What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Critical bus/wiring EFII type system

Freemasm

Well Known Member
I'm a bit hesitant to open this one. People tend to want to validate their own choices.

There’s a lot of different electrical architecture philosophies as everyone knows. For Electrically Dependent aircraft, the design approach broadens even more so.

Nuckolls = (Deviations of) Parallel Direct wire to Battery bus(es) for Electronic Ignitions. Nothing with a coil between bus and critical components (attachment 1). This was a change as earlier approaches had diode bridges and cross-feeds most anywhere in the system.

CF and others = Isolatablity (if such a word) of major electrical components for safety and troubleshooting. Failure of any one component/branch should not require pilot action to continue safe flight etc. (attachment 2)

EFII = Diode bridge cross feed and parallel relay/switch, etc. Also an ECU test function for pre-flight check of ECU. I’m not considering nor debating their bus manager.(attachments 3 & 4) though I may add a version of their automatic (but isolatable) fuel pump activation on falling downstream manifold pressure. Not sure this would bring that much benefit (see swtich placement philosophy, next paragraph.

Please feel free to shoot holes in the preliminary philosophy in attachment 5. My though was to try and gather the aforementioned approaches, save some panel space and weight, create a no-thought/automatic pilot action (all related engine switches “up”) should the situation arise. The Nuckolls Z-28 parallel path simplicity is upset by the coil and injector power singularity; hence the diode bridge. At that point, why not size it for teh full critical load?

Trying to capture the parts I like from the various experts but I gotta draw the line somewhere. I know there’s no perfect solution/have to get to manageable/acceptable risk level, etc. Am I off or not seeing a SPF. I need to know this part of my approach so I can get my firewall and related components laid out. Thanks.
 

Attachments

  • attach 1.jpeg
    attach 1.jpeg
    95 KB · Views: 135
  • Attach 2.jpeg
    Attach 2.jpeg
    75.6 KB · Views: 131
  • Attach 3.jpeg
    Attach 3.jpeg
    58.1 KB · Views: 124
  • Attach 4.jpeg
    Attach 4.jpeg
    130.2 KB · Views: 136
  • Attach 5.jpeg
    Attach 5.jpeg
    53.4 KB · Views: 114
Last edited:
Scott,

Your write up almost looks like a prompt on a system design exam! ... there are clearly a lot of considerations and you pose many questions. And like you said at the end, it may not be possible to get everything you want in any one system, even combining the best of many.

I am a fan of Nuckoll's plans. My -7 runs a Z-19 with fuses (yes, fuses like you see in your car) with EFII ignition (two batteries, one alt). Can't say if it's the absolute best system ever designed (I have no gripes with it) but it's been running flawlessly for 150 hours and I didn't spend tons of time trying to design my own.

Just my two cents ... hopefully someone else can shed some more theory here, I"m sure they will.

Cheers
 

Attachments

  • 20210310_154415.jpg
    20210310_154415.jpg
    375.7 KB · Views: 113
  • 20201227_183052.jpg
    20201227_183052.jpg
    320.1 KB · Views: 104
  • 20201012_183938.jpg
    20201012_183938.jpg
    357.6 KB · Views: 159
The circuit breaker protecting the alternator "B" lead is single point of failure. It could be replaced with a current limiter.
Once the engine is running, the alternator powers the aircraft electrical system, not a battery. There are 3 unnecessary failure points in the alternator circuit: circuit breaker, contactor, and relay. Simpler is better. Many have tried to improve upon Nuckolls' architecture, few have succeeded.
 
Scott,

EFII adds a lot of issues that need to be carefully reviewed and if needed, mitigations developed.

The quick blush on your diagram #5 the one thing that jumps out is you provide isolation from each battery to feed a single buss. This tends to defeat a design opportunity available with two batteries. Why not have each battery feed half of the engine (one ECU, one Fuel pump, etc.)? I assume the engine will run on half of the stuff, but if not then I’m afraid you are left with what I consider unnecessary risk.

Side note - don’t forget the rest of the plane. Keeping the engine running is a good idea but if a single failure takes out your panel in IMC you have problems. Electrical designs must address the airplane as a system, which leads to not just adding stuff to address single issues. Example is the tendency to add a backup battery for everything.

Other side note - those who have designed in a backup mode of running the ship with just the alternator (no battery on line) I recommend you carefully test this mode to make sure it works.

Carl
 
Scott,
The key to good systems engineering is to really understand the mission, know the level of human intervention needed to change the configuration in the event of a component failure, the level of system monitoring required and the ability to do a complete and thorough failure modes and effects (FMEA) analysis with realistic component failure rates. Looking at your power distribution diagrams I wonder if a two generator, two battery architecture with no switch over would provide overall higher levels of availability with far less monitoring,switching and risk of bringing the whole system down. I did look at EFFI ignition very seriously but in the end went with dual PMags. The complexity and overhead for the electrical system drove me away from EFFI. I have a 60 amp main belt driven alternator and a 30 amp gear driven alternator for an IFR configured RV-7A. The end objective is to be comfortable and confident in your chosen configuration and be happy that it will meet your mission needs.

KT
 
Well

“…The quick blush on your diagram #5 the one thing that jumps out is you provide isolation from each battery to feed a single buss. This tends to defeat a design opportunity available with two batteries. Why not have each battery feed half of the engine (one ECU, one Fuel pump, etc.)? I assume the engine will run on half of the stuff, but if not then I’m afraid you are left with what I consider unnecessary risk…”

In the EFII system, each ECU controls half of the plugs all of the time. This means that an ecu failure will not cause the engine to quit due to ignition loss.

Because there is only one injector per cylinder, only I left ecu controls the injectors at any one time. An ecu failure could cause engine stoppage if the failed ecu was in control at the time. Switching to the other ecu restores injector control.

As there are three coil pack for the six cylinder engine, you can’t split the power in half, however, powering all three coils from a single bus from two independent sources provides a way to accomplish the same result.

The fuel pumps could be separated into two separate power sources.

There are several threads discussing this topic and you will find every opinion you can think of in them. Do you research, make your choices, and build what you want…
 
And, the current limiter then becomes the single point of failure.
True, but current limiters don't have contacts or moving parts like circuit breakers do. Bob Nuckolls now recommends a fuselink instead of a current limiter to protect the alternator "B" lead from battery current.
 
One advantage of using SDS (at least on 6 cylinder models) is to have completely separate power sources to feed one ECU, fuel pump, coil pack and allow fuel injector power and triggers to be switched to one ECU and one power source or the other through the injector relay box. Don’t think you can do that with EFII and their three coil packs.
 
Actually

Actually I think you probably could since the only single point item would be the injectors and the ecu select switch changes injector control from one ecu to the other. If the ECUs were powered separately, you would have essentially the same thing…
 
Manual versus automatic switching

Actually I think you probably could since the only single point item would be the injectors and the ecu select switch changes injector control from one ecu to the other. If the ECUs were powered separately, you would have essentially the same thing…

I am sure there are a number of configuration that would allow transfer of fuel management from one ECU to the other - from simple manual switching to more complex automatic control. Simple control through a manual switch is likely to be more reliable with an overall lower failure rate. The question is - in a critical high workload situation, like on final approach is the pilot going to take the correct actions? This gets right back to previous arguments about all electric configurations in simple single engine GA aircraft. Its the same discussion about fuses, circuit breakers and electronic systems like VPX - its perception, personal choice and knowledge.

KT
 
@Mich48041 "The circuit breaker protecting the alternator "B" lead is single point of failure. It could be replaced with a current limiter.
Once the engine is running, the alternator powers the aircraft electrical system, not a battery. There are 3 unnecessary failure points in the alternator circuit: circuit breaker, contactor, and relay. Simpler is better. Many have tried to improve upon Nuckolls’ architecture, few have succeeded.


On the M side of FMEA = a big yes. The alternate approach is to minimize the E. The two effects are linearized in the process (too much so, IMO) but it’s still a valuable approach. Losing an alternator for any reason (there are many) shouldn’t affect safety as a bare minimum. (Dual) redundancy minimizes the SPF effect. I won’t insult the Cult of Nuckolls. I’ve learned a lot from his references. He’d probably agree that any design can be improved upon; even his own by none other than himself. I disagree on your point about alt vs battery as “the” power source. It is a dangerous assumption and I’d argue it’s better to conceptualize it the other way. Even if the system were to survive an associated test with a faulty battery, different operating parameters may get you a different result the next time. Similarly, a simple change in load (flap actuation/radio xmit/etc.) could take everything down. Would love to be wrong here.

@Rocketman1988. I’ve studied your drawings as well. Yes they are safe but it’s my nature to ask questions and try and improve things. I owe you some cold craft brews if we ever meet.

@Carl. “The quick blush on your diagram #5 the one thing that jumps out is you provide isolation from each battery to feed a single buss. This tends to defeat a design opportunity available with two batteries. Why not have each battery feed half of the engine (one ECU, one Fuel pump, etc.)? I assume the engine will run on half of the stuff, but if not then I’m afraid you are left with what I consider unnecessary risk.”

Can easily Z-28 it but as mentioned (not clearly), the diode bridge for coils and injectors is needed IMO and referenced in the OEM’s approach. They don’t draw current unless powered by one of the ECUs, BTW. Since the dual feed was there, why not feed a common critical bus as well? That was indeed the basis of my question. Yes, as is later pointed out, you can run off of a single ECU, Fuel Pump, coil. I’m only considering the critical system at the moment as I’m trying to get my FW componentry laid out. I am considering the overall approach/architecture but this is the first bite the elephant. My overall architecture will probably look very similar your attach 2. I owe you a few as well though IIRC, it would be Bourbon versus craft brews. You can take a man out of the NAVY but…..

@keitht. I am of complete understanding. There is no single best approach or there would be no discussion here. Your architecture is probably 90-95% identical to what I’m contemplating. There’s a reason why I am where I am. Not important here but the question has been asked by more than one person.

@MEIstien. Got your PM want to learn more. No that wasn’t me. I’m basically logging zero hours these days. Will write or call tomorrow.

@Katanapilot. “One advantage of using SDS (at least on 6 cylinder models) is to have completely separate power sources to feed one ECU, fuel pump, coil pack and allow fuel injector power and triggers to be switched to one ECU and one power source or the other through the injector relay box. Don’t think you can do that with EFII and their three coil packs”

It’s appreciated but not relevant for me here. I could design such an architecture but it swims against the KISS current. The electrical system reqmnts and functionality are pretty close.

@KeithH (again) I am sure there are a number of configuration that would allow transfer of fuel management from one ECU to the other - from simple manual switching to more complex automatic control. Simple control through a manual switch is likely to be more reliable with an overall lower failure rate. The question is - in a critical high workload situation, like on final approach is the pilot going to take the correct actions? This gets right back to previous arguments about all electric configurations in simple single engine GA aircraft. Its the same discussion about fuses, circuit breakers and electronic systems like VPX - its perception, personal choice and knowledge.

Late comer comment. Detection would be the tough part; automatic switching the easy. My KISS thoughts were a default “all engines switches = up” verification for the engine critical buss stuff and “change state” for the ECU control switch. As of now, my thoughts were to position that switch horizontally above/between the ECU power/test switches. It would fit my mindset for the related functionality. I can see this thought being a bit controversial.

@wcalvert "I am a fan of Nuckoll's plans. My -7 runs a Z-19 with fuses (yes, fuses like you see in your car) with EFII ignition (two batteries, one alt). Can't say if it's the absolute best system ever designed (I have no gripes with it) but it's been running flawlessly for 150 hours and I didn't spend tons of time trying to design my own."

I've learned a lot studying Mr. Nuckoll's diagrams and appreciate his contributions. There's a lot of lessons learned to take advantage of; a main basis for this thread. I've designed a lot of mech and fluid systems in my life. It isn't hard to improve on someone else's work. We all stand on the shoulders of someone others. I will probably be using a fuse block for some of the non=flight critical stuff.

I should have asked different questions to start this thread. For those not completely bored here; advantages/disadvantages of parallel linear buses (ECU, FP) e.g. the Nuckolls Z-28 drawing approach (attach 1) versus the combined single diode isolated bus (attach 5); basically Carl’s response. My apologies for not getting it. My head is swimming.

To the EFII guys. No comments on the combined ECU power and test switch functionalities? I though this had promise.

I'll have to reread this flow of consciousness later to see where it doesn't make sense. Been a long day of fighting engine baffles. I'm whipped.
 
Last edited:
Back
Top