ELT service

December 10th, 2023

I'm such a slow builder that the batteries in the ELT have timed out, so I need to replace them to be legal. There are three in total, starting with the one in the remote indicator in the panel:

In addition to making entries in the aircraft logs, I put stickers showing the battery expiration dates on the side of each component:

The main battery pack for the ELT itself is huge and expensive:

Despite the expense, a replacement ACK ELT battery pack doesn't come with the required o-ring, so I re-lubed and reused the old one:

The third battery is in the audio alert unit, which I quickly regretted installing on the aft side of the F-707 bulkhead where it couldn't be seen:

When I reinstalled all these units I moved the audio alert unit to the front side of the bulkhead, so it will be slightly easier to service next time:

And while I was crawling way back in the tailcone, I fitted the ELT antenna static suppressor that's now apparently required by a service bulletin:

Cowl fasteners

November 19th, 2023

I have been putting off finishing the cowl fasteners for no good reason, so I decided to tackle this job at last. I am using the standard hinges on the vertical portions of the lower cowl at the firewall, machine screws behind the spinner, and Skybolt quarter-turn fasteners everywhere else. In total I needed 42 sets of studs and receptacles to attach my cowl:

I put the cowl on the airplane, made sure everything was still fitting well, and drilled the existing pilot holes up to #30:

Then I removed the upper flanges from the firewall, and used the provided templates to drill all the required holes. Drilling the center holes from #30 to 11/16" with the unibit generated a tremendous amount of shavings, and required some concentration to keep the bit from wandering off-center:

After preparing the flanges, I riveted everything to the firewall. Afterwards I realized that this might be the very last structural riveting of the whole project… sure there will be additional rivets for nutplates and so forth, but I think this is the last bit of primary structure that was still left un-riveted.

Due to the limited space available in places, I chose to attach the Skybolt receptacles to the flanges with countersunk blind rivets:

Back to the upper cowl, I drilled the #30 pilot holes along the aft edge up to 15/32", again watching to make sure the unibit didn't wander:

After installing the studs and grommets in the upper cowl, I did a test fit and was happy to see that I had everything lining up well so far:

Since the mounting flanges along the bottom of the lower cowl were already riveted to the firewall, I had to drill the holes in-place:

To avoid drilling through the engine mount or various nearby components, I made the above holes using a 5/8" Rotabroach cutter, with a washer and a rubber o-ring acting as a makeshift drill stop:

Using the Rotabroach cutter for the bottom flange holes gave me an idea for a better way to drill the holes in the fiberglass cowl. Instead of separately drilling the holes in the cowl and the underlying flange and then hoping they line up, I realized that I actually had a way to make these holes perfectly concentric.

On the cowl side flanges, I used the Skybolt template to drill the #30 holes for the rivets and lock pin, but I left the center hole at #30:

(yes I realize the edge distance is not great at the front, but it's too late to extend these flanges now)

Here's the entire process in one image: I installed the cowl with clecoes, then removed one cleco, cut through the fiberglass using the ejector pin on the Roatabroach as a pilot, and replaced the cleco with a 1/8" fender washer underneath. Repeating this process for every fastener location gave me a row of perfectly-placed holes in the cowl, even where the holes in the fiberglass were wonky or buggered up.

The Skybolt grommets require a 15/32" hole and the closest Rotabroach cutter I have is 7/16", so I cut all these holes slightly undersized in-place and then enlarged them the last 1/32" on the workbench with a unibit. Then I removed the lower cowl and used a 5/8" Rotabroach cutter to make the holes in the flanges for the receptacles:

Doing it this way guarantees concentric holes, since you're using the same pilot hole to guide the cutter in both cases… really a special kind of match-drilling.

After deburring everything, I riveted the remaining receptacles to the lower cowl, and installed it back on the airplane:

And sure enough, every one of my fastener studs engages perfectly with its receptacle:

Adjusting the depth of the receptacle inserts takes some trial and error. You have to install the cowl, mark the ones that are too high or too low, remove the cowl, unclip and adjust each receptacle, and repeat. Each half-turn of a receptacle adjusts the depth by 1/64", so although you can't get every one perfectly flush, you can get pretty close:

After finishing the Skybolt fasteners, I bent the ends of the lower cowl hinge pins upwards, and riveted these little clips to the inside of the cowl flanges. Now the pins are secured from rotating or backing out, but they can easily be sprung loose when required:

The airplane looks pretty good with the cowl attached. I still have some minor fiberglass work to do, but that job will wait for a future work session:

Wingtip lighting & final installation

October 1st, 2023

I spent a couple evenings installing the eighty or so nutplates that will be used to attach the wingtips. Once again I slightly under-squeezed the rivets to avoid cracking the fiberglass, and I was none too picky about achieving countersinking perfection. These holes will all be hidden when in use, so as long as everything fits together and the nutplates are nominally attached, it will be good enough:

The Whelen nav/strobe heads are designed to be screwed to a mounting plate, which is in turn affixed to the wingtip. However, they don't give you much of a hole for the wiring, which is sort of annoying. I passed the wires through the mounting plates, then installed Deutsch DTM connectors, which will have to be de-pinned if I ever want to separate these components:

I have known about Deutsch connectors for a long time, but only recently picked some up to play with. Once you understand the installation process, these connectors are just as easy to use as Molex connectors, and are apparently vastly more reliable. Over time I'll probably replace any problematic Molex connectors with Deutsch, but I can't be bothered to go back and redo any of my old work at the moment.

I cut holes in the wingtips to allow the connector to pass through. Luckily this will be hidden behind the mounting plate, as will the extra hole I previously drilled to help with fitting the lenses:

I attached the mounting plates to the wingtips with large-area washers so the fiberglass won't crack. Since rough fiberglass is abrasive, I used some snake skin on the exposed wires to prevent chafing:

With zip ties, grommets, snake skin, and clamps, I cleaned up and secured the wingtip wiring. Later I realized that these might actually have been the very last wires I needed to finish on the whole airplane:

With nothing left to do to the wingtips, I aligned them to the control surfaces and drove in all the screws, then installed the lenses:

Here's a short video showing the airplane all lit up. The beast is beginning to stir…

Wingtip lenses

September 20th, 2023

The two lenses for the wingtip lights come as one conjoined bubble, which you have to split in half as a first step. I had kept this part packed in bubble wrap, and was pleased to see that it survived the cross-country move with only a few small scratches:

There is enough material here that you don't have to be super precise about the initial cut – just dividing it at the approximate midpoint is sufficient:

The best way I found to cut this material was with a round blade in an oscillating multitool – buzzer saw, shaker saw, or whatever you happen to call it. With this method I was able to make controlled cuts pretty close to the desired line. It is loud and it does produce a lot of dust (or is it fumes?) that seemed inadvisable to breathe, so all the protective gear is a must:

For the initial edge straightening after a cut, I used my handheld belt sander, converted to a sketchy benchtop unit. This is more solid than it looks, as the wood crosspiece is screwed to threaded mounting bosses in the plastic handle. But just in case it managed to get out of control, I used a cheap deadman switch to power it:

Then it's an iterative process of marking, trimming, and sanding. I used a dry-erase marker, which worked well. For removing more than about 3/16" I used the buzz saw and sander, otherwise I just went after it with a sanding block. The material is thin and pretty soft, so it machines quickly. I've even heard of people using tin snips to cut the lens material, which I think would probably also work if done carefully.

Eventually I had the lenses fitting about as well as I figured I was ever going to get, so I sanded the edges to 320 grit, hit them with a scraper, and then polished them with a felt wheel. Then I drilled one mounting hole in each corner:

I don't know what the plans say about mounting the lenses, since I seem to have lost the loose instructions sheet that came with the lens bubble, but usual method is to use nutplates. These are often held in place with pop rivets, due to the geometry of the wingtip and the hazards of driving solid rivets in fiberglass. I decided to take a different approach, still using nutplates, but with their mounting ears twisted roughly 45 degrees:

I simply gooped the nutplates in place with blobs of epoxy/flox, which is inelegant but definitely worked. The purpose of the bent nutplate ears is to give the flox something more solid to grab onto. Now I have the nutplates mounted permanently, and I'll never have to deal with wear issues related to pop rivets in thin fiberglass:

The fit of the lenses at this point was "okay", but if you zoom in on this picture you can see that it could be better. The profile of the lens and wingtip don't match perfectly at the apex, and the edges of the lens don't match the wingtip recess:

I know the latter must be due to the molded recess having a wavy edge, because the lenses themselves are dead-flat if I set them on a table. To improve the fit, I decided to add some filler where necessary.

I applied clear tape around the outside perimeter to keep the epoxy off the plexiglass, but I couldn't get the tape to lay down nicely along the complex curve of the inside surface of the lens. Instead, I brushed on a layer of PVA mold release:

I spread an epoxy/microballoons mixture on the wingtip recesses, screwed the lenses in place, and then added more filler on the outside where there were low spots:

After letting everything dry – while worrying that I had just made a huge mistake – I managed to remove the lenses from the wingtips without damage, but it was a real struggle. I eventually resorted to drilling a small hole in each wingtip so I could blow in compressed air, which helped pop them loose. I'll repair the holes with flox later.

After sanding off the excess filler, I now have perfectly-matched wingtips and lenses:

I still have some smoothing to do in a few spots, but some of that can be left for the painter to worry about. Meanwhile, I'll consider the wingtip lenses finished.

Wingtip reinforcement

September 3rd, 2023

As-delivered, the wingtips are fairly flimsy and prone to oil-canning. To rectify this I decided to follow the same trail previously blazed by others, adding stiffeners to the insides to make them less flexible.

With the wingtips clecoed in place, I squeezed various spots to identify the weakest areas, and used masking tape to mark some proposed locations for internal stiffeners:

I bought a bushel of 1/4" dowels – which turned out to be more than I needed, but thank goodness for generous return policies:

I marked out where I wanted the dowels to live, cut them to length, and beveled the edges. Then I temporarily attached them to the fiberglass with CA glue, using weights to keep them in place while the glue dried:

Sort of a whalebone corset effect inside the wingtips – I guess in aviation terms these would be stringers:

I mixed up an epoxy/flox mixture, leavened with a bit of micro, and used it to more permanently glue the stringers to the fiberglass. I tried to incorporate a nice fillet as I went:

Then I laid a strip of glass cloth over each stringer:

It doesn't convey in a photo, but the wingtips are now immeasurably stiffer, and demonstrate much greater resistance to flexing:

I'm totally happy with the outcome of this modification – for the cost of a few work sessons and some inexpensive materials, the wingtips are now much less likely to crack due to aerodynamic forces and ground-handling mishaps.