Fuselage

Hooray for shelves! OK, I didn’t finish them yesterday like I thought, but that’s cause I ended up not doing airplane things yesterday. So I maintain that this counts, you just have to interpret “tomorrow” as “the next workday.”

The first part of the day was just figuring out how to get these quarter-turn fasteners installed. I attempted to find some documentation for these Friday night, and about all I ran across were some engineering drawings. So I did some mocking up and tinkering today just to wrap my head around how these should work. The overall idea is obvious: two rivet holes to attach the base to its mounting surface, and one larger one in the center to accommodate the actual fastener.

What’s interesting is that the specs on McMaster-Carr seem to indicate that the center hole should be a different size for the two panels – but that makes no sense, as the fastener has to go through both panels. So I just started experimenting with scrap. A 1/4” hole was a good size to accommodate the fastener shaft with a bit of wiggle room.

Next up was how to actually locate that center hole. I couldn’t just use a base as a drill guide, since the center hole isn’t really round. Instead, I worked out the rivet hole spacing, and just made myself a drill guide using a piece of scrap angle. With this, I could drill three #40 holes in a piece, then enlarge the center hole to 1/4”. I did a test run drilling a set of holes in another piece of scrap angle, and it worked perfectly.

Next I drilled a 1/4” hole in a strip of sheet, and used the fastener to attach them. This worked, but didn’t hold the pieces tightly together. This wasn’t automatically an issue, though; in practice, there’s a little retainer that goes on the fastener, which is what keeps it captive in the panel. I measured one of those and found that it was .040” thick. Just right for me to use a piece of scrap hinge as a pseudo-retainer to test the final fit. (I couldn’t just use the retainer; they’re a one-time use thing, and I had no spares)

So finally, after probably an hour and a half, I was satisfied that I could actually mount these things. Next I had to actually fit them. The challenge here was accurately drilling the center hole in assembly between the shelf and mount brackets, in position in the airplane, clamped together tightly. What I ended up doing was first laying out those holes in the shelf. To do this, I installed the shelf in the plane (for, I dunno, the fourth or fifth time) and marked the outlines of the four brackets. Then the shelf came out again and I marked the counterpoints for each bracket, then drilled a #40 hole at each spot. Then the shelf went back in again, got clamped in place, and I drilled the center holes through the brackets. Finally, it all came back out again, I used my drill guide on the brackets, enlarged holes as necessary, countersunk for flush rivets, and then it was time for primer.

By this time it had been raining steadily for a while, so I ended up shooting primer on the brackets inside the hanger. It’d been windy all day, so ventilation wasn’t really an issue:

While the primer was drying, I went to work finalizing the shelf itself. I still hadn’t final-trimmed it to length, so there was that, and I also wanted to add a small flange to the aft side. The sheet has a bit of a curve to it, which has been annoying while fitting it. In practice, the hinge would keep one side straight, and the quarter-turns ought to take care of the other, but I really wanted to have it be naturally straight. So I laid out a 1/2” flange, did the cuts necessary to set that up, cleaned up the edges, and got the flange bent.

By the time that was done, the primer was dry, and I got the bases riveted to the brackets and clecoed those back into the fuselage. That made it time for a real live test fit, albeit without the fasteners actually captive in the shelf. And…it fit! Whew. So I installed the retainers (necessitating removing and then reinstalling the shelf again), and gave it another test fit.

The only thing I didn’t like was that the fasteners didn’t quite seem super snug – it took hardly any force whatsoever to turn them. It seemed to me that, in this state, they could very possibly vibrate loose in flight, which would not be a good thing with the shelf being right above the stick. I already had a loose plan for this from my earlier experimentation – I went into one of my O-ring kits, got four little 3/8” ID rings, and put one on each fastener above the retainer. With those in place, the fasteners fit nice and tight. At this point, i was satisfied enough with everything to take the final step of riveting the brackets to the brace.

So finally: pictures! Here’s the finished shelf stowed, with the brackets still clecoed. The wings on the fasteners are just visible, as is the flange in the aft end of the shelf:

And the shelf lowered, showing a little more detail on the brackets and bases:

Next, it was time for some live mockups. After having worked on placing the electrical distribution stuff on cardboard cutouts for a while, I could finally look at it in place:

This was a worthwhile exercise. The E-bus alternate feed relay (top right) has vertically-oriented terminals. It’s not obvious from the photo, but the forward upper brace is in somewhat close proximity. I think that large wires going into and out of the relay would be problematic with this layout. Fortunately, this seems easily solvable just by reorienting the battery bus a bit.

A part of the layout tinkering, I laid out some cable guides in their rough locations, and then decided to go to town annotating a photo of the setup to demonstrate the planned wire routing. All wires enter/exit the shelf near the hinge line, necessary to allow the thing to pivot properly. Power feeds enter on the left, while all the individual circuit lines exit on the right.

The red line represents switched power from the master contactor. Normally, this line will be supplying almost everything on the airplane, with the exception of a couple things on the battery bus. The switched power feed first goes to the main bus fuse block on the left, and then continues to the E-bus fuse block, going through a Schottky diode. More on that in a second as well.

The blue line represents an always-hot line, directly from the battery – where it will be protected by an ANL fuse. The line feeds to the battery bus, then continues to the E-bus alternate feed relay. This relay is controlled by a switch in the cockpit, and it allows for feeding the E-bus directly from the battery bus. In the case of a master contactor failure, or any sort of electrical fault with the main bus feed, this will allow keeping minimum equipment online using just the ship’s battery. This situation is also the reason for the large diode between the main and E-bus; without this, closing the alternate feed would also energize the main bus, which is not desirable for a few different reason.

Finally, the green lines represent all the individual wires from circuits, making a nice little bundle. That bundle will also include a switched ground wire that will control the E-bus alternate feed relay.

(now the only question is: will it look anywhere this neat when I’m done with it?)

(also that turned into a lot more of a lecture than I intended)

So now…what next? I think I need to start working up a fastener inventory for actually mounting all this stuff. Between that and the need to get the ELT ordered, I ought to have justification for a Spruce order. I wonder if I should also go ahead and order the large-gauge wire needed for the bus feeds here. With those in hand, I think I can finalize the component locations here and get stuff mounted.

In the meantime, I guess now that I have avionics on hand, I can fabricate the ADAHRS mount in the aft fuse; that’s pretty independent of other stuff, and I could also finalize the static system routing – by which I mean cutting one more piece of tubing probably a couple feet long…

I think I’m going to stop typing now.

So a couple weekends ago I made my first failed attempt at getting the mount holes done in the switch console. That, of course, ended in me ordering a replacement part. Well, the replacement (plus some raw stock for making a better switch retainer plate) came in last weekend. This week I’ve again been sort of dodging the mosquito population, though it’s finally died off enough that I did some evening work tonight, and I decided to pick back up on the new console.

Now, the issues last time mostly revolved around the holes in the console not lining up with the holes in the retainer plate. At the time I’d decided that when I redid this, I’d rivet the plate to the console to make sure everything lined up. In the intervening time, though, I decided that that wasn’t necessary, and would add decent complexity getting those rivet holes lined up. The real issue wasn’t so much a wandering step drill as it was my dumb decision to make those holes while holding work pieces by hand on the drill press.

So I decided to have a go at it doing it the same way as last time – drilling the pilot holes in assembly between the two pieces, but enlarging the holes to final size separately. This time, though, I’d make sure the work pieces were clamped securely while drilling. To facilitate this, I built up a sort of jig with a couple 2x4s. This made a piece tall enough for the console to sit on, and I next used the console to drill holes in the block matching the screw holes, which would allow the console to be clecoed in place while working.

Then I made up another retainer plate strip, laid out the pilot holes on the console, drilled those pilot holes, and transferred those holes to the retainer plate. Here’s the console with pilot holes clecoed to the block:

Next up was laying out and drilling the small retainer holes in the backing plate. Since I wanted to use that same jig block for drilling the retainer as well, I drilled the retainer plate with it clecoed to the block using the pilot holes. This allowed me to cleco both pieces to the block while drilling.

Then it was over to the drill press to start opening up the mount holes. This worked a lot better than last time, and reclamping the work for each hole was nowhere near the annoyance I might have thought it was. Here I am partway through opening up holes on the retainer plate:

Finally, after making a whole lot of metal shavings and sawdust, it was all done and it was time to see if they matched up. If they didn’t, it wouldn’t be the end of the world actually – since I didn’t mess up any of the console holes (unlike last time), worst case I could make another retainer and trace the holes from the console onto it, It’d be really painstaking to open those holes up by hand, but it’d work.

But it wasn’t necessary – everything lined up perfectly this time, and I was able to get all the switched, circuit breakers, and the PWM dimmer installed:

And of course, with that done, I just had to cleco the console into the plane. You know, just to check fitment. Very important. Plus I could lean over the cockpit and flip switches like I was doing my preflight check. Also very important.

So that was a good night’s work, and it’ll set me up nicely to start putting together the wiring harness for the console tomorrow.

In other news, manic ordering of parts has continued. I got my finish kit ordered this week, mailed off order forms for the engine and my seats, and in really big news, found out from Whirl Wind yesterday that my prop is going to ship next Tuesday! That’s barely a week-and-half turnaround on something I expected to possibly take months (based on other stories I’ve heard about Whirl Wind). To say that I’m please would be an understatement, though this also means I’ve got to find somewhere to store that thing for several months.

Next up on the ordering front is the CPI2 ignition; I shot an email off to Ross at SDS today, but didn’t hear back; I’ll probably give him a call on Monday to get that rolling. Oh, and I also got both my voltage regulators from B&C this week – another component I’ve got on hand to work out the behind-the-panel stuff.

Now I’ve just got to work on getting all these parts to fly in close formation…

Well, it was another day of road blocks of a sort, but I did my best to get stuff done in spite of it all. I’d intended to try and get most of the switch console wiring done, though I knew going on that I had at least one roadblock. One thing I replaced in this second console attempt were the circuit breakers; when I was ordering my regulators rom B&C I saw they had mini Klixon breakers, much more compact than the ones I got from Digit-Key previously. Since I was worried about the clearance between the breakers and the side of the console, going with those was a no-brainer. The problem is, these breakers need a #6 ring terminal, unlike the #8 for the original breakers. Guess what ring terminals I don’t have on hand? Yup.

But I figured I’d at least get started on the wiring stuff anyway; I’d probably not get to the point of loading up the connectors or bundling the harness, but I could at least get some terminations done. So I drug all the assorted wires needed out on the workbench, and here I was confronted with another problem: I’ve been intending to terminate the console harness with a pair of Molex Mini Fit plugs; this was the best way to get a fairly compact connector overall, since Molex doesn’t do more than two-row connectors in this form factor.

The problem is, I saw that the current rating of those connectors was fine for everything I have, but didn’t consider the wire gauge the pins could accept. And it turns out the biggest they’ll take is 18AWG, whereas I need both 14 and 12 for the seat heater and pitot heat, respectively. I did find that there’s a Mega Fit series that will work with this wire size, so I’m going to rework my connections here into a 6-pin Mega Fit connector for the high-current stuff and a 20-pin Mini for the rest. More stuff to order…

Oh, and that still wasn’t the end of it. I also didn’t have the right blade terminals for the 12AWG wire needed for the pitot heat circuit, so I couldn’t terminate those to the switch either.

I still decided to go ahead and cut the wires that I could and crimp on the terminals needed to connect them to the switches, but that was as far as I went. I’ll have to wait until I have the rest of the stuff I need before I can turn this rat’s nest into something that looks actually organized:

So – what to work on next? Part of me wanted to sit down and start getting more parts orders together, but I can do that any time, and maybe wait until tomorrow in case I find more stuff I need. I still don’t want to finalize anything on the forward shelf until I get the CPI2 ignition stuff on hand, but I’m quite confident I can get everything I need up there – and that means I can start finalizing the rear shelf.

What actually started me on this path was trying to think of other wiring things I could do fairly independently, and the one thing that came to mind was the Skyview network connections in the tailcone – specifically the runs from the network hub back to the ADAHRS units. I won’t want to bundle those until I have the trim and strobe runs going back as well, but I can still get through all the pinning. But first I needed to have the hub location nailed down.

Before that, though, I needed to actually make the shelf attachable. So I carefully positioned it, got the four screw holes drilled, and installed the nutplates in the longerons. Then I pulled out the pertinent components and worked on a decent layout for them – which I’d already roughed out, so this was just finalizing things. Nothing much changed here, except that I think I had the hub and ADS-B receivers switched previously. This new layout makes it easier to get tot he receiver’s harness connection, and positions the hub close to where the pitch servo will live. By doing that, I’ll be able to just terminate the wires on the servo with a D-sub, and plug them directly into the hub – no intermediate cable needed.

So that was pretty much today’s work – I didn’t quite get to laying out mount holes in the shelf before I had to go to an afternoon appointment. Tomorrow I should be able to get this stuff mounted and at least start on those network cable runs. And I guess place some parts orders…

Today I picked right back up with the rear shelf. I was happy with the locations of all the components there, so I marked off and drilled all the various mount holes. The one interesting bit out of all this was the ELT, whose manual insists that I should have drilled #19 mount holes, despite all the preexisting holes in the tray being obviously sized for a 3/16 bolt or screw. So I drilled them #12 instead. This annoyed me for a bit, because I thought I’d need to order yet more fasteners for this, but wait, I did order some 10-32 screws already. Good thing I bought extras in an assortment of lengths.

Anyway, there was just an extended grind of making those holes, laying out nutplates of assorted sizes for all of them, and getting those nutplates riveted. But then it was done and I got all my stuff attached.

That allowed me to move on to the original thing that prompted wrapping up the shelf: the network cables from the network hub (second from right above) to the ADAHRS. This would be my first time assembling a Skyview network cable – hooray for new things! Now, it wasn’t strictly necessary to make these – Dynon sells prefab cables, and those probably would have been a time-saver – but by making my own, I can make them exactly the length I need and also save a decent bit of money.

They are a bit labor-intensive, though. Each cable consists of four twisted wire pairs (two power/ground and two data pairs), plus a ninth single wire. Here again, you can buy pre-twisted wire if you want, but I didn’t do that either. Twisting your own isn’t really that hard – Stein from SteinAir has a pretty good video on his Youtube channel showing how to do it with a drill, and it works well.

But first, I had to figure out how long to actually make the cables. One idea I’ve seen is to rub ribbons through the fuselage and use those as a way to measure future wiring runs. In this case, rather than going looking for ribbon somewhere in the house, I decided to use my huge surplus of static tubing instead. I figured it was a little easier to have this hold the expected routing than ribbon. Looks kind of ridiculous, but it does the job:

Now, the prefab network cables have each wire a different color, to properly differentiate between the identical pairs, but I didn’t want to spring for a bunch of special color wire. Instead, I used red/black for both power/ground pairs, and green/blue for both data pairs. I used my printable heat shrink to label the pairs so I could be sure to pin the connectors properly. After labeling one end of the wires thusly, I stripped each end, crimped on the D-sub sockets, and then very carefully pinned my first connector. Some heat shrink at the strain-relief point, and into the backshell it went:

Next up was bundling up the cable. My initial intent was to go to town cable-lacing the thing, but after just a few knots, I didn’t really like the way things were laying up. I think the lacing probably works better for non-twisted wires; in this case the twisted pairs seem to just make for a lumpy bundle between ties:

Instead I decided to use the 1/4” flex wrap I bought a while back to just enclose the entire length of the cable. This provides the sort of smooth overall finish that I’m going for, at the cost of some annoyance – getting that stuff over a bundle of wires has a sort of greased-pig quality to it. Once I got the sockets crimped on the other ends of the wires, I took it over to the fuselage and got it routed.

The question at this point was whether to go ahead and pin the other end and add the backshell, or wait until later? Leaving it as-is meant having the end of the flex wrap to possibly fray. Previously I’d imagining just having the network wires live in the larger bundle it’ll be routed with, but now that I’ve wrapped the whole thing, there’s nothing to be gained there. So I decided to go ahead and finish the cable. Later, when I pull the other wires to the tail for the tail light and elevator trim, I’ll just lace them into a single bundle in place.

But before committing to heat shrinking everything together, I tested the cable, making sure that the corresponding pins on both ends were continuous. Definitely a better-safe-than-sorry moment. There were no issues, so I finished the cable and plugged it into both ends of the run. At the moment, there’s a lot of slack, but the routing at the shelf end will be more circuitous than what’s represented here:

In the end, it was definitely tedious and time -consuming, but this was also my first time doing any actual d-sub stuff, so I imagine that the next cable will go a fair bit faster. One thing I’m thinking about doing is building up a couple dummy connectors to allow for easier testing of these cables. To do this one, I had to grab two unused pins and insert one into the location being tested on each end, and probing those with the multimeter for continuity. The pins were necessary because my probes are to large to go into the sockets. With a dummy connector, I could chain the individual circuits together and basically test them all at once. I’ll have to think about that one.

Next up, I get to repeat this exercise, except the next cable will be 2” shorter than the other one. I guess that’s when we’ll get an idea how much quicker subsequent cable build-ups will be. Now it’s time to go put together yet another Digi-Key order. Maybe a Spruce order too. Oh, and possibly SteinAir…so many things…

Nothing too exciting tonight; I did some tidying up and then got to work on network cable #2. As expected, this one went a good bit faster than the first, though I also caught myself getting a little too comfortable – I started pinning up the second connector on the bench. Fortunately I realized my error before seating any pins, so I didn’t have to actually do any defining.

Testing was a lot quicker this time too. I realized that my previous idea of making a sort of loopback test plug wasn’t smart – all I’d really be testing was that al the circuits were continuous, not that they were actually paired up correctly. Instead, since I was putting together another Digi-Key order early this week, I added a couple of solder-cup DB9 connectors. I can just plug those into both ends of the network cable and use the solder cups for continuity testing.

In any case, now I’ve got network cables to both of the ADAHRS units:

Since my Digit-Key order came in this week, and with those two network cables done, today I was able to get back to work on the switch console. It was a bit of a late start though, since there was an EAA fly-in across the runway that I spent most of the day at.

I got the terminals crimped on the remaining wires, got them in place, and commenced to bundling everything together temporarily so I could figure out the Molex plug locations and get all the wires trimmed. I got the terminals crimped on all the wires, and even got the high-current ones finished and pinned into their connector, but before I could do the same for the low-current wires, the mosquitos drove me inside. Between it being dusk and me having the hangar lights on, it was turning into a bloodbath. I guess I need to go spray out there or something.

Anyway, at some point during the work session, Josie came out and I commented along the lines of “see, it’s looking less like a rat’s nest,” to which she replied with a skeptical look. I guess I can see where she’s coming from; it looks better to me, but certainly still far from organized:

Tomorrow I ought to be able to get everything into the connectors, and then with some cable lacing this ought to morph into something that actually looks organized.