Adding redundant fuel and spark to auto/sled/motorcycle conversions.

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
4,740
Location
US
If we are converting an engine that is already equipped with modern electronic ignition (EI) and electronic fuel injection (EFI) systems, what are the options for adding economical, lightweight, reliable redundancy for assuring fuel and spark continue to arrive? I understand and agree that these modern systems are highly reliable. Maybe from an objective standpoint redundancy isn't warranted (after all, there are plenty of other single points of failure on any engine: we don't have a redundant crankshaft, water pump, oil pump, throttle cable, etc). But if we nonetheless want redundancy in fuel and spark, what's the best way?

I'm thinking of the Aeromomentum Suzuki right now: 1 plug per cylinder, it has an aftermarket ECU (Microsquirt) that provides sequential FI (one injector per cylinder) and ignition.

Options I can see:
1) Add another Microsquirt ECU that has a completely redundant set of sensors (crank position, throttle position, temperatures, MAS, etc) and injectors, with a set of pluq wires going to the single set of plugs. Turned off in normal ops (except for a check before takeoff), and toggled on when needed.
PROS: Fully functional backup that results in normal performance
CONS:
- Expensive, complex, and not very light (for all the additional injectors, sensors, wires, etc)
- Electric power will still be required by both systems, so true redundancy requires redundant power

2) As above, but the auxilliary ECU is very minimalist: A single injector in the throttle body, the bare minimum of sensors, set up to function in "limp home" mode.
PROS: Cheaper and lighter than Option 1
CONS:
-- Degraded performance in "limp-home" mode
-- Same as Option 1 regarding electrical power

3) Caveman approach:
-- Fuel induction would be from a simple carb throat/venturi ahead of the throttle body (or, simpler, a Revflow/POSA/AeroConversions style "metered leak" that drips fuel into the intake based on the position of a cone-shaped pin. The pin would move in and out with the throttle all the time, but no fuel would flow because the mixture is turned off mechanically.) With either a venturi carb or "drip," in case of engine failure the main FI is turned off (either with a separate switch of a microswitch that is activated when the mixture knob on the backup system is move from "off).
-- Spark: Magneto, or the fixed-advance magnetron system used on Sonex (and B&S lawnmowers :)).

PROS: -Technologically simple
- Will provide fuel and spark in case of total electrical failure (if gravity feed of carb is available)
CONS:
-- Lots of things to individually engineer and fit into the existing engine package. So, not very simple in practice
-- Can a spark gap be found that will work well with both the "main" EI and a much lower energy magnetron?
-- Engine will not be running efficiently in backup mode. Fuel delivery is bound to be too rich or lean until adjusted, magnetron ignition does not provide spark advance, etc.
-- Probably fairly heavy (carb body, new fuel lines, magneto, etc)

Based on the little I know, I'd lean toward Option 2. I don't know what a minimum set of "limp home" sensors would be, however.

Opinions (and spears) are welcome and expected . . .
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
12,594
Location
Memphis, TN
What is the end game? Save your bacon or redundancy to continue oblivious to the failure? Save bacon, duel ecus, single injectors and plugs. All injectors or plugs do not fail. One maybe on a heathy engine. If you were flying around the world nonstop, double the injectors and plug setups. How long are the flights and how far away from home? Less than than two hours, I would go single everything.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
4,740
Location
US
What is the end game? Save your bacon or redundancy to continue oblivious to the failure? Save bacon, duel ecus, single injectors and plugs. All injectors or plugs do not fail. One maybe on a heathy engine. If you were flying around the world nonstop, double the injectors and plug setups. How long are the flights and how far away from home? Less than than two hours, I would go single everything.
Single everything--including ECU?
My objective is to be able to land safely in a way/place that poses only minimal chance of damage/injury. Say, 30-45 minutes of available flight time with the ability to safely climb at a moderate rate. I don't care if I don't reach my initial destination.

I'm curious why the duration of the typical flight would affect your desire for redundancy? If I'm over solid forest, rocks, water, etc and the engine goes quiet, it seems unimportant whether I typically fly for six hours or two. In deciding whether redundancy is required, the only two factors I considered were:
1) Do I need to arrive at my initially planned destination? (My answer: No)
2) What would be the consequences of a forced landing? This would depend on the terrain overflown, the characteristics of the aircraft (esp stall speed), even the number of people likely to be aboard. IMO, a single-place ultralight operated over a dry lakebed is a different case than a Cozy flown with the family aboard and operated over the Appalachians.

Thanks for the input!
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
4,740
Location
US
dual ignition is not that difficult and adds power in addition to redundancy
I would think drilling for another set of plugs would be hard in a 4-valve-per-cyl head plumbed for water cooling. And maybe, given the modification to the head, more likely to cause problems than to reduce them. Plug failures are pretty rare these days, and the high energy of the modern EI systems provide usable spark even with some pretty ugly deposits.
 
Last edited:

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
3,749
Location
Calgary, Alberta, Canada
Single plug engines are harder to backup ignition on unless you can find some reliable diodes that work at 45K volts to isolate your second set of coils. If you can do that and tee the plug wires together reliably then you can use a second ECU to drive them, ditto with the injectors. Where do you draw the line on redundancy with so many single points of failure already present as you pointed out?

Very hard to place a second set of plugs in most liquid cooled heads. Drill anywhere else and you're likely to hit water as you said.

I'm not sure why many people believe mechanical devices are any more reliable than well designed and proven, modern electronics.

The Microsquirt ECU doesn't have much of a track record in aviation. Is that why you don't trust it?

I wouldn't care how the backup system makes the engine run as long as it gets you back down safely.

Seems like option 2 is best if you want a backup.

We have about 1500 single ECU systems out there and only about 5 of those to my knowledge have some sort of mechanical backup to get fuel and spark to the engine if the ECU takes a dump. These have something like 400,000 flight hours on them to date. I can think of about 4 which shut down (3 to poor wiring connections and one due to a dead alternator and finally dead battery 15 minutes later). No actual ECU failures that I'm aware of.

We have another 200 dual ECU setups out there with around 100,000 flight hours on them. I can recall 2 people telling me they switched over to the backup after the engine started running funny and 1 of these came back smooth on the backup. Again, bad grounds were found to be the cause in these.

Of course, this is one brand of EFI and has no bearing on how some other brand might perform.

I'm not comfortable in my older age in ANY single engined aircraft at night, IFR, over the rocks or big bodies of water no matter what powers it or what provides fuel and spark but I'm fine flying over areas where I can likely carry out a reasonable forced landing with a single ECU. Everybody has a different risk threshold.

The complexity of dual ECU setups makes you think again if it's really a good idea or not. Lots more wires and relays to isolate the failed parts.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
4,740
Location
US
I'm not sure why many people believe mechanical devices are any more reliable than well designed and proven, modern electronics.
I'm not squarely in that camp, it depends on the device and situation. I wouldn't feel comfortable flying over rough terrain with a single magneto, either. OTOH, like most folks, I've flown wit a single carb and never even considered the idea of redundancy for fuel (though I'd want two pumps if pumps are needed).

The Microsquirt ECU doesn't have much of a track record in aviation. Is that why you don't trust it?
Partly. Also, frankly, I don't understand the details of how these things work, and what happens when things fail. Which sensors are critical to flight? What happens when various sensors or injectors fail? How do I troubleshoot that in flight? Has this thing been thoroughly wrung out to assure there are no "blue screen of death" crashes when various data is out of limits, when voltage is squirelly, etc? A carb or magneto is easy to understand, but I have dabbled enough in programming to appreciate the tremendous difficulty of debugging and crashproofing complex code.

I'm also a bit concerned about the hardware: Capacitors do fail (esp in the heat), PC boards and the soldered connections on them can crack due to vibration, etc. Having a second unit (or a completely independent unit, even if in the same box) would help alleviate this concern.

We have about 1500 single ECU systems out there and only about 5 of those to my knowledge have some sort of mechanical backup to get fuel and spark to the engine if the ECU takes a dump. These have something like 400,000 flight hours on them to date. I can think of about 4 which shut down (3 to poor wiring connections and one due to a dead alternator and finally dead battery 15 minutes later). No actual ECU failures that I'm aware of.

We have another 200 dual ECU setups out there with around 100,000 flight hours on them. I can recall 2 people telling me they switched over to the backup after the engine started running funny and 1 of these came back smooth on the backup. Again, bad grounds were found to be the cause in these.
That's interesting and useful, thanks. So, about 90% of your customers are entirely dependent on a single SDS setup to keep their engines running. I'll paw through your web site, I'd like to learn more about how your system handles sensor failures, etc.

The complexity of dual ECU setups makes you think again if it's really a good idea or not. Lots more wires and relays to isolate the failed parts.
Yes, I can appreciate that. And plenty of potential for one of these isolation components to fail and make the two systems be less reliable than a single one.
 

BBerson

Well-Known Member
HBA Supporter
Joined
Dec 16, 2007
Messages
12,986
Location
Port Townsend WA
I never heard of a redundant fuel system.
My V-twin has a separate mag coil on each cylinder. And dual separate carb venturi with shared bowl.
For more cylinders, dual carbs could be fitted to each bank of cylinders so only half of the cylinders would fail.
I would be more concerned about the PSRU, prop, crank, etc.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
4,740
Location
US
Take the thermostat out or drill small holes through it. Oh wait... that was a different thread.
But still useful!

I would be more concerned about the PSRU, prop, crank, etc.
True. But there's only so much I can do about those things. If there's a relatively simple way to avoid a possible big problem with fuel or ignition, then I'd look into that.
Maybe it's something about the >nature< of the ECU and other parts that makes me wary. That little brittle PCB, those tiny solder joints, lots of little electronic parts soldered in, the connectors, all the things in the injectors and each sensor--maybe a hundred mission-critical things, none of them amenable to inspection to prevent a problem. Heck, a timer board on my AC unit failed yesterday, I've temporarily jumpered across it to keep the house cool. I know they can be reliable--never had one fail in any of my cars, and lots of other folks have had similar experience.
 

BBerson

Well-Known Member
HBA Supporter
Joined
Dec 16, 2007
Messages
12,986
Location
Port Townsend WA
I like the simple stuff. My Honda V-twin has a magnet that rotates past a coil and a wire to the spark plug, no impulse, no computer ECU, no sensors. Just simple.
Unfortunately, I don't know if the V-twin has enough power.
A bigger version but just as simple would be nice.
Auto conversions are too complex to interest me.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
4,740
Location
US
My Honda V-twin has a magnet that rotates past a coil and a wire to the spark plug, no impulse, no computer ECU, no sensors. Just simple.
Yep, that's the "magnetron" of which I spoke under "option 3" in the OP. The Sonex "Aerovee" , and Revmaster use them and lots of lawnmowers and industrial engines use them. They are very simple devices, cost about $15 each, and produce a usable spark as long as the flywheel is rotating fast enough. The advance can't be easily adjusted (Sonex is fixed at 28 degrees BTDC), so there's that. For use as a redundant ignition system (backing up a sophisticated EI that would do a lot of things in a more optimum way), the biggest concern would be the spark gap. The magnetrons get the job done, but it's not a very energetic spark, the plug gaps need to be set fairly small. That's fine if they have their own plugs, but if we have just one plug, the gap would be too small for use by the higher energy EI.

I think I remember reading about spark plugs with two independent positive leads that could be gapped independently. That might work for the present challenge, if they really do exist.
 
Last edited:

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
3,749
Location
Calgary, Alberta, Canada
I know nothing about the Microsquirt stuff as far as how they handle sensor failures. On our EFI, we have defaults entered at the high and low ends of the parameters so that the engine continues to run pretty normally at high power if the MAP, air temp or engine temp sensors fail. Only rpm is critical and that's provided by a Hall Effect sensor which I've never seen fail electronically in something like 8000 systems and millions of hours.

With circuit boards, lots depends on the soldering process they go through, pre-heating of the board, shadowing effects of tall components etc. Again I know nothing about how other manufacturers load and solder their boards. We've used the same company for years now to do ours and they've done millions of boards collectively for many clients. The quality is excellent and we've never had a component fall off yet in environments much more severe than aviation (think offroad desert racing for example).

Properly designed electronics are far more reliable than mechanical parts in my experience. I like to cite the case of one of our ECUs used for programmer and MAP sensor testing. It ran 24/7 for nearly 14 years- 145,000 hours. We have several clients with over 2000 flight hours each, no issues and some military UAVs probably with more than that.

People fly with one carb, sometimes one mag, one fuel pump and think nothing of it but they're mortified to trust one ECU. To each their own.

With any backup ECU scheme, be sure to be able to isolate the failed driver or the backup may do you little good.

I'd ask Aero Momentum how many hours they have on the fleet of Micro Squirt controllers and what if anything has failed on them. Try to get some independent feedback as well. They could be very good. Also ask how they cope with sensor failures. This is a very important aspect as you surmise.
 
Last edited:

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
12,594
Location
Memphis, TN
Just like your car, the electronic stuff works past the life of the car. 100,000 miles is about equal to 1500 hrs of an airplane engine if counting rpm. Once past the early mortality rate of the electronics, it will be good a long time. Connections, bad fuel, old fuel, just plain sitting and not being used will be where the problems come from. The problem can be what do you do down the road? The thing has been so reliable but it is time to overhaul, and they don't make your stuff anymore. Having a spare sitting on a shelf would have been the thing to do. If you make it to overhaul, you will want to start with fresh parts. Wires, injectors, sensors, pumps, ECU, Everything just to get rid of the cycles. You do end up starting all over with individual reliability testing but you know the system works.
 

BBerson

Well-Known Member
HBA Supporter
Joined
Dec 16, 2007
Messages
12,986
Location
Port Townsend WA
The battery powered electronic ignition from a R/C gas model could work for independent backup. One to four cylinder systems, I think. Tiny spark plug for a smaller hole in head.
 

gtae07

Well-Known Member
Joined
Dec 13, 2012
Messages
1,938
Location
Savannah, Georgia
We have about 1500 single ECU systems out there and only about 5 of those to my knowledge have some sort of mechanical backup to get fuel and spark to the engine if the ECU takes a dump. These have something like 400,000 flight hours on them to date. I can think of about 4 which shut down (3 to poor wiring connections and one due to a dead alternator and finally dead battery 15 minutes later). No actual ECU failures that I'm aware of.

We have another 200 dual ECU setups out there with around 100,000 flight hours on them. I can recall 2 people telling me they switched over to the backup after the engine started running funny and 1 of these came back smooth on the backup. Again, bad grounds were found to be the cause in these.
I'll admit I've been thinking more about a single-ECU system, maybe with dual power feed (if that's possible).
 

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
3,749
Location
Calgary, Alberta, Canada
I've flown with a single ECU for 15 years now. Not something that I worry about.

Wiring is the #1 fault we see on customer installations. Basic power and ground connections MUST be 100% if you expect to have reliability.

We recommend the pumps be changed out at 2000 hours even though we have experience running them to 5000 hours. I see no need to change the temp and MAP sensors since there are no moving parts in them and they have been extremely reliable in service. I probably would change the TPS at 2000 hours though since it has contacting/ moving parts inside.

Injectors, we never see issues with them, I've seen plenty with 10,000+ hours on them still working fine and some decades old ones from cars with probably over 20,000 hours on them, still flowing and sealing well when tested on the bench. This is with OE brands, not El Cheapo junk which some people peddle.
 

Daleandee

Well-Known Member
Joined
Sep 11, 2015
Messages
992
Location
SC
For use as a redundant ignition system (backing up a sophisticated EI that would do a lot of things in a more optimum way), the biggest concern would be the spark gap. The magnetrons get the job done, but it's not a very energetic spark, the plug gaps need to be set fairly small.
I can't speak to the overall reliability of the SDS systems but Michael Crowder had a Simple Digital System EM-5 electronic fuel injection system installed on his Jabiru 3300 that crashed after an engine failure. From speaking to him and reading the NTSB report the fault was in a poorly routed wire to the crank sensor that burnt and shorted out ... not a fault of the system. Full narrative here:

https://www.ntsb.gov/_layouts/ntsb.aviation/brief2.aspx?ev_id=20130909X40737&ntsbno=ERA13LA405&akey=1

My installation is a bit of a mixed bag (as you know) with a electronic ignition trigger with backup points in the modified distributor. My fuel is gravity fed into a Marvel Schebler carb. With 176.0 hours on this engine it's beginning to earn a bit of trust from me. ;)

Dale
N319WF
 

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
3,749
Location
Calgary, Alberta, Canada
A Titan T51 experienced the same sort of issue and forced landed in a tomato field. The owner had decided to make a set of aluminum (instead of the usual steel) exhaust stacks. Predictably, they melted and the hot gases melted the crank sensor wire through.

Subsequently, we provided the sensors with a 3 foot length of fire sleeve. We now provide Tefzel cabling plus firesleeve as this is a critical component.

What sort of fellow would think it's ok to use aluminum for an aircraft exhaust system though?
 
2
Top