What's new
Van's Air Force

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

Power system architecture for EFI

...So how did you wire yours?

Not counting the nonessential (LOP, ECU pri/sec, etc), I have a single switch that controls the engine:

ENGINE ON/OFF

This switch lights up the ECU package which includes the ECU and controller, main fuel pump, and both coils. As ALL of these components need to be operating to make the engine run, I see no benefit (and many downsides) to independent switching from a PVI perspective.

I've read this and similar threads with great interest. I have developed the philosophy that favors simplicity and reliability over an absolute fail safe in all possible scenarios. I understand that even the most robust, high quality circuit CAN fail, but I believe that I only have to push that probability down to meet or exceed the many single point failure opportunities in the engine itself. Thats not a particularily high standard.
 
Slight deviation from the topic, but I see several references made throught this thread to separate switches for the various EFI components (pumps, coils, brain, etc). I dont see the logic in this, wondering if I'm missing something. In my mind, the engine can't run without the pump, brain and coils lit up, so why more than one switch for the engine control function?

Switched coils for the same reason as switched magnetos--you can check and verify each one is working. However, that's for a four-cylinder; I don't know how the six-cylinder engines are set up.

Switched pumps (at least one needs a switch regardless) so you can alternate use of them, verify both are working, and turn them off if you want to for maintenance or programming on the ground (you could also pull a breaker or fuse for that if desired, I guess). They draw a lot of current so if not needed it would be nice to shut them down.

Other than that, assuming you use a power bus that is completely turned off when you shut down the airplane, I'd imagine you could leave the ECU(s) and injectors "hard wired" on. You'll still have the switch for the ECU select (well, injector control source select) on a dual-ECU system, and the "lean switch" if so desired.
 
If that one switch fails, the engine quits. On the other hand, if there are two
electrical supply paths and one fails, there is a back up path.
There are two fuel pumps. If one fails, there is a backup. Same for ignition
and etc. The objective is to keep the engine running no matter what single item fails.

I understand and appreciate that line of thinking, but there has to be a line drawn somewhere. Yes, that single switch CAN fail in flight, but what is the probability? Is it higher or lower than the probability of the crank gear bolt coming off or throwing a rod or swallowing a valve? Further, does the very presence of a second switch increase or decrease your chances of a favorable outcome from a PVI perspective? After all, there are MANY accounts of aircraft crashing due to "fuel starvation" when there was in fact plenty of fuel on board - but the pilot failed to select the correct tank.

PVI, redundancy and system reliability are all in play with a complex system design. You can't focus on one element in the hopes that it will overcome deficiencies in the others. Its a balance, and not an easy one.
 
Switched coils for the same reason as switched magnetos--you can check and verify each one is working. However, that's for a four-cylinder; I don't know how the six-cylinder engines are set up.

This is where I think the paradigm shift happens with EFI/EI... Aside from verifying function, there is little value in the traditional "mag check" at the EOR. All you really need to do is verify all plugs are firing and thats pretty easy to do in the chocks by pulling a single circuit breaker or just being really in tune with your EMS display. A savvy pilot can see a failed plug or coil pack just by observing the display and picking up on a deviation from normal. We cling to many aviation traditions handed down from the days of the radial engine... Maybe we need to revisit some of them.

Switched pumps (at least one needs a switch regardless) so you can alternate use of them, verify both are working, and turn them off if you want to for maintenance or programming on the ground (you could also pull a breaker or fuse for that if desired, I guess). They draw a lot of current so if not needed it would be nice to shut them down.

Thats easy. If you turn the ENGINE switch to "ON" and hear the pump and see pressure, OPS check good. Otherwise the engine doesnt start and there is no flying. 100% safe. In my case the primary pump requires zero pilot action (just like cars and legacy airplanes). The secondary pump OTOH is activated with the same switch that my legacy boost pump used. Decades of training and muscle memory in this particular ship will hopefully guide me to the correct automatic response to a primary pump failure: BOOST PUMP ON.

As for maintenance, that's not a primary consideration. I'm willing to forego some convenience in exchange for system reliability. And to accomodate alternate use of the pumps - every CI I'll switch the wires in the pump module. Every year the pri pump reverts to boost duty, and vice versa.

Other than that, assuming you use a power bus that is completely turned off when you shut down the airplane, I'd imagine you could leave the ECU(s) and injectors "hard wired" on. You'll still have the switch for the ECU select (well, injector control source select) on a dual-ECU system, and the "lean switch" if so desired.

In my case, the ECU bus is switched to the ECU battery with the ENGINE switch. This system is completely independant of the ships power system. I can shut the ships power completely off and (assuming a healthy ECU battery) the engine will run, just as a legacy piston single. The only tie between the ship and ECU is a crosstie relay that feeds the alternator to the ECU bus. Went back and forth between a diode or relay for this and went with the manual control.
 
Last edited:
Not counting the nonessential (LOP, ECU pri/sec, etc), I have a single switch that controls the engine:ENGINE ON/OFF
This switch lights up the ECU package which includes the ECU and controller, main fuel pump, and both coils.

Better show us a wiring diagram.

Yes, that single switch CAN fail in flight, but what is the probability? Is it higher or lower than the probability of the crank gear bolt coming off or throwing a rod or swallowing a valve?

Well, that firmly places you in the "Does not believe in fault tolerance" column.
 
Better show us a wiring diagram.

With all due respect, I'd rather shove bamboo shoots under my fingernails.

...but the concept is simple: if the ECU is "ON", so are the pump, coils and injectors. Since all of these components are required to make the engine run, there is no possible circumstance that I can think of that requires a separate, pilot activated toggle for each.

...Well, that firmly places you in the "Does not believe in fault tolerance" column.

As operators of airplanes built in a home shop, powered by a single piston engine, we are ALL in this camp. Its only a matter of degree. Remember that most of the pilot population thinks that we are absolutely out of our minds, insane risk takers. And they have some pretty good statistics to prove it.
 
Last edited:
Michael's ideas have some merit. If you have some breakers you can pull to disable for maintenance activities, you can eliminate some switches and also some reaction time in an emergency.

Many people seem to think something WILL happen in their plane. The stats show it's highly unlikely unless you have one of those alternators with a high failure rate. The backup systems become something that most folks will never use.
 
My auto background is longer than I want to admit. I have managed a shop for 30 years and have owned my shop for 20 (I?m 53). Management is wearing all the hats, including working on the floor doing the heavy troubleshooting and problem solving. Between the two shops I have run through more than 50,000 invoices. We work on any car my customers own, all makes and models, 2017 back to the 50?s, more than a few with 300,000+ miles. Approximately 40% american, 40% asian, 20% european. We stay very busy, only accept referals for new customers, and am hidden with no signs.

1.GM: fuel pump relays on occasion, melted fuel pump relays on occasion, blower relays often (ongoing since the 70?s), melted blower motor circuit connections often (ongoing since forever)
2. Ford: EFI and fuel pump relays common in the past and still some of both, melted fuel pump fuses common lately, blower relays on occasion
3. Chrysler: EFI and fuel pump (power shutdown) relays on occasion
4. Toyota: EFI and fuel pump relays on occasion, many ac compressor clutch relays
5. Nissan: ECM relays on occasion, ac compressor clutch relays on occasion
6. Honda: ECM relays on occasion, ac compressor relays on occasion
7. Others: more of the same

When troubleshooting a no start or an intermittant problem my guys are instructed to NEVER touch the relay until being sure what is going on. Many a relay has been touched and now things work, just a little tap with a screwdriver or wiggle to remove and again, things now work. Now the **** thing will not mess up and there is no way to prove what happened. I do not need to bitch at them, they know what they did wrong or had no choice but to finally touch the relay. We check the complete circuit, replace the relay, it never messes up again. The perfect crime ?DanH?.

High current heat type connection issues are mainly on head lamps, cooling fans, blower motors and fuel pumps, in that order. Lots of headlight connections in the last few years on all brands. High flow cooling fans have a lot of connection issues on all brands. I predict we will see some fuel pump connection issues on EFI systems. May take some time.

Not many cars last for 300,000+ miles. Chevy trucks and cars, Ford trucks, Honda cars, Toyota cars and trucks, Mercedes cars, few others. There are exceptions, but these are the standouts. Problems not seen on other cars can be found here. On occasion we get an older car with weird problems. Have had intermittant shorted injectors shutting down the injector or computer and not blowing the fuse, unprovable computer problems solved by installing used replacemant computers, crank signal wiring harnesses not providing the proper signal when oil soaked. I have seen strange things, sometimes not totally explainable.

In the future, when these EFI planes have multiple years and hours on them, we will go through the same things. There will be unexplained intermittant problems that will be difficult to troubleshoot. Computers and sensors will become obsolete and unobtainable new. Something else to think about.

There will be more than one wire design that works, different designs for different missions, different designs for different risk tolerance, different designs for different aircraft. But, there will be some standards that apply to all. Hashing it out here will help.

This has has been an enjoyable thread to follow. I hope to help and contribute.

George Meketa
RV8, V-tail Bonanza, Cessna 180, PA12 Super Cruiser. All at the same time

ps. Sorry for any typos, using an ipad remotely.
 
Interesting. Do you see a lot of these issues on cars less than 5 years old?

300,000 miles equates to something like 7000-8000 hours for most drivers.
 
With all due respect, I'd rather shove bamboo shoots under my fingernails.

Chicken ;)

Actually Mike hasn't proposed anything very radical between the EFI bus and the EFI components. As compared to the standard SDS diagram, he has eliminated only three switches...two coil, one pump. Here's the reworked version:

20pq4o7.jpg


Personally I would not eliminate those three switches, as they add only benign failure modes, not critical failure modes, the point of analysis before philosophy.

In return, coil switches allow clear runup pad checks of coil and ECU function (in dual ECU setups, each coil is driven by a different ECU).

An unswitched primary pump is workable. However, IF it is desirable to alternate pump usage so they build operating hours equally, it's a whole lot easier to do it with switches. Again, the additional switch, if failed, is a benign result. The required pilot recovery (close a pump switch) is the same.

Power supply TO the bus? Yeah, single wire/single switch architecture is, ummm, radical. Mike, you're an engineer. Ditch the crank gear philosophy. With respect, it's trash talk. Is there an accepted industry method, numerical or logical, which would demonstrate that the single wire/switch power supply is somehow superior?

My opinion? It fails the most basic analysis. Either an open or a short in that segment results in a critical failure....the big fan stops, and there is no recovery.
 
The idea that additional switches for required (redundant) components adding a benign failure mode works on paper, but ignores the much more difficult to address PVI perspective. Knowing how pilots react in an emergency (personal experience here), removing "nice to have for maintenance troubleshooting" choices from the normal cockpit flow increases the chances of a successful outcome. That said, my coils are controlled by resettable CB's, labeled and easily accessable. I can pull them if I need to test something, but they are NOT part of the normal engine operation sequence. Your comment about balancing the fuel pump hours has merit, but again, I'm moving that out of the pilot workflow and transferring that responsibility to the maintainer. One less thing for the stressed out pilot to get wrong. And despite the fact that a single switch failure (short) can take down a traditional magneto equipped aircraft, the very remote possibility of my ENGINE control switch failing in flight can be overcome with the addition of a separate circuit and a "Hail Mary" switch pretty easily. This will be a guarded, but easily accessed switch that is still outside my normal PVI flow. My EP for engine out goes from one step (BOOST PUMP ON) to two (EMER POWER ON). So thanks for the reality check on that.
 
Last edited:
The idea that additional switches for required (redundant) components adding a benign failure mode works on paper, but ignores the much more difficult to address PVI perspective.

Uhh, what's a PVI?

....my coils are controlled by resettable CB's, labeled and easily accessable. I can pull them if I need to test something, but they are NOT part of the normal engine operation sequence.

Yeah, that was the point. Other than studying EGTs or pulling breakers, there is no available preflight check of dual ignition function...a normal (and highly accepted) operation.

And despite the fact that a single switch failure (short) can take down a traditional magneto equipped aircraft....

With separate toggles, a switch failure takes down one magneto, a benign failure. The fan keeps turning.

....the very remote possibility of my ENGINE control switch failing in flight can be overcome with the addition of a separate circuit and a "Hail Mary" switch pretty easily. This will be a guarded, but easily accessed switch that is still outside my normal PVI flow. My EP for engine out goes from one step (BOOST PUMP ON) to two (EMER POWER ON). So thanks for the reality check on that.

Ok, so now you have dual power supply to the EFI bus. Assuming you supply your EMER PWR switch from different battery, the architecture is straight from the standard SDS diagram.
 
Uhh, what's a PVI?.

Pilot Vehicle Interface... A human factors issue that has become its own engineering sub specialty in recent years.


...With separate toggles, a switch failure takes down one magneto, a benign failure. The fan keeps turning..

But there are more than a few certified airplanes that funnel the two P leads into a single Bendix switch... In fact, its the industry norm. Sure, its a rare ocurrence in practice, but still will not survive your critical deconstruction on a schematic. An internal short of that single switch is critical and unrecoverable. Yes, an easy "fix" is to simply use two independant switches, but its funny how the industry adopted a single point failure as the norm... Is it because of PVI concerns?


...Ok, so now you have dual power supply to the EFI bus. Assuming you supply your EMER PWR switch from different battery, the architecture is straight from the standard SDS diagram...

Yes, addition of a second switch and wire is easy. But so is the addition of a short length of fuel tube to bypass a clogged fuel filter or a broken fuel fitting...

See where this can go if you're not careful?
 
Pilot Vehicle Interface... A human factors issue that has become its own engineering sub specialty in recent years.

Ahhh. Good subject.

Here we have two diagrams. On the left (best I can tell from the description, circa post 212) is Mike's EFI bus supply architecture. On the right we have the same system, with the addition of two diodes located (specifically) at the EFI bus connections.

erlvms.jpg


First, a partial wire-by-wire fault check.

Left diagram.

Conditions, leftmost switch closed, in flight:

1 open: no power to EFI bus. PVI: Close right switch

1 short: left fuse pops, no power to EFI bus. PVI: in order, open left switch, close right switch. (closing right switch first causes unrecoverable failure, CRITICAL)

2 open: no power to EFI bus. PVI: close right switch

2 short: left fuse pops, no power to EFI bus. PVI: UNRECOVERABLE closing right switch pops right fuse. CRITICAL

The above notes are the minimum PVI's specific to the circuit segment condition. Since the segment conditions are unknown to the pilot, the actual PVI for all failures would be (left switch off->right switch on->wait for a response->if no response, pump switch->ECU select switch)

Now the right diagram.

Conditions: BOTH power supply switches closed, in flight:

5 open: 7-8 supplies power. PVI: None

5 short: left fuse pops, 7-8 supplies power. PVI: None

6 open: 7-8 supplies power. PVI: None

6 short: left fuse pops, 7-8 supplies power. PVI: None

No power supply circuit problem results in power interruption; all failures are benign. The PVI following engine power failure would be (pump switch->ECU select switch). If the physical fuel system will allow both pumps to operate at the same time (aux pump ON for takeoff, the conventional PVI), the entire power interruption PVI becomes (flip ECU select switch).

When an additional component is necessary to design for fault tolerance, it is simply treated as a part of the circuit segment. In this example, an open diode is simply an open in segment 6 or 8. An internally shorted diode may be treated as a wire, which would negate its isolation benifit. In reality, if that isolation became necessary the diode will act like a fuse. For example, assume diode 6 will pass electrons both ways, and wire segment 6 shorts to ground. The diode will at as a fusible link, as shorted amps in either direction would exceed its current capacity. No change to the wire-by-wire check.

Yes, addition of a second switch and wire is easy. But so is the addition of a short length of fuel tube to bypass a clogged fuel filter or a broken fuel fitting...

See where this can go if you're not careful?

Into a philosophy argument? I'd rather teach 'em how to fish.
 
Last edited:
Rusty Crawford Photos



Here's a couple of photos of Rusty's ECU, relay and MAP sensor layout for those who may be interested.

 
Ross and group

I work on cars from near new to very old age, just out of warranty to worn out. We have replaced relays with as little as 40,000 miles and I personally have cars with 400,000 miles that have everything original except the water pump, brakes, and maintenance. What happens while under manufacturers warranty is not usually experienced by me, but plenty does goes wrong. I hear the stories. Mileage and age are not the only factors.

1. We have replaced more than a few Bosch type alarm relays where the normally closed contacts are bad and the starter will not engage. As stated on a previous post, we have replaced hundreds of relays with problems with the normally open section. A relay is just another switch. Again, I feel they are becoming over used in circuit design.

2. Has anyone here ever had a Honeywell TL switch failure? I have seen several Carlin switches fail at the rivets (Can send pictures). Only Honeywell with screw terminals on my planes. Same for B&C alternators and starters.

3. I have seen strobe and fuel pump switch fast-on connections with heat damage, along with Bus brand fuse boxes with heat damaged fuses and melted plastic (can send pictures). All crimps and terminals appeared to be OK. Fast-On's are OK in general, but I feel EFI requires higher quality where possible. Redundancy can make up for the risks. Two EFI bus power sources, two independent fuel pump circuits, one independent backup ignition source, backup fuel system (see: Dave Anders install information, Cessna 185 missionary backup fuel system).

4. We have been running fuel injection on mechanical fuel pumps, with the electric pump only used for starting and TO/Landing. Now they will be necessary, not a backup. I have concerns and would want two independent pump circuits, not on one switch or relay, and carefully check electrical connections at every annual. Time will tell.

5. I see strange electrical failures with low mileage older aged vehicles. My RV8 is 16 years old. What will we do with old EFI computers, injectors, sensors. I have witnessed EI pickups, LS ignition modules, ignition coils fail and have witnessed intermittent problems. These planes will not be new forever. What are the plans here? Run until failure?

6. Terminal quality and crimps are important. A Daniels sub-D, environmental splice, Thomas Betts Sta-Kon WT145A style crimpers along with tefzel strippers are a good start, can be purchased on E-bay, and will provide high quality connections. Here is not the place to be cheap, there is a difference. Larger cable ends can be done with an inexpensive hydraulic crimper purchased on E-bay. They do a great job and are easy to use. An RV10 crashed with fatalities because of an improper crimp and Van's has a Service Letter for improper cable crimps.

7. Quality relays only. The ones on post #101 spec well and one can double up the contacts for load sharing. Same goes for switches.

8. Some of these diagrams have no redundancy and lack pilot control. I would have two battery power sources that could be used simultaneously or isolated, along with EFIS monitor/aural warning of both sources. I would want independent control of the fuel pumps allowing both for takeoff and landing. Both on the Battery EFI Bus, maybe one on a breaker or relay controlled by the ECU and the other on a stick grip controlled relay or panel switch by the throttle for takeoff, landing, emergency. I would run the coils and ECU from a dedicated EFI breaker panel, fewer switches and still maintain control.

I woke up this morning thinking I could find no good reason to install an EFI system on a Lycoming equipped airplane. Any benefits would be minimal at best and the complexity and chance of a dangerous fuel system failure is far greater. I like modern technology, but not for the sake of technology. Some will make the decision and install EFI, some will pursue (like alternate engines) and go back to the old tried and true Bendix style injection. Just go in, realizing the risk and responsibility, with eyes wide open. Sorry, but I will still continue to try and help to the best of my ability.

George Meketa
ASE master tech, A&P mechanic
 
7. Quality relays only. The ones on post #101 spec well and one can double up the contacts for load sharing. Same goes for switches.

Carl's original link:

https://www.alliedelec.com/schneider-electric-magnecraft-92s7d22d-12/70185034/

I pulled up the data sheet, and there is a spec detail I'm curious about. Take a look at the NC rating:

https://www.alliedelec.com/m/d/fdee0862eba999cc55833ca90d4f03ff.pdf

Are they saying the NC contacts are only good for 3 amps at 28V?
 
George wrote,
"Bus brand fuse boxes with heat damaged fuses and melted plastic"
I took a Busman fuse block apart with the intent of
modifying it so that it could be fed power at both ends for redundancy. While it
was apart, I noticed that the fuse clips do not have much gripping strength. I
decided not to use the fuse box in my plane because it was a poor design; weak
fuse clips, cheap plastic could melt or burn, wires exited to the sides, not being
compact takes up too much panel space. So I made my own fuse panel using
THESE fuse clips. They are board mounted fast-on connectors.
 
Are they saying the NC contacts are only good for 3 amps at 28V?
Yes. Normally open contacts are held tightly together by a strong
electromagnet. Normally closed contacts depend on a spring to hold them
together. The spring is not very strong compared to an electromagnet.
N.O. contacts might bounce less when closing.
Some automotive relays are rated 40/30 amps. 40 amps for N.O. & 30 for N.C.
 
Yes. Normally open contacts are held tightly together by a strong
electromagnet. Normally closed contacts depend on a spring to hold them
together. The spring is not very strong compared to an electromagnet.
N.O. contacts might bounce less when closing.
Some automotive relays are rated 40/30 amps. 40 amps for N.O. & 30 for N.C.

Thanks Joe. I was aware of NC de-rating (like 30/50 and 15/20)....

https://www.delcity.net/images/linedrawings/73980.pdf?v=20170424182940

....but 3 amps NC with a 30 amp NO rating? I had no idea the NC rating could be that small a percentage. My "something-learned-today". The linked order page didn't offer a clue:

https://www.alliedelec.com/schneider-electric-magnecraft-92s7d22d-12/70185034/
 
Ross and group

I work on cars from near new to very old age, just out of warranty to worn out. We have replaced relays with as little as 40,000 miles and I personally have cars with 400,000 miles that have everything original except the water pump, brakes, and maintenance. What happens while under manufacturers warranty is not usually experienced by me, but plenty does goes wrong. I hear the stories. Mileage and age are not the only factors.

1. We have replaced more than a few Bosch type alarm relays where the normally closed contacts are bad and the starter will not engage. As stated on a previous post, we have replaced hundreds of relays with problems with the normally open section. A relay is just another switch. Again, I feel they are becoming over used in circuit design.

2. Has anyone here ever had a Honeywell TL switch failure? I have seen several Carlin switches fail at the rivets (Can send pictures). Only Honeywell with screw terminals on my planes. Same for B&C alternators and starters.

3. I have seen strobe and fuel pump switch fast-on connections with heat damage, along with Bus brand fuse boxes with heat damaged fuses and melted plastic (can send pictures). All crimps and terminals appeared to be OK. Fast-On's are OK in general, but I feel EFI requires higher quality where possible. Redundancy can make up for the risks. Two EFI bus power sources, two independent fuel pump circuits, one independent backup ignition source, backup fuel system (see: Dave Anders install information, Cessna 185 missionary backup fuel system).

4. We have been running fuel injection on mechanical fuel pumps, with the electric pump only used for starting and TO/Landing. Now they will be necessary, not a backup. I have concerns and would want two independent pump circuits, not on one switch or relay, and carefully check electrical connections at every annual. Time will tell.

5. I see strange electrical failures with low mileage older aged vehicles. My RV8 is 16 years old. What will we do with old EFI computers, injectors, sensors. I have witnessed EI pickups, LS ignition modules, ignition coils fail and have witnessed intermittent problems. These planes will not be new forever. What are the plans here? Run until failure?

6. Terminal quality and crimps are important. A Daniels sub-D, environmental splice, Thomas Betts Sta-Kon WT145A style crimpers along with tefzel strippers are a good start, can be purchased on E-bay, and will provide high quality connections. Here is not the place to be cheap, there is a difference. Larger cable ends can be done with an inexpensive hydraulic crimper purchased on E-bay. They do a great job and are easy to use. An RV10 crashed with fatalities because of an improper crimp and Van's has a Service Letter for improper cable crimps.

7. Quality relays only. The ones on post #101 spec well and one can double up the contacts for load sharing. Same goes for switches.

8. Some of these diagrams have no redundancy and lack pilot control. I would have two battery power sources that could be used simultaneously or isolated, along with EFIS monitor/aural warning of both sources. I would want independent control of the fuel pumps allowing both for takeoff and landing. Both on the Battery EFI Bus, maybe one on a breaker or relay controlled by the ECU and the other on a stick grip controlled relay or panel switch by the throttle for takeoff, landing, emergency. I would run the coils and ECU from a dedicated EFI breaker panel, fewer switches and still maintain control.

I woke up this morning thinking I could find no good reason to install an EFI system on a Lycoming equipped airplane. Any benefits would be minimal at best and the complexity and chance of a dangerous fuel system failure is far greater. I like modern technology, but not for the sake of technology. Some will make the decision and install EFI, some will pursue (like alternate engines) and go back to the old tried and true Bendix style injection. Just go in, realizing the risk and responsibility, with eyes wide open. Sorry, but I will still continue to try and help to the best of my ability.

George Meketa
ASE master tech, A&P mechanic

I had Carling switch failures in my RV-9A leading to burnt wiring and terminals in 2006. I published a forensic analysis on VAF and Aeroelectric but it is now obscured in the depths of the interweb. Conclusion: Never use Carling switches. The Honeywell TL switches are better. In my Rocket, I eliminated all load carrying switches and went to an array of relays controlled by low current switches.

As George mentioned, the rivets on the Carlings allow vibration, stress and current to degrade the connection. This can lead to thermal runaway with certain loads and smoke in the cockpit.

FMI http://www.aeroelectric.com/articles/Anatomy_of_a_Switch_Failure/Anatomy_of_a_Switch_Failure.html
 
Last edited:
In the past cube type relays often had NC de-ratings printed on the body. Not much anymore, but I do have a Potter Brumfield (T9AP5D52-12) panel mounted relay in front of me with a 10 amp NC / 20 amp NO spec printed on it.

The linked relay has no NC section? That is one reason, along with the silver contacts, that I like it. Specs are good, but quality materials and construction go a long way in the decision process. Some of the diagrams and posts here make much use of the NC section. Bad idea. Never for something like a fuel pump.

https://www.alliedelec.com/schneider-electric-magnecraft-92s7d22d-12/70185034/

Like I have been saying, relays are not the miracle devise many here believe them to be. Cut one open and look at the small contact areas. Use them sparingly and for function. Stick grips, dynamic braking / ground function for flap motors, sensor / warning lights, remote contactor functions, avionics annunciation, etc. I recommend a switch for important items: ECU power, at least one backup Bus power, at least one fuel pump on an EFI system, etc.

The electrical failure examples given here are clues to where we will see problems. I often tell my guys when they encounter some small subtle clue that "the car is talking to you", "are you listening?".

George Meketa
 
Buss 15600-08-10 ATC Fuseholder internals

George wrote, I took a Busman fuse block apart with the intent of modifying it so that it could be fed power at both ends for redundancy. While it was apart, I noticed that the fuse clips do not have much gripping strength. I decided not to use the fuse box in my plane because it was a poor design; weak fuse clips, cheap plastic could melt or burn, wires exited to the sides, not being compact takes up too much panel space. So I made my own fuse panel using
THESE fuse clips. They are board mounted fast-on connectors.

This is what a Buss 15600-08-10 ATC Fuseholder looks like inside https://photos.app.goo.gl/ZKjbqnU3DfkUKNex2

I'm not making an argument to use it or not; just showing the insides.

It's easy to add a 2nd power feed stud as shown in the photos.
 
Last edited:
Two potential issues occur to me about using a fuse position as power input to a bus. One is a caution that a single blade position will have a current limit lower than the stud input location. The other would be the care needed in clearly labeling the unconventional input. Maintenance at a later date, particularly by a subsequent owner, could be somewhat hazardous if they weren't aware of the 2nd power input.
 
ATO/ATC Fuse Holders and Bare Terminals



I've used these Cole Hersee ATO/ ATC fuse blocks for many years with no issues in race cars and my aircraft. With the proper terminals used (below), you pretty much need pliers to pull the terminals off the blades.



These are the terminals I use for wiring. After crimping and pull test, I use double wall, glue infused, shrink tubing over the barrel and at least 3/4 inches back over the wire for strain relief and additional retention properties.
 
Two potential issues occur to me about using a fuse position as power input to a bus. One is a caution that a single blade position will have a current limit lower than the stud input location. The other would be the care needed in clearly labeling the unconventional input. Maintenance at a later date, particularly by a subsequent owner, could be somewhat hazardous if they weren't aware of the 2nd power input.

Thanks Charlie... your comments inspired me to add a second 10-32 stud.

More photos at https://photos.app.goo.gl/ZKjbqnU3DfkUKNex2

i9gadzsYzffr6zG0q8QuE91bgGuxlQfMFvaSxiJlhyQdq0KoygL8L9Gs0p3kJE206DAesJWadDpYUiP2r7bBR2Y92lfsWCkq-dagXEqrd3jHuJFilQ9bRV1riYsZjVGlcP1L-ZF0zrGKr5FXHQElfzRbTEbBgJvR5mJlIP41uWHPpDPdb25MzedfsqX6IvL3Uv33FFtVUFD3E792XQ5s3NFztI0VrajUMPlwOdz3Mnpfge6N65WQtkBsTCa6yHWl_QAkFazknsMU01-rak-h5EWgBr6GXcL6dcMdNKkRPWTIfdIcw7NSnSt6_7g0hDbC0cCRXPuh0N0XMCh1F1nCCDBNWfC3-EyOff3lSaq6_ikSvvHX8lLiiAVp3K7W0lHW7CJFPCeezZS1-a8DWW_eyorA2A4wysGoeNLoTMSUDJqOnR1Pa_CZJKKHTEzWSPVjuxm1mdNwJU6BKr7xq-d1ymERfOIZuqD_1hH5TlWEzO2-wKYoTtVUs5jBIJKjYUqcd8NgR7UT6z_giR9OzCvQWAG9YolmhHxPIMctpKTp5RyPQxgBeUThT17gBkCSoeG0X1ehsc6bkXw8I3R9V-tiBPfldRCguRnUq25ch-OiRKRqnm4ylEYgb1GlvbL8j0LEB7y7LNoyS1mvxzrEjIZaXxjMm8Jb0fCa=s1917-no


kH-k96qzblrEDDBl8tVC60k9GpInhtRMXzsreynZ9ktvwcK17e8xN7g3JdAVNz6hNJMbb9F_fvFLHPRuPNDFzfGb_olKlXvbfUUqL--HMjzEzuUAf6BJQyGeg0RPCr7j6Rr70E8W0UB8LH8rTyCUQoBo3jzIaptubp6OAEOf1fLjGce0UD2pVLaE2vOvAn6EQrQfRETYEMXpoEOQSFO7T1oEgKk6hK9sEVVx862BCYHADXDnOplGzktxE8JVh5p54TmRn7jLB9U1qV8Cnt-EuEMPgkRq6pdieYyYYQCrGYDJU6DaMFl4Af7bVvxRHcATnK7QQjxssM0bEX-8EzFJ19Xr4IiBF1nqgEV6APE0hsjrzn04rEsMdv-evEX05_5gEi9ngYPeAic2gGKAnOzsgahs7kdfNw_wtI7wZkwp88Si-fNJFvXxTpSMe4DntufAqVtOqaN8AbMOqVCdXUFwLov37-blq6Qv9O5iBNvK4GMDiq4h3FMPkVzrpZ5Z1SmeuMliASGvAYi6VUX0mR0Ql-x9hEFpg-MVoMkYS7Lfme9AAx8wjrb7qNgesuQEN1ch4HPpy8zGdCEeygBfKOBooWwPlXTHorj3K2_kFE1TvsLQ7KTBtOTA2uAzoscnbo_LqAN0sf4hE_6Wn9l6r5RrkoNlROdMfZyQ=s1917-no


2490nR7NGjNVfq0pXKfHGUxghQJyiTZCMvhnX78weQhFH0NP2YCFvmohBVxNWfZoG3uXYIz_RuoXDp9nFOk-NUbWzYqhwHmf6SBj5cD2sHqOCa3ogNCZLoFu_npPY_G1tskt5uw1yeyH6QydN1TC0YoZfVDF75e5WGHNChHhfr1luPIwYJC2A1sEFgtqB8pKoxS-2QxfAhmUHQzIH8PWYXqAWClasDkY_Yify8pYLwTEJEQXo3nbR9OMOUai779DiazRhxNCSs2dUCxy7qxC10fTzx_ICfq7LjL3bfEVDV_IiD5XmZlpacyi-OTyUR1rJ9qEToP_shvqiR-Nz70fU_nQORYgLLuvznmdAKbsCJtjgv1Aaa9hn8-5HG2Qw7DZcWadbAqHQ4b_e4hFPVJsnyzNCN5sfOw9CQs8UAKauGYVt9dMYT2BEipuenPN39RVuzB6J_KMHS96ITJdqWVOPYfUf_tayA9Ziqv6P_8y_CK9vfNYTifgrffZxbl7G6msymIe63MohlGntFmeBFt-HvMvMvQlyD8Fb14h6iSihI8ZC92B93aC_Rqnl5Q6B4VShUCB8FT1hEmTNZosguZPKhVFT5gszv6PndVbik2gsGvojSwAY0BMBK_nBT7wBppiIWLpHvgMJe7BLvEBf8HEROeRIVOeQVUZ=s1917-no
 
Forgive my ignorance but why wouldn't you just stack the terminals for the primary and secondary power inputs on the stud provided? Seems like there is plenty of room.

The concept is that if there is somehow a failure at one stud there will be a backup stud. There will be two independent power sources into the bus and it gives greater reliability to have separate studs also.
 
Last edited:
I can follow his logic. If you feel you need redundant power feeds to a bus, the single stud would be a single point of failure for both feeds if it came loose.
 
As I posted in the CPI thread, if you're concerned about a single stud, just split your power supply setups with 2 smaller fuse blocks, say a couple six slot ones instead of a single 12 slot.

I'm not sure how the stud would fail on the Cole Hersee one I showed, since it's actually a bolt penetrating the power strap, unless you way over torque it. Are you folks worrying about all the single points of failure inside that single engine up front too?

Maybe we should be looking at fitting a couple of these to drop out of the belly if the electrical system or powerplant takes a dump?

 
Last edited:
What the failure mode on that system?:rolleyes:

Yes, I see, we'd need independent ignition and fuel systems for each turbine plus a separate door and actuator system to drop them into the airstream independently. Plus a fire warning system, fire extinguishing systems and N1 indicators...:rolleyes:
 
Last edited:
and...

Then we would need to look at the switches. contactors, and relays, too.

Also would need to account for insulating the cavity that contains the turbine for heat, sound, and the potential for an unconfined turbine failure...not to mention the the separate fuel system for the turbine and preventative measures to ensure that no kerosene gets into the main tanks or gasoline into the turbine...

That's just a short list but it's enough to get started!:D
 
Well, duhhh.

It would be amusing to think about all those guys who poo poo alternative engines, and then throw something like that, with such obviously un-thought-out design work, on a 'traditional' engine and call it good. But for how scary it is...

Charlie
 
Let's treat the information in the spirit of aviation safety culture...lessons to be learned, "There but for the grace of God", etc.

Too often, "duh" is only obvious in retrospect.

The docket should be interesting, when available.
 
Last edited:
Dan,

From the installation manual, here are all the instructions I could find for power to the system:
11. Ignition power (these wires stay bundled and fed from a separate IGNITION breaker).
12. ECU power (black to ground bus, red to 5A ECU PWR breaker).


The closest thing I could find to a wiring diagram is a photo of the harness.
 
more...

If you are speaking about the EFII system, in the ignition install manual, right below the photo of the harness is this quote:

"...
The Ignition Power wires provide +12V power to the ignition coils. Provide +12V power to these wires through a single 10A circuit breaker.
The red ECU power wire needs to be connected to a source or +12V through a 5A circuit breaker. The black ECU power wire should be connected to your ground bus using a reliable crimped or soldered connection method..."

In the EFII install manual, this quote also appears:

"...The single ECU EFII systems require three +12V circuits and associated circuit breakers. The dual ECU systems require an extra breaker for the second ECU. If you plan on using an EFII Bus Manager, you may also need panel space for a Fuel Pump Mode Switch, and Start Battery Select
Switch. The circuit breaker requirements for below:

System ECU breaker(s)

EFII-4 5A
EFII-4R 5A (x2)
EFII-6 5A
EFII-6R 5A (x2)

the different versions of the EFII system are listed

Ignition breaker
10A 10A 10A 10A 15A 10A 15A 10A..."

The table format did not paste properly but all the info is there in the install manuals, if you care to check it out...
 
I think you're looking at the *ignition* manual; not the full ignition/injection controller manual (subject of discussion).

I guess I should have included the link.
https://www.flyefii.com/media/EFII_Installation_Manual_rev9-13.pdf

And I don't see a proper wiring diagram, in either manual. I did find the text below, under 'grounding the vehicle systems'. (?)

"The single ECU EFII systems require three +12V circuits and associated circuit breakers. The
dual ECU systems require an extra breaker for the second ECU. If you plan on using an EFII Bus
Manager, you may also need panel space for a Fuel Pump Mode Switch, and Start Battery Select
Switch. The circuit breaker requirements for the different versions of the EFII system are listed
below:
System ECU breaker(s) Ignition breaker Fuel Pump breaker
EFII-4 5A 10A 10A
EFII-4R 5A (x2) 10A 10A
EFII-6 5A 15A 10A
EFII-6R 5A (x2) 15A 10A "


No indication of where the injectors get their power; it mentions breakers for ECU, Ignition, and fuel pump.
 
Docket items:

https://dms.ntsb.gov/public/61000-61499/61262/615998.pdf

https://dms.ntsb.gov/public/61000-61499/61262/615999.pdf

https://dms.ntsb.gov/public/61000-61499/61262/616000.pdf

https://dms.ntsb.gov/public/61000-61499/61262/616002.pdf

Speaking strictly about injector power, there are many lessons.

The investigation tends to focus on an obvious proximate cause, an undersized breaker. In this accident, the instrument panel vendor installed a single 5A breaker for injector power supply based on what may have been an offhand comment from the EFI vendor. That recommendation was accepted or overlooked by the owner/builder, the builder assistance crew, and the engine vendor, who helped with the installation. The lesson? Like everything else, final responsibility rests with the builder....trust but verify. When in doubt, measure the operating current draw, size the supply wire for a reasonable margin, then size the breaker to protect the wire.

Don't be fooled by average current. Depending on the device, there may be a peak current followed by zero current (here, the typical injector), or a short inrush when first flicked on (the typical incandescent light), or a higher than normal current when under a temporary heavy load (like a DC motor). The lesson? Understand the device you're wiring.

A single circuit breaker to drive the coils and the four injectors is an extension of the "single crankshaft nut" philosophy. What it ignores is that if any one of the six component feeds (two coils, four injectors) becomes shorted to ground, all six components go offline. In this case someone wanted to separate the injector power from the coil power. Since that doesn't improve reliability, we can only conclude that again, it was a decision based on hubris rather than analysis. This contrasts with using six individual circuit protection devices, so that if any one is shorted, the other five remain functional. Most RV's will fly just fine on a single ignition, or with an injector loss, i.e. with one dead cylinder. The lesson? Subject every part of the wiring to a basic "short or open" analysis, as discussed elsewhere. Ignore the probability of a short or open, but rather, understand what happens if it does short or open.
 
Last edited:
There was a fatal accident with our EFI installed on an Egg 6 engine a number of years back, which I helped the TSB up here investigate since they were unfamiliar with these engines or EFI. After going over the obvious, I found the cause in about 15 minutes for the power loss. Injector power was mis-wired to the wrong breaker value (adjacent breaker). In this case a 3 amp one.

It did not show up in extensive ground running as the rpms did not get quite as high as they did with the prop unloaded a bit in flight (current draw a bit higher at higher rpms). Engine stopped at about 300 feet after takeoff and a landing was attempted on the crossing runway. Stall/spin and my friend test flying, was killed.

As Dan said, this is vitally important stuff.

For Vertical Power users, please read our warnings about setting breakers values on these devices when using our EFI.
 
Last edited:
Back
Top