Electrical System Finished (mostly)

I gave up onĀ  logging hours. As I don’t have a 51% requirement, it’s a renovation not a kit build, there is no regulatory need to show amateur effort percentage. Of course, other than the engine rebuild, it’s 100% my effort anyway.

Panel completed

The panel is ready for installation. All its bits and pieces are installed, the instrument harness and pitot static tubing are all connected up.

Electrical system wired up to switches

The electrical system is based on a Bob Nuckolls design. Unlike the traditional Cessna master and avionics-master, it had two separate buses with no single point of failure. This is achieved by having a separately switched cross-feed from the battery to the “Essential Bus” and the traditional contactor-driven master switch sends power to the “Main Bus”. There is a feed from the main to the essential bus through a diode that prevents back feed from the eBus to the main bus.

All that sounds more complicated than it really is. In practice the critical (or essential) systems run off the eBus and everything else off the main bus. In the event of an alternator failure you switch off the master: that’s it. This leaves only the cross-feed connected to the battery, thus all your heavy loads are off and the critical stuff is on. The endurance on eBus only with an Odyssey PC680 is over 4 hours which is similar to the fuel endurance of the aircraft.

The idea here is instead of dead alternator = land immediately (pan pan, I have 20 mins to dead battery) it becomes: continue to destination / appropriate airfield, switch on the master when on approach and land as normal. To achieve this endurance every incandescent bulb has been replaced with LEDs and the only real current hog is the pitot heater.

Annunciator wired and tested

How then do I know the alternator bit the dirt? What if the cross-feed isn’t turned on?

Annunciator – in need of labels.

It started out in the schematic drawings as a couple of warning lights and ended up with these five. It was fairly straightforward until I realized that three of them are activated by a +ve signal and the other two -ve. That means the three have no voltage until the switch is on, the other two have voltage always but should only be lit when the ground connection is made by the failed device.

The warnings and cautions are like this:

  • Alternator OFF (Red)
  • Pitot Heat FAIL (Red)
  • eBus Feed OFF (Red)
  • Starter ON (Amber)
  • Fuel Pump ON (Amber)

Once all that was wired up (on paper) I came to the conclusion that I need to be able to check that the lamp itself is working during preflight. While they are LEDs and should last nearly indefinitely the circuit might fail somewhere, better to know. I also needed a push-button for start as I’m not using the Off-Left-Right-Both-Start key switch; mostly because the one that was in it is in rough shape and I’m too cheap to buy a new one. The key-switch I used from a reputable UK manufacturer is intended as an industrial panel lock out switch and was 20% of the price of the aviation one. The mags are on a pair of toggle switches just because. It has three positions – centre is off, left momentary (springs back to the middle) is for test and right momentary powers the starter contactor.

In the test position all the LEDs are powered to make sure they are still alive!

The test position lights the lamps and grounds all the test LEDs on the fuse panel showing if any of the fuses are dead.

So far so good. The problem I discovered as I was wiring it up was that when I powered up the test circuit to make all the annunciator LEDs come on it would then connect power to the +ve signal devices so the fuel pump, starter contactor and eBus would get power via the 22gauge test circuit: not the plan.

The result was a set of diodes. I found a couple of circuit diagrams of “common cathode” and “common anode” lamp test modules online which it turns out equate to the +ve and -ve signals mentioned about above. I used KiCAD (open source electronic CAD package) to magic up a PCB and uploaded it to aisler.net and along came the solution.

Annunciator lamp test module

This allows the key switched lamp test to work and made wiring up the annunciator lights, signals etc. much easier.

Wired for pitot heat

With the addition of 12AWG aircraft wire the supply to the main bus and the wiring for pitot heat is in. I put an IP67 rated automotive disconnect at the wing root as with everything else that routes into the wings. That should make the wing installation literally plug and play.

USB Charging

With everyone using tablets and phones in flight these days there is always a need to plug them in. There will be no smoking, so no need for the cigar lighter socket with a converter jammed in and wires all over. I was unwilling to use the expensive certified charging sockets, especially as I wanted 5 charging points. The result is a “transport category” isolated DC-DC converter that turns 12V into 5V which in turn is fed to simple USB sockets.

This is all run through a single piano key, so the whole USB charging system can be shut off.

Ready to install

The picture shows why I gave up logging hours. The panel is relatively simple from the front despite its capabilities. I wanted it to look sort of period, the purists will of course object to any kind of glass, while having the capabilities and safety of modern avionics especially solid state AHRS. The complexity is all behind the scenes.

The panel is now tested and almost ready to go into the aircraft. I intend to add a door on the fuse panel to look a bit like the glovebox on the original. Once the new heating ducts are bolted in I will install and connect up the panel.

Panel Progress

The CAD file has finally been applied to metal. 6061-T6 Al .0625in thick.

The work was done at Watermark Engineering in Dublin; the owner is an avgeek who is plans-building a Fiesler Stork replica. Using their punch press we (the operator with looking on) punched out the panel and the shock-mount instrument sub-panel.

First up was the shock mount which has the multi-function display in the middle. I’m trying to keep the spirit of the original panel while adding the safety of engine monitoring and electronic AHRS. While I’m using it as an MFD, the unit (MGL iEFIS MX1) is actually a full EFIS with primary flight display, engine management and even an autopilot. The beauty of this unit is the level of configuration possible for the screens. I have designed several custom screens and started work on interactive checklists.

This is the stock Primary Flight Display Screen. It packs a lot onto a small screen and is not how I intend to use it. The checklists are able to show the relevant data directly on the list. The fuel level is manually entered at the start of the flight (as the tanks don’t have electronic sensors and I’m not opening the wings). The system then uses flow rate to calculate and display the remaining fuel in real-time.

The shock mount below is ready for paint.

From left to right will be: Airspeed, AHRS, Altimiter, EFIS, VSI, DI/AHRS (backup function), RPM (Tach). So if the 6-pack is supposed to be in front of the Pilot this is missing VSI, DI and Slip/Skid. However, the AHRS has the DI built in, so heading is shown on the DI. Thus all that’s missing is VSI and skip/skid.

The iEFIS let me build a screen that has all the engine management display, the VSI, slip-skid, AOA, and G-meter so rounding out and exceeding the 6-pack, while also allowing me to control the COM, NAV, and Transponder from the top bar of the screen.

All of that shock-mounts into the panel below which retains the piano keys and hides the fuse panel behind the glove-box door. The slot on the left is for my glasses and the (backup) paper chart I don’t tend to look at as I mount SkyDemon on an 8″ tablet on the yoke.

The punch press “nibbles” the shapes using various round and rectangular punch and die pairs. There was quite a bit of filing and sanding to do to finish it up. It certainly beat hand cutting it!

The result so far is shown below with the piano keys test fit.

If the panel looks too deep (tall), that’s because it is. There are two bends to be made along the bottom to reinforce by forming a u-channel along the whole bottom of the panel. The shock mounting points (tabs) need to be folded also. That’s beyond what I can do here, so it should be done some evening this week at Watermark.

From the back you can see the rebuilt piano key mount with new switches and the (seriously overbuilt) electrical system. Attached to the map box are a stratux (ADSB-in) and 5V power supply for 5 USB charging points to power cameras and tablets. Of course that’s about as tidy as it is ever going to look as the wiring will be a bit of a mess. Have to finish up the folding and painting before I get to that point.

Prepping – No not that kind!

I’m not a good painter. Which is why of course I decided to do it myself. However, I’m not really painting the aircraft, just patching and matching the donor tail. I installed some “curtains and I’ve ordered in some new LED flood lights, so at least I can see what I’m doing. I spent a lot of time cleaning up the H-stab which was reused from N3832V. There were quite a few dents and divots on the underside but nothing worrying. After lots of filling and sanding the parts are ready for Alodine and then paint.

Best news was finding the trim tab actuator that I thought was missing. It’s sad the things that #avgeeks get excited about; in my defence, they’re not really easy to find and very expensive!

Bending, Drilling, Cutting, and Filing

I’ve never so much as bent a piece of metal before this project so getting from zero to making panels, the new heating system and various doublers has been interesting. Since I got her I’ve been looking up at a raggedy hole in the roof of the cabin where the NAV antenna used to be attached. It doesn’t look like there was ever a backing plate and certainly not the doubler it should have to support it. I decided to patch the original hole, using AC 43.13, and relocate the antenna.

Unfortunately I didn’t save the before picture but here’s a few along the way:

Clecos are now my favorite tool! The patch has two pieces of .040″ 6061-T6, one sized to the hole and the other oversized to allow it to rivet around the hole.

The result is a flush patch. Not that I’m especially worried about drag optimization in a 48 rag wing, but if it’s worth doing … I’ll clean it up and give a lick of paint to tidy it up when I get to painting stuff in the next few weeks.

Next up was the new location for the antenna. I’ve put it just aft of the rear spar. This far back may add to “shadow” from the cabin, but I decided I liked the stronger position better. Also given the radio is in the back (controlled by the MFD) I’m able to run the coax up the inside of the control cable box, keeping it out of sight.

I put the COM antenna further back (at the last cabin frame) so it is adequately separated from the NAV antenna. Again, shadow may affect performance; if it does I’ll move it.

So, she now has a GPS puck, NAV, COM and ELT antennas all installed.

And shortly they’ll be hooked to the avionics rack.

Electrical System “Overkill”

My ILAS inspecter has stopped highlighting and emphaisising that it is critical to comply with AC-43.13 and do everything right. His new word seems to be “overkill”!

I probably went a bit mad with this, but I decided it was an opportunity to update the electrical system without being constrained by the certification system.

The existance of certified aircraft is a very good thing. It ensures that a member of the public can get into an aircraft and have a reasonable expectation that it is what it says on the tin. If I rent a C172 or a PA28, I can expect that it will behave as designed and hasn’t been hacked about. However, the unintended consequence of certification is how it has grown out of control as a self-sustaining bureaucratic exercise. The cost of certification is so ridiculously high that many advances in performance and safety have never made it into general aviation as the 1950s design “works OK and is certified”.

As a result things like electronic ignition systems are not used and single points of failure ilke the “avionics bus” persist. We’re still turning off electronic systems to protect against “spikes” which they are required to be able to withstand. This makes no sense when the EFIS or at least engine analyser must be on before startup in a glass aircraft. I enjoyed Bob Nuckolls’ book which may be purchased or downloaded from the AeroElectric Connection web site. I based my new system on one of Bob’s samples (in the back of the book) and modified it as required.

The result is a power system that has greater electrical endurance than the plane has fuel endurance.

  • All bulbs now LED
  • Strobe system (Whelan) removed and replaced by LED strobes
  • 100% of wiring replaced. Except in the wings which I’m not recovering so can’t access the wires. They are not original having been replaced about 15 years ago when the wing was last recovered.
  • Combination of main and backup battery can run the endurance bus for 4hours + in the event of alternator failure
  • Backup battery in the cabin (back of the firewall can run the endurance bus for min 1 hour
  • Avionics all remote units on a rack in the back
  • All fuses rather than breakers (reduced risk of failure at lower cost – what’s not to love?)
  • Test circuit provides test of annunciator lamps (idiot lights) and fuses during pre-flight
  • PIDG connectors everywhere in the power haress and crimped pins in the DB25 connectors to the electronic systems
  • Grounding using a “forest of grounds” on both sides of the firewall. Only the wing and tail nav lights are grounded through the body, everything else uses a ground wire back to the firewall.

3D CAD for your panel? – Recommended!

Learning a CAD package is non-trivial. I have spent many hours in Solidworks and on occasion thought to myself, “this is a waste of time, I should be working!” However, I have now found 4 separate errors in the panel that would have resulted in having to recut it. Given I have to get 6061-T6 from spruce in the USA, shipped by UPS, it comes out to about $150 per 2’x’4 sheet delivered. Then the cutting it out labour on top of that means remaking it would not be cheap!

Solidworks is very expensive – I just bought it for work and a single seat in the UK is GBP4K + support! Joining EAA gets you a student engineer license as part of your membership. I recommend it!

As I can’t get to the aircraft at the moment due to the Covid-19 lockdown, I decided to “waste” some time tracking down 3D models of instruments to stick into this model. It turned out not to be a waste of time at all. I had left plenty of space for each instrument based on the hole patterns until I tried “installing” them in Solidworks and they snagged on the shock mounts and the edge of the cut-out for the subassembly.

I didn’t have a model for the MGL Blaze unit I’m using as DI and AHRS so was going to assume it must be the same as the other instruments – they’re all standard, right? Wrong. I knocked up the Blaze part and installed it in the subassembly which required more tweaking to get everything clear of snags and obstructions.

In short – all the time spent in CAD has saved money, aggrevation and the delay that would have occurred if I had to buy more Al in from the US.

Still Here!

The work has continued and progress has been made. (We’ve had an election so the passive voice seems to be in fashion.) This is the results of moving the panel from 2D CAD (QCAD) into SolidWorks (thanks to the EAA for the low-cost licence).

Much more to come and as I’m in documentation mode, I’m planning to update this over the next few days!

Omnibus Progress

I stopped logging as it was more and more tearing things out and work etc. has been crazy. Almost everything is out of the fuselage that’s coming out. The windscreen still needs to be popped out which will hopefully happen next weekend.

All wiring, control cables, pulleys, fairleads are out of the fuselage. There is still some cleaning to be done under the floor and the cabin interior needs sanding re-priming and painting. I’m not putting back a headliner so the frames will be exposed and need to be clean and smooth for painting. The rear of the baggage compartment will need a single piece cover which I intend to fabricate from .5mm(ish) Al sheet and cover.

Aft of that the consensus is don’t paint it, lather it in ACF-50, so 4 Litres of that is on order and the weed-killer sprayer has been “rescued” from the shed.

I have not got inside the wings yet; need to build a wing stand. Next week!

Wing spar follow-throughs

Micro-camera in the wing spar follow-throughs showed some surface corrosion. I’ve pulled one of the spar mounting blocks for a better look and it may be that it looked worse on the little camera than it is. I’m going to gently scotch-brite the interior of the box section to clean it out as much as possible and then stick the camera down again.

Weekend tidy up and more windshield

Lots of tidying up last weekend. Got shelves set up and all (most) aircraft parts off the floor.

Got the remainder of the fastenings out of the windshield bottom fairing. There is a wealth of silicone. I expect that despite the lack of bolts in accordance with the STC, this thing was going nowhere ever. It is bonded to the fuselage. It looks like it will require gentle heating around the periphery to get it out. To be continued.