Sonex 1339 Electrical System

[NWade]

So last week I put in multiple orders with B&C, Aircraft Spruce, Sonex, and MGL as - despite knowing what I wanted for the panel - I kept discovering more components were needed for my electrical system... Frustrated by the wasted dollars on extra shipping fees, I sat down to plan out the whole system in detail. The result was an entire weekend spent in front of the computer, bouncing between drawing programs, Avionics websites, and supply houses like DigiKey. :roll:

But I think I've got things pretty-well planned out now. I would love feedback and constructive criticism of the design!

I want to give thanks and tip my cap to Frank Alvarez for his wiring diagram; which was a strong inspiration for my own drawing.

In general, I'm using a combination of principles from Bob Nuckolls (of AeroElectric Connection and B&C fame) and from Greg Richter (from Blue Mountain Avionics). My aim is to have a straightforward, reliable system that doesn't require stocking too many different spares or dealing with too many different types of component. So you'll note that the same gauge wire is used in lots of places and a couple of common fuse sizes are used, with an eye for protecting the wiring (not sized for each individual load; except in a few cases where the manufacturer recommends doing so).

Engine
AeroVee Turbo

Instruments

• MGL iEFIS Discovery Lite w/RDAC-XF-MAP & SP-6 Compass
• MGL V6 Radio
• Trig TT22 Transponder
• ARTEX 345 ELT (406MHz) w/panel-mounted remote switch & LED
• MGL Infinity ASX-1
  (backup airspeed & altitude instrument, just in case of total electrical failure during night VFR flight. Can power from a 9V alkaline battery)

Primary Circuit Protection

• 25A Main Breaker (prior to main bus/fuse-block)
• 10A Avionics Breaker (prior to avionics bus/fuse-block)
• 10A Lighting Breaker (prior to Nav/Pos/Strobe Switch)
• Crowbar OV circuit for Alternator w/disconnect relay
• 12 ATC-style fuses, mounted in Primary or Avionics bus (one for each switched circuit)
• 1 in-line fuse for alternator/voltage warning light

Switches

• Master switch (Batt Contactor ground)
• Starter pushbutton switch (to AeroVee Starter Solenoid)
• Alternator switch (to Alternator relay)
• DPST Mags switch (to Mag ground)
• Secondary Ignition switch (to Coils)
• Avionics Master switch (to Avionics bus/fuse-block)
• Nav/Pos switch (to Aveo Powerburst lights)
• Strobe switch (to Aveo Powerburst lights)
• Landing Light switch (to Duckworks light)
• Dimmer switch (to red LED strip-lights under panel)
• DPDT switch for ASX-1 (connects both power & ground to either Avionics bus or 9V alkaline battery)

Take care,

--Noel
Sonex #1339
Center-Stick, TD, AeroVee Turbo, Acro-Ailerons, Flush Pulled Rivets

[jjbardell]

Noel,

I have sent you a PM with my contact information. I have some of my personal thoughts which I would be happy to share regarding your design.

As for your setup, please note one important aspect to your MGL/Trig setup. Although it is advertised as ADS-B compliant, it will not be operating correctly if you purchase and install anytime soon. I have worked with the FAA and Matt (MGL) to understand why the unit is failing some of the "out" protocols. Without going into the details, the results are the need for another GPS receiver to be installed inline of the CAN BUS. This will require additional ground, power and the antenna itself. I would not buy any MGL hardware until you plan for that and ensure your unit will be provided with the new GPS receiver. They still are working on that design. Until then, you will receive ADS-B data, but your out will be inaccurate for other planes and radar systems.

Additionally, since I have the exact setup you have, I would be interested in pulling my system out of my plane (EFIS, Trig, V6, etc.) to sell you my entire system.

Cheers, Josh

[NWade]

Thanks for the offer, Josh.

And yes, I was aware that an MGL iEFIS + Trig system is not fully ADS-B Out compliant.

The surprise for me (after I already ordered a Trig from my sailplane avionics supplier) is that controlling the Trig via the iEFIS requires a special $190 adapter harness (with a built-in PCB to do some protocol conversion). Not only is there that unexpected cost, but the main Trig control head cannot be attached when you use the MGL CAN harness with a Trig transponder. On the one hand this saves panel space; but on the other hand it means you don't have a separate control unit in the panel for the Trig - you *must* use the iEFIS touchscreen for all transponder interaction. :/

As for being ADS-B compliant in the future: I fully expect that whole shebang to continue evolving as we get closer to 2020. The FAA made a hash of it, and things like newer Drones and the Navworx debacle are going to affect the market between now and then. I am not in a rush! With the Trig TT22 unit, I have a fallback position if I need to: I'll have the Trig Transponder control-head sitting around. All I need to do for ADS-B Out is to remove the MGL CAN harness from the Trig, add the Trig controller to the panel, and buy a Trig TN-70 WAAS GPS. That makes the transponder completely independent of the MGL system and fully supported by Trig as a regs-compliant solution.

Take care, thanks again,

--Noel
Sonex #1339

[Darick]

Yikes! Too complicated for me. I'll stay out of Charlie and Bravo airspace!

[sonex1374]

Noel,

I have a few comments.

1) Several places you show an 18 AWG wire protected by a 10A fuse or CB. You might want to revise this. 18 AWG wire is good for 10 amps, then it heats up to the point of potentially failing. The idea of a fuse or CB is to guarantee the failure will occur there, not in the wire. Right now, it's a toss up which will fail first. You may want to either decrease the fuse/CB rating, or bump up the wire size to the next size.

FMI: http://www.aeroelectric.com/articles/wiresize.pdf

Examples of this combo: Avionics 10A CB and 18 AWG Avionics Bus supply; NAV Light system; Landing Light system

2) The crowbar system needs to work in conjunction with a CB, not a fuse (see position 1 on the Primary Bus). When the crowbar detects an over-volt condition, it immediately shorts itself to ground. This is guaranteed to burn up something until the short has relieved itself, so the logic is to give it a nice low CB (i.e. 5A) to pop rather than anything else. Your fuse will work fine, but whenever the crowbar activates you'll need to change the fuse. Here, a CB is easier to reset rather than change out a fuse.

3) The starter solenoid is shown as 12 AWG wire. This is overkill. The solenoid itself only sees a couple amps to activate the starter. Anything small (18-22 AWG) is fine. Likewise, the 10A fuse is also overkill. I recommend an 18 AWG wire and 5A fuse.

4) The Avionics Bus is getting its power through the Primary Bus, and as such adds the total avionics load to the Primary Supply loads. The Primary Supply is protected by a 25A CB, so this means that only 15A are available to the Primary Bus. A better routing scheme would be for the Avionics Bus to get power directly from the Battery Contactor, with its own wire (14-16 AWG, or even 10 AWG to keep the wiring order simple) and a 10A CB. This also guards against a problem in your Primary Bus taking down your Avionics Bus.

5) The starter is supplied with a 4 AWG wire, and grounded with an Eng Ground Strap. I'm sure the strap itself will be large enough to carry the starting loads, but if you attach the strap to your firewall-mounted Grounding Block, all that starter load will flow through the 10 AWG wire to the grounding block, and that's too much load. You can either increase the wire size to the grounding block to 4 AWG, or connect the Eng Ground Strap directly to the battery negative terminal.

Just a few things to think about...

Jeff

[NWade]

Jeff - Thank you for the detailed review and commentary! I do sincerely appreciate it. I actually deliberately chose to do many of the things you commented on, so let me explain my reasoning below. Then feel free to tell me I'm being an idiot if you still think I'm going astray. :D

1. I respectfully disagree with your assertion that an 18awg wire will fail with a load of 10 amps. 10 amps is a lot of load for an 18 gauge wire to take; but it should not fail the wire unless the wire-run is very long or is buried in a really thick bundle of other hot cables. Keep in mind I'm not intending to run 10 amps on these lines continuously; they just have to take that load long enough for the fuse to blow.
   I'm attaching a few files to this post, including the wiring chart that comes with Aveo PowerBurst lights and shows an 18 gauge wire is capable of handling 10a over a 16-foot circuit. Additionally, in the article you already linked-to in your message, Bob Nuckolls uses a chart showing a 10a load on an 18 gauge wire over 11 feet and causing a 35-degree-Celsius rise in wire-temp. That's enough to make a wire pretty warm, but not enough to fail it. Oh, and if you'll look at the attached screenshot from the AeroVee assembly manual, you'll see Sonex themselves recommending a 10a fuse and 18-gauge wires for the coils.
   

   

   Wire Size Chart from Aveo Engineering

   
   

   

   AeroVee Manual showing 18-gauge wires

   
   ...Still, after taking in your comments and doing some online research, I've made a few revisions. I've moved the Avionics supply wire (more on that later), and I've changed the landing light fuse to 7.5a (since the 55 watt bulb shouldn't draw more than about 5 amps anyways). The 10a CB for Nav/Pos/Strobe remains, because I don't see a better compromise solution... In theory the LEDs are low-draw units (requiring a max of ~5.5 amps during a strobe pulse with nav/pos lights on at both wingtips). But the Aveo units are provided with the 20-22 gauge wire built into the lights themselves. There's no way I can put a CB in that's short enough to protect those tiny wires on the wingtips from melting while also providing enough current for the ~35 foot round-trip circuit length, and not suffer an excessive voltage drop (because of using a tiny CB - tiny breakers impose larger voltage drops). And while a fuse would get around the Vdrop of a breaker, I want something resettable so that a partial failure of the lighting system can be corrected and the flight continued/concluded with partial lights (i.e. if the strobes short-circuit I can turn them off, reset the breaker, and still go land somewhere with Nav/Pos lights working).
   --
2. If my voltage regulator fails in flight, I'm not messing with it until after I land. A fuse is a 50-cent part to replace on the ground. It seems like a bad idea to reset a breaker in-flight and expose my electronics to high voltages for a second time (if the regulator is truly failed and producing runaway voltages). I agree that a CB is more convenient; but it doesn't seem like an obvious best choice. The only time that I'd want a CB is if the OV protection was overly sensitive and causing nuisance trips of the circuit-breaker. But in that case I want to figure out why and fix it! Not fly around resetting a breaker every hour or two! :roll:
   --
3. I kind of agree with you that the starter solenoid wire seems like overkill. But then I looked at the latest AeroVee Assembly manual (they revised it again in December) and they clearly show a minimum 14-gauge wire for this operation. See attached screenshot. Since I'm already using 12-gauge wire elsewhere it was easier to design around a common wire size, rather than buy a short length of 14-gauge wire for this single run... The weight difference is pretty negligible over this distance anyways... *shrug*
   

   

   AeroVee Manual showing start switch wire size

   sonex_recommended_wiring_starter.jpg (29.51 KiB) Viewed 9390 times

   
   --
4. OK, so I had an initial reaction to your comments: The Avionics bus comes right off of a solid metal block on the Primary bus, and every circuit on the primary bus has its own fuse or CB so it should blow before the entire primary bus goes offline. Furthermore, I don't want the Avionics Bus to be powered when the master switch is Off.
   ...But then I looked again and realized that I still get all of the same robustness by running a wire directly from the battery contactor to the avionics bus/fuse-block (via the Avionics CB and switch). When the master is off the battery contactor is disconnected (duuuuhhhh), so having each bus powered independently is a better solution. Thanks!
   --
5. Ah, this last point of yours is just a misinterpretation of the symbology. I admit that it was a bit ambiguous - the program I'm using shows horizontal lines coming out of the battery symbol as part of the battery itself. So I think you interpreted that to mean that I joined the 4AWG grounding strap to the 10AWG grounding wire on the firewall. That is not what I intended to do! I've added some lines around the battery to make it clear: The 4AWG grounding strap comes straight off of the battery post! Its a separate grounding wire from the 10AWG wire that runs to the firewall. :)
   

   

   Clarified Grounding Strap Wiring

The minor changes I've noted above have been rolled into Revision 2 of the drawing, which I'm also attaching to this post.
Further feedback and critique are welcome!

--Noel

[sonex1374]

Noel,

1) 18 AWG won't fail at 10A, but that's at the upper limit of what's normally accepted. My point was to draw your attention to this. The best engineering solution is to make the structure only as strong as it needs to be, and not a bit more. However, sometimes a healthy margin of safety is applied to guard against the unknown. You as the designer get to set your own margin of safety...

Your discussion of the combo LED lights is also interesting. Is any failure of the light units likely to ever pop a 10A breaker before the unit itself burns out? If not, then that 10A CB will never get called upon to do its job. Perhaps a 7A CB would be better suited here?

2) The only time you'll have an issue with your voltage regulator is when 1) your battery is very depleted and the alternator is working extremely hard to recharge it, or 2) the VR itself is failing. For situation 1, you'll need to size your fuse correctly to prevent this inrush of current from burning out that fuse, otherwise every-so-often you'll have a nuisance trip. A CB is only a convenience here, as it gives you one in-flight reset to see if the problem was a temporary nuisance trip. If it pops again, you've confirmed the problem is more than a transitory effect, and can divert to take care of the problem. With a fuse, you lose this diagnostic tool and your only option is to divert immediately to take care of either fixing the problem or changing the fuse. For situation 2 (the main concern here) either a fuse or CB work.

It might sound like I'm pro-CB here, but that's not the case. I favor fuses, and if fact have exactly 1 CB in my entire panel. That CB is to protect the voltage regulator for the reasons I mentioned above.

3) Follow the recommendations of Sonex. Common wire sizes are good, so I get that. No harm in going bigger.

4) I think that's a better solution, but almost hesitate to offer another consideration to think about. Is a 10A Avionics CB really necessary? What fault is it guarding against? Under what circumstances would it ever do its job? Each load is protected by its own dedicated fuse that will blow long before the CB will. It's likely that the CB will sit in your panel and add no real level of protection, and actually offer one additional point of failure (a single point of failure at that).

5) I couldn't tell what you intended from the original diagram, and wanted to ensure that you weren't going to attach the ground cable to the ground block. Sounds like you're tracking. Your diagram is a bit non-intuitive here.

I hope this doesn't sound like I'm nit-picking here. Designing for simplicity and durability is tough, and you're doing the right thing by carefully considering every item you put in, and remove everything that doesn't pass muster.

Jeff

[NWade]

Jeff -

Thank you again. And don't worry, nit-picking is a good thing in this context! :D

Regarding circuit breakers: Your points are duly noted; and in general I'm on the same page as you. The reason I have an avionics CB is because I want something to protect the wiring from the battery to the avionics bus, and since the electric instruments are important to me for flight I want the possibility of resetting the circuit if it trips. Without a CB I'm stuck with either a fuse (that could take out my entire panel due to a single momentary spike) or I have an unprotected power wire that could carry a lot of current if it ever has a short.

After further consideration of the wire lengths & sizes, though, I've tweaked the lighting setup again. I decided that limiting the current that can reach the small-gauge wires in the Aveo units is a good goal. I took the CB with 2 switches ganged together on a single circuit and split them onto two independent circuits. That way I can protect the "important" lights (Nav/Pos) with a low-amp CB and also protect the Strobes with a reasonable-amp Fuse.
[It all started as a single switch on a CB, but I realized that without separate switches a failure of one LED component would continuously trip the whole circuit. So I split them out into two switches on a single CB, but that ran me into a corner in terms of amp-limits & circuit protection]

"Simplicate and Add Lightness" is hard to do, but always a worthwhile goal!

--Noel

[racaldwell]

Noel,

I am also designing my wiring diagram right now. I see you have the MGL Discovery Lite on the Avionics bus so when you are cranking the engine you will have no engine data when the engine starts until you turn on the Avionics. Does MGL require this? I currently have the Discovery Lite on my Avoinics bus design but was going to ask MGL the question if it can be on while cranking. In my RV6 I have the Rocky Mtn engine monitor on during cranking so I can see what is happening right at start up and have had no issues with it in 17 yrs and 950 hrs of use. I hope the MGL is the same.

Rick
Xenos 0057 panel planning

[NWade]

Noel,

I am also designing my wiring diagram right now. I see you have the MGL Discovery Lite on the Avionics bus so when you are cranking the engine you will have no engine data when the engine starts until you turn on the Avionics.

Rick -

I figure it takes several seconds for the engine to register oil pressure, and several more seconds for the engine to register CHT/EGT/Oil Temp. So not having that information during cranking is not a problem for me - UNLESS the boot sequence for the EFIS is a long one. I have been assuming its less than about 10 seconds; guess I'll have to put it on the bench and time it.

Take care,

--Noel