What have I gotten myself into!: Ch 22 - Electrical/Avionics

Roll servo installation

I sure wished the currently installed Navaid servo motor had been compatible with the GRT digital autopilots, unfortunately reality is seldom that simple, and as I had dreaded I’d have to replace it.

To be completely honest, a very small part of me was actually glad about having to do this, since the few practice GPS approaches I shot with the Garmin GNC-300XL (driving a VFR Trio autopilot, which in turn used the NavAid servo) turned out to be quite scary once one got down to about 500’ AGL. I had expected that much, and I wasn’t too surprised.

This Trio EZ Pilot has served me well

NavAid servo motor in its original location

Yeah, I know what you are thinking and I had the same reaction initially, however even after 15 years of sliding on the nylon corner, the rudder cable had neither worn itself nor made a mark on the nylon... I can't argue with that.

Now for the good-news bad-news part… turned out that the Navaid servo and the GRT servo bracket had the same bolt pattern. How cool is that!? I should have been able to take the old one down, put the new one up, run some wires, and… Boom!… Done!

So, where’s the bad part?

Larger GRT servo attached to the NavAid servo mounting bolts

Are you starting to see the issues?

Yeah, as much as I had hoped for this to work, it just didn’t. The steel cable rubbing on the digital servo was not gonna happen, plus the adjustable rudder link was now occasionally snagging on the servo (extremely bad situation).

If that wasn't enough, I had planned on enclosing the servo to give it some protection from the radiant heat of the exhaust pipes, and this would only have exacerbated the issue.

Enclosing the GRT servo

An RV-9A donated the aluminum sheet to make this box possible

Pardon my metal fabrication skills, they have not received the same amount of practice as my glassing ones.

Adhesive heat barrier taped on the sides...

... and on the top.

This open side attaches to the wing spar

The nylon corner piece was reused on the new servo box

Oh my... this actually looks frightening!

The new angle at which the cable entered the roller wheel worried me a lot. It just looked like an accident waiting to happen. The way I figured, one day I’d step on the rudder during a crosswind landing, and the wire would jump off the roller and jam the controls the worst of times. To avoid that I’d have to relocate the servo mounting bolts, the very thing I was happy I didn’t have to do.

Crap... here we go!

There are hardly any words to convey how tight a space we are talking about, or even how bad the access to it is. I must have worn a bald spot on top of my head from all the times I banged it on some part of the engine trying to get close enough to actually see what I was working on.

But it’s even worse than that. With no real plan on how to do what needed to be done, but knowing it would require some fiberglassing, I worried about how keep the loving epoxy happy, now that the hangar high temperatures were barely in the 40s (5º to 10ºC).

This wasn’t going to go down easily, but let’s look at it one problem at a time.

To begin with, I needed access to the back of the wing spar minus the aluminum heat shield, so I started by removing everything in the way until only the shield remained, then took that down as well to take stock of what I had to work with. Three of the four bolts that held the servo came from the other side of the wing spar, so I just took them off making a mental note to later seal their holes, the fourth (top left in the photo) was a click-bond that doubled as a holder for the shield, so I left it in place.

First time in 15 years this spar sees daylight

A total of three click-bonds were glued to the spar, and my plan was to repurpose the top right as one of the servo bracket anchors, then machine and install three more click-bonds of my own design (a bit larger than the ones commonly available for purchase) made out of stainless steel, and drilled.

Home made stainless steel click-bonds

Of course none of this would even begin to happen without having control of the engine compartment temperatures, and complete confidence in a reliable source of heat, which would have to prove itself first by maintaining epoxy curing temperatures over the wing spar overnight.

This required some experimenting…

Experimenting with heat lamps in the engine compartment

Testing the effectiveness of the heat lamps

While the two heat lamps did well during the daytime high temperature, with overnight lows dropping into the 30s (-1º to 5ºC) they proved insufficient, though not by much. While the temperature output of the lamps was good, the losses were just too great to be usable as they were, and the ability to retain a little more heat was necessary.

I decided to try a cardboard box folded into a corner piece, lined up with reflective aluminum tape, and a towel on the forth side to slow down the escape of hot air as long as possible, then added an halogen light below for good measure.

Covering more or less of the forth side with the towel allowed me to fine tune the spar temperature by trapping varying amounts of heated hangar air.

Heat... box?! 😬

The winning setup

Closeup of the box

The new setup seemed to work well, and another cold night proved its effectiveness.

With the final piece of the puzzle in place, I was ready to start glassing, and the first order of the day was attaching the click-bonds.

Spar sanded and taped up

Using the taped pitch servo's bracket to locate the click-bond positions on the spar

Flox over the click-bonds

Spare bracket in position held up from the one pre-existing bolt (top right).

Click-bonds needing some cleanup after curing (next day).

Next, I would be adding a few plies of fiberglass over them.

Flox sanded, and silicone tape over threads for protection.

Three plies of BID and peel-ply over the click-bonds.

Few hours later I removed the silicone tape, added the spare bracket.

Nuts were tightened slightly in order to mate the BID to the bracket.

Peel-ply removed and surface lightly sanded the next day

Later, I pulled new wires from the engine compartment through the wing spar and into the “hell hole” (space behind the passenger seat), then through a conduit buried into the fuselage foam all the way forward of the instrument panel.

Pulling through eight new servo wires all the way to the nose

Forward in-fuselage wire-conduit hole. Red thread used to pull the wires through

Next, I reinstalled the heat shield…

Heat shield back in place on the spar

… the servo…

Roll servo mounted on the wing spar

… and promptly ran into another issue.

Rudder cable link connected

Repositioning the servo to the right ended up placing it very close to the adjustable rudder cable link, and while that might have actually worked okay, I didn’t like it. I did consider stringing new rudder cables, but I didn't think it was necessary.

Since I had the rudder link adjusted all the way in anyway, I decided to make a new non-adjustable one that included stainless steel bushings, designed in such a way to remove any possibility of bolts and nuts ever touching the nylon slider.

Nuts and threads fully contained by the shuttle-like rudder link design

Bolt heads also captured and hidden by the new design

Starting machining the bolt side rudder link half

2.5D machining completed

3D machining completed

Finished bolt side half

Opposite side

Starting on the nut side

Ball end-mill working on the 3D surfaces

New vs old. Weight is about the same

The red silicone tape saves my fingers from getting stabbed by the cable end

The new rudder link proved successful, and has worked flawlessly since this servo installation.

There was still one more change I wanted to make before this project could be called completed... the switches on the stick needed some minor tweaking.

I wanted the Autopilot Disconnect switch where the Primer switch was mounted. I reasoned that in a panic situation, the loss of fine motor skills would make squeezing my fist a much easier task than pressing a hard to reach switch with my pinky. Besides, that's the same place the Boeing 767 has it, so I'm already used to it.

Prime and A/P switches needed to be swapped

Furthermore, the A/P switch needed to be changed from an ON/OFF to a momentary ON/OFF switch. While at it, I'd make the A/P disconnect button red to highlight its importance, the fuel pump switch green to mimic the fuel pump light on the annunciator cluster, and the Primer button blue, same color as the 100LL fuel.