Thanks for the quick responses!  I think we have the basis for a good, educational discussion and hopefully can carry these forward to a final set of recommendations.  With regard to specific comments:

 

Hamid wrote: I think the importance of good workmanship and proper component selection has not been emphasized enough.

 

I agree entirely.  If you put a $4 Radio Shack toggle switch in to control a critical system, I think you are playing Russian Roulette.  I recall years ago talking to the shop manager at Lancair Avionics (several shop managers ago).  He had tested many rocker switches, and found most to be not especially reliable. 

 

Can some of you contribute suggestions on what constitutes a satisfactory component (like the diode mentioned) for various applications?  Can we compile some suggested work standards for the electrical work in our airplanes?  Many standards exist.  It would be great if they could be distilled by those who work with these standards every day, and assembled in one place for all of us to reference. Regarding the question of panel switches, I followed Brent Regan’s lead and put in high quality, environmentally sealed Honeywell/Microswitch sealed toggle switches.  (Eaton makes a similar line.)  These are qualified for use in tanks, Humvees and other vehicles operating in awful environments.  Not cheap, but since switches represent potential single point failures, I believe their selection is wise.

 

Hamid also wrote: Looking over your design, I see one issue: Will the bus 1 to bus 2 diode keep you from reliably getting the low voltage warning on bus 2?

Beats me, Hamid, I am just a dumb mechanical engineer.  What do your think, and what would you recommend?

 

Also, to all reviewing my dodgy sketch, you missed that I forgot to insert the second diode to feed the avionics bus.  See the missing diode now?  (My sketching workmanship standards need improvement.)

 

 

John wrote: If one uses the B&C voltage regulators and the circuit therein to power a caution lamp, it would be adequate to detect a pending low voltage on a buss. No? I'm not sure how the Chelton gauges are wired, but they should provide voltage readouts from the two busses, and probably current as well.

 

Sounds right to me, John, but I am not specifically familiar with the B and C hardware and not qualified to evaluate it.  I can tell you what I have done.  Based on recommendations received, when I upgraded my early generation (circa 1999) Chelton hardware I tossed the EAU (engine monitoring unit) and purchased an Electronics International MVP-50.  Among many other things, I have it configured to monitor electrical load on both the A and B buss (amperes being drawn by the aircraft in each system) and voltage levels on each buss.  The unit permits one to set trip levels for alarm, and the alarm is both audible and visible.  As long as the cross feed is not engaged, failure of one alternator should result in low bus voltage and an alarm.  But because I have one little alternator and one big one, the subsequent response strategy and load shedding depends on which alternator continues to operate.  Once I understand the fault and shed load, I will head for the ground.  Note that Columbia (and probably Cirrus) have two big alternators. 

 

John also wrote: One philosophical slant for me is the use of the essential and avionics busses. They complicate the design and increase the parts count, with their consequent effect on reliability. With a reliable crossfeed contactor, you can have the same benefits of both of these busses. The only thing for the pilot would be to manually reduce the electrical load as required.

 

I understand your point, but suggest you consider the following: What about the case of a cross feed contactor (or its switch or its wiring to and from the switch or contactor attachments points….) failing?  A single point failure and poof- benefits evaporate.  Would you test the cross feed system at each run up to make sure it works?  Even if it does, a subsequent single point failure in the cross feed hardware, and you have a problem.  Consider also a bad short on one buss that takes it off line.  If you hit the cross feed as your first response you might bring down the second buss as it leaps into the short circuit.  Now the airplane is dark and everything turns off.  Oops.  Consider this and other possible failure modes.  A complete “failure modes and effects analysis” (FMEA) considers literally hundreds of combinations and permutations of this type.  Today it is done for FAR 23 certification.  Most of us are not capable of doing such a thorough job.  So we should steal good designs from others if we are confident that they are “better.” 

 

Scott (and via Scott, Klaus) wrote lots of good stuff including:

 

• It was reported on the Lancair Mail List that Light Speed Engineering is no longer shipping/selling dual ignition systems and only recommends a single system with a mag backup because of too many experimental aircraft's unreliable electrical systems.  To which Klaus replied: There was a concern for a while when we had a few sensor failures on the Continentals used on Lancairs. The engine builder would set up the proper clearance per our instructions, test run the engine and deliver it. The proud owner would then install his own baffling between the case and our mounting bracket. This would reduce the clearance to the point that the sensors could get damaged.

I made an assumption as to the cause of the caution (faulty electrical systems).  I have to remind myself how to spell ASS-U-ME.  L  I am pleased that the problem has been isolated and corrected.  Apparently my Continental engine was caught in this cautionary period, and the engine builder put on one mag and one Lightspeed ignition system.  Now my engine is making its way across the Pacific this way.  Alas.

 

• Elder* Cessna aircraft (I have some experience) use crappy components and connectors and are frequently full of architecture and design shortcuts used to save MONEY, not lives.

Absolutely correct, in my direct experience.  I could not believe the crap that Cessna put in my brand new TR-182 when I bought it new in 1979.  That includes switches, circuit breakers, and other important electrical components.

 

• Sorry, I cannot trust that the FAA knows any better than an experimental aircraft builder that actually built his/her airplane and studied available information on how-to-do-it.

Ah, Scott, you are being cynical, but your cynicism is not misplaced.  I agree that a builder familiar with the topic should be able to build an electrical system equal to or better than that specified by the FAA.  But I am a little familiar with some current certification processes via the CAA (Australia’s equivalent to the FAA).  CAA follows FAA on FAR 23, and in my recent discussions with CAA I learned about requirements that manufacturers present a quantitative safety case that subjects system designs to thorough data validation and mathematical analysis.  This is a post “elder Cessna” phenomenon, and probably more than we as homebuilders can hope to achieve. 

 

I merely suggest that FAA-approved designs for MODERN aircraft provide a good STARTING point.  If you are limiting yourself to day VFR, your standards can (should) be relaxed.  But since some of us are apparently flying night, IFR, single pilot, over the mountains at high altitude, I suggest we get the best input we can obtain.  Hence my call for an exchange of thoughts and preparation of some recommendations and example workmanship standards that can be used by all to improve the quality of our projects.

 

• One of the things the builder cannot fix with hardware is the loss of life from serious lapses of judgment, training and knowledge in the other single-point failure, the pilot.

Excellent point. We should ever lose sight of the fact that most of the time pilots kill pilots, not airplanes.

 

Let the adventure continue!

 

Fred Moreno

AKA Captain Tuna, Chicken of the Skies