and some storage boxes. I also picked up a digital buffer which claims to have speed control like the expensive Makitas – we’ll see how it holds up. They also had some great parts storage boxes with removable compartments which made great boxes for rivets and clecos.

Cleco organisation Rivet organization

Starting Tail Kit

My tail kit was delivered in mid September but with family visiting from England it was only this Sunday that I was able to start the build. I started by laying out all the pieces of the horizontal stabilizer. Despite all the images I’ve seen in other people’s blogs, it was exciting to see vaguely airplane shaped parts laid out on the workbench in front of me for the first time.

The quality of the laser cut parts was great with hardly any flash or burrs. I decided to start with the T03-01 main spar assembly. The parts cleco’d together in no time, I just used a one in every other hole. The most challenging part was removing the little while part number labels from the factory but I found flux remover spray released the adhesive instantly so they could easily be peeled off.

T03-01 clecoing T03-01 clecoing 2

I did notice one small issue with the assembly – the T03-03 attach angle wasn’t sitting flush against the spar as it was riding up on the radius of the extrusion. There is a chamfer on the part to prevent this but it was not deep enough. I decided it was important to have good contact between the the surfaces for shear strength so I filed down the chamfer a small amount until the part sat flush.

T03-03 issue T03-03 filing

After clecoing I up-drilled to #30. I’m really glad I decided to get an air drill, its really small and light compared to an electric drill of the same power. There is nothing like a brand new drill bit through aluminium. With the high speed air drill it cut amazingly cleanly and easily, despite me being bit nervous to do the first irreversible operations on the parts.

Next I got to try out all the various deburring tools I had amassed. Some of the holes had a slight ridge that would have required quite a lot of material removal with the speedy hole deburr tool. I found the Pan American Tool medium surface conditioning discs in the die grinder made quick work of all the burrs around the holes and I followed up with a single turn with the speedy deburr tool, applying almost no pressure, just to make sure any chip inside the hole was cleared. The medium discs were quite aggressive and left fairly deep scratches that I’ll smooth out with a Scotchbrite pad. I ordered some very fine and super fine discs which hopefully will do a better job.

On the edges I experimented with the 6″ Scotchbrite wheel in the bench buffer as well as the 1″ general purpose wheels from Nebraska Surplus. The 6″ wheel was good on the long straight edges but was quite aggressive on the corners. The 1″ wheel was very soft but left a great finish without removing too much material. For the details around the corners I found a combination of a needle file and the 1″ wheel worked great.

The plan next is to rough polish the inside of the spar channel under the rivets to make it easier to polish up later.

Time: 5 hours

Converting an airfoil from DXF to a coordinate file

Just thought I would quickly post about a useful tool I just learned about.

I was working on tracing the Cozy main wing airfoil templates with splines in SolidWorks (more on this in a later post) and wanted to bring the traced airfoil section into XFLR5. SolidWorks unfortunately has no ability to export a list of coordinates from a spline, the best I could do was to export it as a DXF with splines converted to polylines. The problem now was to turn this DXF into a coordinate file.

It seems there are a lot of tools and information available on making a DXF file from an airfoil coordinate file but very little about doing the reverse. It all seemed hopeless until someone told me about Profscan. Its a nifty little utility that can import a DXF (or a number of other formats) and export an airfoil coordinate DAT file.

One tip is it seems like Profscan only supports older versions of the DXF format so in order for it to successfully load the DXF file I had to convert it to Release 2.5 format using SolidWorks DWGeditor.