Avionics

For the time being, I’m stepping away from the pitot layout stuff, in favor of a more general look at systems. As I alluded to previously, I’m getting to the point where it’d be nice to have some electrical supplies on hand for stuff like extending the pitot heat harness. But if I’m going to order wire, I’d like to go ahead and do it in bulk, so I’m pivoting to working out my avionics layout so I can then at least rough out all the harness routing and then get a materials estimate so I can order a whole bunch of wire and stuff.

So I started actually a couple nights ago with the arts & crafts portion of this program – making mockups out of cardboard for assorted boxes. At first I was just going to sorta mass-produce mockups of everything, but then it became clear that in most cases, I don’t need a full three-dimensional mock; just a flat box with connection locations identified is probably good enough.

In any case, after making flat mocks of both voltage regulators, I moved on to the ADAHRS units. For these, a three-dimensional mock was called for, since vertical clearance will matter, especially for maintenance accessibility. Then, after making these, there was kind of a pause for some documentation reading and plans changing.

Both ADAHRS units connect to the Skyview system using the Skyview network connection. My original plan was to run a single network cable back to the tail, then add a splitter cable to accommodate both ADAHRS units. However, I’ll also have the pitch servo for the autopilot back in this same area, which also accepts a Skyview network connection, albeit with a couple of pins modified. So if I use the splitter, I’ll still have two cables running back here; I could change that by using a Skyview network hub, which allows tying together up to five devices. I got curious as to how much more expensive the hub was than the splitter cable, and discovered…it’s actually cheaper. Huh.

So yeah, I’m going to put a hub back here, and then I can essentially run a single network cable for everything in the tail, and then branch out from the hub. That, in turn, gave me something else to make a mock for (though just a flat piece in this case).

Finally, I wanted to mock up the shelf on which the ADAHRS and the hub will live. Cardboard didn’t seem like it’d cut it here, but I had some thin plywood scrap that would work well. I ended up having to cut it with a hand saw, since my bandsaw kept throwing the blade (seems the lower wheel needs a new tire). But with that done, I could try laying out the units here to see how it fit together.

Here’s the shelf in place. This is just behind the aft baggage bulkhead, which is convenient for maintenance. Accessing these will only require removing that bulkhead, which is just held on with a few screws. (This glosses over the part where I’ll have to be lying inside the fuselage. The word “convenient” is kinda relative here.)

Anyway, with the bulkhead out of the way, I’ll be able to easily get in with a screwdriver to remove a unit for service, and having the hub here will make it easy to add service loops on the network cables. Since the cables attach to the backs of the units, this will allow me to detach a unit, then remove the network cable after pulling it out of the bay a bit. The next photo shows what this looks like from the baggage area, just forward:

I figure the network feed cable from the forward fuselage will come back along the left side somewhere, then climb up the bulkhead to connect to the network hub. From there, another cable will descend back down the bulkhead to the pitch servo. The two cables to the ADAHRS will just have a short run.

On the front of the ADAHRS units, the pitot/static/AOA lines will connect. That’s the one thing I want to think about before I commit to this – what will that plumbing look like? Each of those lines will need to be teed somewhere in this area to connect to both units. These lines will also come from forward on the left side of the fuselage, since the pitot/AOA lines will be coming from the left wing. The static line will probably feed from aft – I need to double-check where the static ports will be located.

Next up will be planning the rest of the aft avionics. Somewhere back here will live the com radio, the remote transponder, the ADS-B receiver, and an ELT. I suppose it’s time to go look at other people’s builds to see what they did for mounting those items. A shelf bridging across the lower longerons seems like a decent approach, but it’s a long span and would need to be pretty well stiffened, especially to mount the ELT.

Decisions, decisions…

Another day spent figuring out where to put things. Today’s focus was figuring out what to do for an aft avionics shelf, something that was a lot more challenging than it seemed like it ought to be. For whatever reason, I had very little success finding photos of avionics bays in RV-8. I had the idea of putting a shelf just ahead of the elevator bellcrank, just behind the baggage bay, and I found a photo of someone who’d done that, but it just didn’t seem workable to me, at least taking the simple approach of running two pieces of angle laterally between the longerons to serve as the edges of the shelf.

The problem is that doing that only gives a few inches of depth on the shelf, and both the ELT and transponder are much longer. This seemed problematic enough that I tinkered around with several other ideas, most notably putting the shelf just below the ADAHRS shelf, bridging the middle longerons. This location would provide good access through the baggage bulkhead, but it would put those avionics far too close to the ADAHRS units. Since those include magnetometers (read: compasses), they need to be away from basically anything with a power connection.

So putting the avionics here would require either relocating the ADAHRS, or else additionally using a remote mount magnetometer. Any other ADAHRS location would have been hilariously inaccessible, and the remote magnetometer seemed like obnoxious extra work, though I still considered it as a backup plan.

Finally, I came up with an idea to get better shelf space in the original location. Instead of using two pieces of angle oriented perpendicular to the long axis, I could only have the forward angle like this, and at the rear, have two angle pieces…um, angle in…from the aft bulkhead to a point forward of the bellcrank. This allows for a lot more depth where needed for deep items, even if it does look a little weird.

The piece of plywood I used to mock this up actually has the cutout around the bellcrank a fair bit overdone. More realistically, either I’ll have the two angled lines simply meet at the centerline, or else add a short lateral piece at the place where the angles meet the cutout below:

This location provides plenty of depth for the transponder (left) and ELT (right), as long as they’re located outboard. Also shown here is a 2D representation of the ADS-B receiver. This location also provides ample space forward of the units to allow for wiring connections and harness routing. Originally I was going to mount the remote com radio back here as well, but since both the control head and the antenna will be located much further forward, I’m going to prefer putting that behind the panel if possible. If necessary I can still probably fit it back here, too…maybe over beside the transponder.

Next up will be figuring out some kind of avionics shelf behind the panel. I’ve got a couple ideas for that area, none of which I’m super enthused by so far, so that’ll probably end up being another bunch of hours spent puzzling over things. I think I’m also going to go ahead and order a few items that I’ll be finding homes for, since they’re relatively inexpensive – I’m specifically thinking of the fuse blocks I’ll be using behind the panel.

Pretty much the same story as last time. Tonight I made up cardboard mockups of the remaining units I need to plan for, and which I at least sort of intend to put behind the panel. Then I got to looking at the area behind the panel with an eye towards making up an avionics shelf. What makes this interesting i that there are two braces that go across the cabin, tying together the tops of the gear towers. These seem like great starting places for a shelf, except one is higher than the other for some reason.

Still, it seems like people are using these for their shelves; one photo I found simply makes the shelf sort of L-shaped, placed at the level of the lower brace and then having a vertical element to tie into the higher one. I’m tentatively leaning towards just having a sloped shelf going between the braces. I kind of like this better than the L-shape thing, and I think it’ll actually make the stuff back here a little easier to work with for maintenance; I can either remove the EFIS units or the entire panel center section, and then have the shelf stuff presented nicely to me as I sit in the seat.

Anyway, for the time being I cut out a piece of cardboard roughly the size of the shelf and tried laying stuff out on it. Everything seems workable off the bat, though I haven’t carefully accounted for the space needed for connectors. I’ll likely end up with some things mounted against the baggage bulkhead as well, most notably the two voltage regulators (not shown in the following photo).

So none of this is final by any means, particularly since I need to account for those connectors. I guess maybe I should make up cardboard versions of those as well; it feels a little weird but this is definitely something I need to take into account. Mostly now I just get to move stuff around over and over again.

In other news, I’m continuing to accumulate parts. Got some tailwheel stuff coming in tomorrow, just ordered my fuse blocks, and it would appear that the prop governor I ordered via a VAF group buy is about to ship. Oh, and I might have a line on a slightly-used exhaust system for about half the price of a new one. Not super confident that one will work out, though – it’s the Vetterman system I want, but it’s for an IO-360 in an RV-4. Part of me wants to think that the packaging in the cowling can’t be that different, but since Vetterman has different part #s for the two exhausts, clearly there are some differences. But maybe the RV-4 exhaust can be made to work…the savings would certainly be nice.

So when I got started nice and early this morning, I decided to go ahead and start fabricating my aft avionics shelf, and I figured I’d start with the most complex part of the thing – the carefully-bent piece of angle to support the sort of wraparound shape of the back of the shelf.

So I cut off a chunk of angle long enough to do the deed, plus some extra because I’m not a complete idiot. Then I used the plywood mockup I had to mark where my bends would be, and cut out notches to allow for those. Those were rough cut on the band saw and then refined with files, and then the bends could be made. It ended up nicely fitting with the template:

Next, I used the template to trim the ends to length, and all I had left to do was to remove one leg of the angle to leave mounting tabs that could bolt to the longerons. And that’s where this went sideways – I set up the trim wrong, and ended up mangling the ends. Welp…learning experiences abound. So I cut off another chunk of angle and started over again. This time I got the mounting tabs done right, and the piece ended up fitting nicely in the fuselage:

Well, except for one thing. I neglected to account for how, where this piece crosses the fuselage centerline, there’s the beginnings of the elevator bellcrank mount, which sticks up enough that the center of the angle contacts it:

I made an attempt to cut a small relief in the angle there for clearance, removing about 3/16” of the 1/4” angle, but it still contact lightly, though it at least sits level now. Still, I’d probably need to make that relief cut twice as deep to ensure no contact, especially under loads in the air. At that point, I’d have removed a lot of the rigidity provided by the angle here, and right at midspan, where I most need that rigidity.

So long story short, I think I’ve got to go back to the drawing board here. What I may end up doing is figuring a way to tie into the nutplates on the baggage ribs, which are there to support a rear battery box, if installed – but mine is going up front. That would at least provide some nice support to the center of the shelf, but making everything line up properly will likely be a challenge.

So yeah, time for some more thinking.

UPDATE: After mulling over this some more, one conclusion I’ve come to is that I am seriously overbuilding this avionics shelf. Basically, my planned structure would consist of aluminum angles to bridge across the fuselage, with sheet rivet to those angles to form the shelf. However, the upper baggage shelf – which will sit right above this location – has no sort of steparate stiffeners whatsoever. It’s just .025” sheet, with a little rigidity provided by way of bent flanges forward and aft.

Van’s doesn’t list a specific max weight for this shelf, but for the sake of comparison, everything I’ll be putting on my avionics shelf is just over three pounds. If I were to end up putting the com radio back here, it’d go up to about 3.75#. I’m sure that baggage shelf can support well over that!

So with this in mind, I’m going to start with rethinking my shelf and just using sheet aluminum with bent edge flanges for lateral support, instead of the complexity of bending angle in fancy ways. I’ve already got an idea how to use this approach to build my strangely-shaped shelf, too. Maybe I’ll give that a try tomorrow.

Alright, so here we are looking at v2.0 of this oddly-shaped shelf. Yesterday was kind of a bust overall, but it eventually got me looking at this problem in a different light. Most of the challenge with this shelf is the way the aft portion sort of wraps around the bellcrank area, and trying to work that with aluminum angle wasn’t doing too great. But last night I eventually concluded that using angle was overkill anyway.

So today I started on a new shelf, this time just using aluminum sheet with formed flanges to provide stiffness. This was also the beginning of a new approach to handling the odd shape. Now, I could have done this all as one piece, but it would have been a bit complicated, especially when it came to forming the aft edge. So instead I decided to take a piecemeal approach. To start with, I’d form a small shelf that went straight across the fuselage, and then form two separate “wings” for the angled portion, which I could then rivet onto the main shelf.

I started with the simply part today – the straight “main” shelf. Material was donated by a surplus wing skin, which was damaged years ago during our move from Atlanta. It’s been just sitting around waiting for the day it was given a new purpose. Basically, I cut off one bay of the wing skin to start with, the used my plywood shelf as a template to get the angles right where they’ll meet the longerons, and finally added allowances for a forward and aft flange. Then there was just lots of cutting and filing and fine-tuning and cleaning up edges, before it was finally time to do some forming:

The tooling holes being used to hold the piece on the forming block are in the locations where screws will attach the shelf to the longerons, but just drilled to #30 for now. (I didn’t want to go look up the right drill size for the screw, I’ll figure that out later)

Getting the whole thing cut out and formed was about an hour and a half of work, but the end result fits nicely in the fuselage, just ahead of the bellcrank:

After attending to other stuff in the afternoon, I found a bit of time after dinner to pick up where I left off and get started with the “wings.” The main shelf didn’t really need a dedicated forming block, since the edges were just straight, but the wings are a different matter. The good news is that since they’re mirror images of each other, I only need to make one block.

Here again, the plywood mockup served as a pattern, though I had to make some slight concessions due to the dimensions of scrap wood I had on hand. Basically, I’ll lose about 1/4” off the back of the wings – which should be no big deal. I decided to call it a night once the block was done, before getting into cutting up more sheet. Here’s a look at the block alongside the main shelf, showing where it fits. The main shelf and wings will simply be riveted together through the mating flanges:

Tomorrow: more cutting and hammering…

Yep, finished up this shelf today, for the most part working in half-hour shifts before retreating to the air conditioning. Houston summer has definitely arrived in full force.

After cutting another section off my wing skin to start with, I laid out the cuts. The nice thing about using a dedicated form block like I did in this case is that it can double as a template – just lay it on the sheet and trace around it, and since my flanges are intended to be 3/4”, the same as the block thickness, the flanges can be traced using the edges of the block. Much faster than doing a bunch of careful measuring. I also laid out my tooling holes in the block and pre-drilled them, so I could use the block to drill those in the pieces and ensure that both parts could be attached to the block.

Some time and a bunch go cutting and filing later, I had the sheet pieces ready for forming:

From there, it was just a matter of clecoing the sheet to the form block, clamping it in the vise, and applying some strategic physical violence with a hammer. The only hangup here is that thank to the material’s springback, I can’t quite get a 90° bend with this block. That would require relieving the edges of the block a bit, and then wouldn’t be symmetrical and thus wouldn’t work for these pieces.

One mostly-finished wing piece:

After squaring up the flanges, it was finally time to cleco the assembly together and see how it fit. Seems decent:

After that there was just laying out the drilling the rivet holes, and eventually squeezing rivets to assemble everything. After a test fit in the fuselage, I needed to work on the outside edges a bit to match up to the fuselage taper, but eventually it fit like a glove:

And here it is with the cardboard unit mockups sitting roughly in their planned homes:

I still haven’t determined if the remote com radio will be located here or behind the panel up front. I realized yesterday that when I was mocking up stuff for the forward avionics shelf, I completed forgot about the ECU unit for the SDS CPI2 ignition. Granted, I haven’t gotten anywhere close to having a real layout for that shelf, so it’s not a huge problem, but it is another thing that could make me decide to move the radio aft.

I’m not convinced that’s going to be necessary, though. I have the beginnings of an idea for basically a two-level setup behind the panel, but it’ll require some tinkering to see if it’ll work. The general idea is to have two low-mounted shelves on the right and left, which will be hinged at their outboard ends and can be swung down beneath the panel for access. The fuse blocks could go on the left side, and the CPI2 controller unit on the right. (I want this to be somewhat-easily accessible, since it has some fuses in it that might need service in a pinch)

Above this would be another flat shelf, high enough to clear the units on the hinged shelves below, and that could house all the other units, stuff that can be less easily accessible. Getting to those would require pulling the EFIS units and/or the entire center panel.

My fuse blocks should come in from Stein Air tomorrow, which will help me start to determine whether this idea is feasible or not. Coming up: lots more photos of cardboard and cheap plywood mockups, I suppose…

OK, so I haven’t been super productive this past week or so, at least in terms of shop time. I did take a giant leap this past Thursday, though…after exchanging a few emails with the folks at SteinAir and running a couple quotes, I finalized and placed my avionics order. Sounds like I can expect to have all that on hand in maybe a week and a half, assuming nothing is backordered at the manufacturer.

This has me back to getting serious about the remaining avionics shelf design, specifically the two shelves I’ll be putting behind the panel. I did a bit of layout work this morning, trying to finalize the design and measurements, but didn’t really get that far. For some reason I’ve been having trouble getting into a work mindset.

This also has me thinking about what else I want to have on-hand to really start laying out and fabricating harnesses. With the main avionics order placed, that takes care of almost all the stuff I’ll need wiring for. The exceptions are the ELT, the assortment of switches and breakers, and the SDS CPI2 ignition controller, so I’m likely going to get those ordered semi-soon as well. This is mostly simple, with the possible exception of the CPI2 – one of the things I’ll need to know before ordering is the required harness lengths.

The main thing that makes that fun is knowing the required lengths for the firewall-forward stuff. I can figure out the required runs from the controller to the firewall penetrations, but I have no idea the required lengths past that. I should probably just get in touch with Ross, I imagine he can help here.

I still need to finish up riveting the forward bottom fuselage. Finalizing the avionics shelf will probably require me to at least rivet the baggage bulkheads into place, and having those in place will make it a bit more fun to get to the rivet locations inside the cabin for the bottom fuselage riveting…