Step 12.5, Wiring continued

Wiring an airplane is a test of dogged determination. Or should I say uncommon perseverance. I guess they're really the same thing.  But this is just a thing that begs saying it twice. Perhaps it's a bit of an exaggeration.  I think not.  It seems that this phase of the project is destined to continue for centuries.

Does this explain why airplanes didn't exist before the 20th century?  Might they have been begun in the middle ages and only finished recently?  No one knows.

One thing I am certain of is that I'm wearing the gloss off of my formerly shiny shop floor wandering about looking for the wire stripper or the crimper.  Not that one, the one that fits these special connector pins.  When I finally get the right tools in place, I forget which pin I'm connecting, spawning yet another trip across the shop to recheck the already thoroughly rechecked plans.

When I get past this phase I'll be crowned the Ultimate Wiring Champion and perhaps have my own reality show.  Or not, either way I'll be finished with this and moving on.

It is not particularly difficult as there really are no inaccessible places or stubborn fasteners to encourage frustration.  It's just a lot of checking, rechecking, making labels, and crimping connectors.

Then there is the oft' repeated episode where I forget to slip on the heat-shrink tubing before the connector goes on.  Some may remember a very similar difficulty I faced when I would forget to slip on the B-nut before flaring the ends of the fuel and brake lines.  Arrrg!


Here's a look at the most common crimp connector pins that I'm using. It surprised me that I was able to assemble all of them in one place for this photo.  These boogers have a knack for staying out of sight when you need them.

A wing makes a handy work surface / schematic holder

And so it goes.  One wire at a time.  When I finish this page, I'll just have 6 more to go.

Step 12.4, Home built avionics

Just when it looks like I might be making some actual progress on this project, I find a new way to slow it down.  One might reasonably assume that I've missed my calling;  That with a skill such as this, I could have been a mid-level manager.  Perhaps.  But I digress...

The most recent impediment to my project's completion is the idea that I should make some of my own avionics.  I'm not talking about anything big, like a radio or auto pilot.  Something really small, I tell myself, should be no trouble at all.  Like an annunciator panel or relay deck.  How hard could it really be?  They're so small.

With this truly terrible idea rattling about in the back of my consciousness, I then compounded the injury to my project by inventing several new solutions that have yet to be matched with actual problems. In all, I've identified 5 circuit boards to make.

The screen on the right shows the circuit layout underway for a few of them.  Clockwise from the top:  annunciator panel, relay deck, and two copies of the control column serializer.  The screen on the left is my instrument panel which was also being designed at the same time.

Now without further adieu.  The five boards my project can not live with out are:

Annunciator Panel:  This is a set of lights that signal changes in various conditions within the aircraft.  There are six conditions monitored by my annunciator: Master warning,  Master caution, Low oil pressure, Low voltage, Canopy not locked, Auxiliary fuel pump on.  The six lights are also push-buttons that are monitored by a microprocessor whose principal responsibility is watching the input signals and updating the status of the lights -- on, off, or flash.  The big advantage of having a microprocessor do all of this is that it is more complicated.

There are some people who think that 8 indicators are the absolute minimum for an annunciator panel, but I've settled for only six because somehow six seems less ostentatious.

WIG/WAG Controller:  A potential problem at an uncontrolled airport is an aircraft or other vehicle pulling out onto the runway when you are about to land.  One way to reduce the likelihood of this occurring is to increase your visibility from the ground.  The wig/wag controller does this by alternately flashing the taxi and landing lights.  A microprocessor interprets the state of three switches from the instrument panel, Landing light, Taxi light and Wig/Wag enable and then sets its outputs appropriately to control the landing and taxi lights.  

Relay Deck: A relay deck is just a collection of relays.  My implementation of a relay deck contains 10 relays and, --- wait for it --- a microprocessor.  Relays such as these are typically used to isolate low current instrument panel switches from higher powered devices like motors. In my case, I need the relays for the trim motors, flap motor, and the push to talk switch.  In addition, I wanted to reduce the size of the wire bundle going down through the control columns by serializing the data emanating from the stick mounted switches.  The serial data input function (on a relay deck) is not available commercially as far as I am aware.  My relay deck has three data channels for input.  One serial channel each for the pilot and passenger, and one parallel channel that permits panel mounted switches to get in on the act as well.  All of this allows me more flexibility on how and when the relays should be activated and by whom.

Control Column Serializer:  This board reads the control column switches that are mounted within the grips and converts this data to a serial data stream going to the relay deck.  The data are packetized and a CRC is computed to validate the data's integrity at the receiving end.

Fuel Capacity Transducer: This board is required to translate the fuel tank's quantitative transducer, which is of the capacitive type, to an analog voltage that the Garmin EFIS (Electronic Flight Information System) can then use to then display the fuel quantity on the PFD (Primary Flight Display).

Pictured below are the 5 bare boards listed above, plus a sixth (bonus) board at right that doesn't do anything -- well, it doesn't seem to do a lot because it simply translates some signals from the Honeywell AML34 instrument panel switches. (how useful could it really be without a microprocessor?)

I've always felt that any problem worth fixing can be made more complex by adding a computer to it's solution.  It's a kind of occupational hazard for me.  I just can't help myself.

At the core of each of these boards are 8 bit micro controllers of various types.  Having a computer right there surely invites feature creep which is the enemy of CPS (Completed Plane Syndrome), so I'll have to watch that.  I've already added the ability for a couple of these boards to listen to the Garmin EFIS RS-232 serial data stream.  Whether or not I'll do anything with that data source remains to be seen. At the very least, it will be something I can play with after I finish the plane.

At this point I've built the 5 boards and have tested 4 of them.  A lot more software will have to be written before they are finished, but at this point I'm just trying to verify the correctness of the hardware before I move on.  I'll have more to say about them in upcoming posts.  By then I will have had a chance to learn some new acronyms.  TTFN.

Step 12.3, Designing the panel

It turns out that making the instrument panel is not such a hard thing to do, but there are a few questions that must be answered first.  Probably the most important question from a monetary point of view is whether or not the panel will be IFR certified (Instrument Flight Rules).  The cost of an IFR panel is largely dictated by the FAA requirement that the GPS and or VOR  and or localizer receivers be certified to operate within the national air traffic system.  This alone can easily add $10K to the cost of the full featured Visual Flight Rules panel.

With that in mind, it looks like I'll be doing a VFR panel, but what exactly does this mean?  A panel equipped for VFR flight means the aircraft is instrumented for flight under Visual Flight Rules. Basically, the aircraft can be operated only under clear weather conditions. This does not seem to be such a large impediment to me, since I don't have a burning desire to fly in inclement weather nor am I certified to do so.

Next question, how will the panel be in instrumented, that is, what style of instruments to choose?

The most basic choice is between a Glass panel or a panel with steam gauges.  Despite the confusing name, a glass panel is not actually made of glass.  It simply refers generically to an instrument panel that displays its data on some kind of screen, usually an LCD like a computer monitor.  Likewise, steam gauges do not actually measure steam pressure.  The term 'steam gauge' refers, in a somewhat derogatory fashion, to the round dials that traditionally adorn the aircraft instrument panel.

One of the big advantages of a glass panel is that all or most of the instruments are displayed on a single screen.  This can simplify the panel and reduce weight.  It also permits new functionality to be added to the display without having to physically change the panel.  On the other hand, one big advantage to the steam gauge approach is that if the power should go off, the steam gauges won't know it -- they don't need power to operate.

Alas, I chose the glass panel approach because it is the more modern way, and it offers the most flexibility going forward.  Still, I'm not entirely comfortable with the little problem of power or the accidental loss of of it, so I will be adding a separate ASI (AirSpeed Indicator) and altimeter, both steam gauges.  

Deciding the location and number of switches and their function was the hard part for me.  Since I had already decided the number of displays I wanted (2) and the brand that I would be using, Garmin, a lot the functionality has already been decided for me.  What remains mostly reduces to turning on or off lights or avionics.  To get started, I made a list of the things I thought I might like to control:

• Master power / Alternator field enable
• Engine start / mag select -- key switch
• Auxiliary fuel pump
• Avionics master
• Autopilot Enable
• Control Column passenger enable
• Strobe light
• Navigation lights
• Taxi light
• Landing light
• flaps
• pitch trim
• roll trim
• Wig Wag enable
• Pitot heat
• Seat heat
• Dome light

To arrange the switches I followed the idea that the switches should be placed in roughly the order they might be used in a typical flight, left to right.  This order makes about as much sense as any other, with the added advantage that this order will utterly befuddle any Hebrew, or Chinese plane robbers :)

With all that settled, I then laid out the panel using online software from frontpanelexpress.com. Their software allows one to import the panel outline, which I downloaded from Van's in dxf format. This way I could be sure that the panel would fit properly.  I was also able to get some pre-made 'macros' (FrontPanel express terminology) for some of the switches and instruments online.  One RV7 builder in particular, Brian Chesteen, was very helpful.  After that it's just a matter of placing the components, or rather, the cut-outs for them.  There is a lot of measuring involved to get the sizes of the switches and other components properly specified.  A fair amount of guessing is required.  Like how much clearance is needed over here or how much to allow for powder coat over there.  And so it goes...

When the design is finished, the front panel express application totals up the machine time it will take to cut the panel and how many tool changes are required.  It then gives you the bad news, which of course is the price. 

This is the point where I returned to the design to try and reduce the number of tool changes, as well as reducing the sizes and the number of labels as well.  After a few iterations, I got the price down some and simplified the panel a bit.  Both of which are good things.  It is a simple matter to place the order right from the application.  All they really need is your credit card and shipping address. Everything else is specified in the application while you design the panel. The cost of the panel includes infilled engraved labels which are done by the CNC router along with all of the other machining.  Sending this work out probably saved me a month of work.

And the finished panel?  Glad you asked:

My next task is to begin attaching the instruments to the panel.  A bit of filing was necessary here and there. but there were no show-stoppers.  I made adjustments to the design file, so if I ever need to order another it should be perfect.

Bracketry to hold the radio stack.

AML34 switches by Honeywell.  Custom engraved caps by engravers.net

Once all of the switches are in, there is still a lot of wiring to do on the back side.

And here is the nearly completed panel being test fit.

Very briefly, the two LCD screens on the left half of the panel will display the flight data and moving map, respectively.  In the center are the backup ASI and altimeter. To the right is the radio stack containing (top to bottom) the auto pilot panel, the audio panel, and two GTR-200 COM radios.  The big hole to the right of the radio stack is the glove compartment.  Finally, across the bottom are the aforementioned switches.

A cover plate below the second radio allows room for future expansion.  The radio stack was sized to allow a GTN-650 NAV/COM/GPS radio to replace one of the GTR-200 radios + cover plate.  This will convert the panel to a fully certified IFR panel.  The panel is ready -- I'm just waiting to win the lottery now.  And if the lottery doesn't pan out I still have a long lost relative from Nigeria that died and left me a pile of cash.  I just need to pay some expediting fees and...

The one item on the panel that was not previously mentioned is the annunciator panel. It's the row of 6 colored switches located just above the second LCD display.   That will be the subject of an upcoming post.

Step 12.2, Trim and Pitch Servos

At this point in the build I'm installing boxes and getting ready to begin wiring.  The last two servos to install before beginning the wiring are the roll trim and the pitch servo.

The roll trim servo allows one to compensate for lateral imbalance.  This could be caused by uneven fuel use between the two fuel tanks which are located in the wings or even just flying without a passenger.  In the extreme, one tank completely empty and one full, the roll torque applied to the aircraft would be on the order of 21 gallons x 6lbs x 4' = 504lbs!  That's quite a lot of force trying to roll the airplane.  Those interested will note that aviation fuel weighs 6lbs per gallon and the 4' figure is just an estimate on  how far it is from the center line of the fuselage to the center of the fuel tank. Because the passenger sits so close to the center line of the aircraft he makes much less of a impact to the balance:  A 200lb passenger x 1' from the center is only 200lbs roll force applied.

Whatever the cause of an imbalance, the roll trim servo is designed to apply and opposing force by moving the ailerons in the opposite direction.  Fortunately for the servo, it doesn't take that much force to move the ailerons -- very little force at all.  Partly because of mechanical advantage in the linkage, and partly because of aerodynamic forces acting on the ailerons.

Because not much effort is required to trim the airplane, the roll trim servo (pictured above) is not so much larger than a large RC model servo.

The next servo to be installed is the pitch servo which is controlled by the autopilot.  The pitch servo moves the elevator via a connection to the elevator push tube.

The control arm length is adjusted  such that the servo is at its center position when the elevator is in its neutral position.

The pitch servo is installed behind the baggage compartment.  Pictured below is the servo and the push-tube that connects the elevator to the control column via the elevator bell crank.

With the pitch servo installed the next step is to attach it to the bell crank with the control arm.

Now the wiring begins.  Having the fuselage on the rotisserie makes accessing the interior a lot easier.

Pulling all the wires is a pretty big job which is made more difficult by the very small spaces that must be accessed.

At this point in the build I found it necessary to stop pulling wires momentarily to step back and complete the airplane's overall electrical schematic. Although proceeding without a schematic is much faster, it will inevitably lead to me leaving out some necessary circuit.

And so, a week or two elapses without much apparent progress as I pound out the schematic in my office.  Its a good time to think through how the various systems tie together.  Fortunately, I don't have to figure out everything myself as I bought some of the harnesses for the Garmin boxes from Stein Air to save some time.

But a lot of the wiring goes to the panel which has yet to be designed.  I can't finish the schematic until I finish the panel designed.  Next time: It's panel time.

Step 12.1, Fairing the canopy

When the canopy is attached to the frame the canopy's edge is flat against the wind.  A fiberglass fairing is constructed to smooth the transition from the cowling to the canopy providing aerodynamic as well as aesthetic advantages.

The process is simple:  keep adding fiberglass layers until the transition forms a smooth arc.

The loose fabric visible on the top is not fiberglass.  It's called peel ply.  The objective of peel ply is to protect the surface of the curing fiberglass when it is planned to add additional layers.  When peel ply is used it is not necessary to prepare the surface with sanding or cleaning prior to bonding the next layer.  Just peeling the peel ply off leaves the surface clean, and the texture of the fabric gives sufficient bite to adhere the next layer of glass.

The first few layers are made black with dye because the fiberglass overlaps the canopy edge and is visible from the inside.

After five or six layers the transition is beginning to flatten out into a nice curve.

The profile of the curve is made uniform with the aid of  the reverse side of a sanding block. 

The thick black tape protects the canopy from the sandpaper.

And finally, the canopy in its finished state.  Note the tie down strap.  It's keeping the canopy from tipping over and falling off of the roll around cart while I finish off the underside.

Step 12.0, Gluing the canopy

Although I've taken the summer off from blogging, I've continued to make progress on the project. This step in the project occurred in late June.

With the canopy cut and the edges smoothed, its time to begin the process of attaching the plastic to the aluminum.  There are two main methods of attachment, glue or screws.  The Van's way is to use screws, which is the simpler of the two as it allows one to fix one point at a time.  The glue method, on the other hand, requires that the entire canopy be fixed at once.  The glue method does offer one important advantage in that no holes are drilled through the plastic.  It has been observed that canopies with cracks tend to begin cracking at a drilled hole.  And thus the genius of the glued canopy is revealed -- no holes drilled equals no place for a crack to form.

Before gluing the canopy on I thought it a good idea to paint the cowling under the canopy black to reduce reflections.  After the canopy is in place, the clearance is too limited to attempt painting in there, mostly due to the slope of the canopy.  And so I mask off the area to be glued and paint the rest black.

Next we test fit the canopy one last time and fabricate some mechanical tie downs that we can rivet in place to hold the canopy while the glue cures.

I use a wire to act as a spacer in between the canopy and the side rail. This ensures that there will be enough glue in place for a strong bond.  The danger here is that without the spacer, the canopy could contact the rail an squeeze out the glue, preventing a good bond.

The next step is to apply the primer to the plastic and the aluminum.

Then the real fun begins.  Squeezing out the glue.

During the gluing process it is necessary to get inside to scrape off the excess glue.  Getting back out is the hard part.  Here I am trying to climb back out through the baggage compartment.

With the glue on and the canopy in place there is nothing to due but wait.

I really enjoy these moments in the construction.  Applying glue brings a kind of clarity to process. There is 'before glue' and there is 'after glue' and nothing in between.  Black and white.  Once you've mixed the epoxy or squeezed out the glue there is just no going back, It's all ahead forward.   It's at these times that one really knows where they are.
  

It frequently happens that people ask me how I'm doing on the airplane.  The engineer in me struggles with this kind of question because I feel obligated to give an exact answer.  Most of the time I have a half dozen different subprojects going on simultaneously each involving many subparts and their attendant issues.  I don't really want to stop and calculate the precise level of completion that the question requires -- partly because I always realize that I'm not as far I along as I had previously thought, but also because the questioner would probably get bored and walk away while I mentally tabulate the answer.

In the end, I know that the questioner really doesn't want that kind of detail so I usually politely answer, "Oh, its going fine."

But on days like this, when I've completed a major milestone, I'm happy to add something like, "I just attached the canopy."

Curiously, I failed to photograph the canopy as it looked while the glue cured, so I add this one which gives a preview of the next step in which we fiberglass the edges of the plastic down to the aluminum cowling to form a smooth fairing.

Why does this dog keep photobombing my blog?

Step 11.9, Cutting the canopy

First there was a big bubble of acrylic plastic.  The canopy, we'll call it.  It's a single piece of blown plastic as it comes from Van's factory.  It must be cut in two -- one to fit the tilt up portion of the canopy and the other, the back window.  So the trick, as any sculptor will atest, is to trim off the bits that don't look like the artwork.  An initial trim is required so that the big bubble will set over the fuselage in something close to its final position.  Once this is accomplished, more accurate measurements can be taken to effect the big cut.

So I eye-balled the initial trim and got pretty close.  Here is the canopy is back on the aircraft where I can mark the location of the big cut.  A mistake here could be very costly, somewhere north of $1500, so I am uncharacteristically careful to double check my trim lines.  

Back off of the aircraft, the line to cut is clearly visible.  All I need now is the nerve to plow into it.  

I tape across the cut as I go to try and keep the stress on the plastic at a minimum.  There have been a number of builders crack their canopies at this stage.  The actual cut is with a die grinder sporting a thin cut-off blade.  I don't have any pics of that as I was alone -- principally, out of concern for the safety of others should I make a mistake :)

And finally the back window is separate from the canopy.

A very relieved selfie

Now the edges are sanded smooth.  #120 to #200 and then down to #400 grit.

It's a bit of work, but the edges are smooth and straight.  Now back on the aircraft for additional measuring.

And some fine tuning of the forward edge and sides.

At this point the plastic work is done on the canopy.