Would you consider FADEC?

[RichB]

FADEC...................would be like putting an automatic transmission on a Harley...

L.Adamson ---- RV6A

Nah, it's more like putting electronic injection and electronic ignition on a Harley... except you don't have to, because they come with that from the factory with now. OTOH, if the FADEC includes prop control, then you might be on to something... but even then I'd compare it to CVT rather than an old slushbox. Still, how often do you change gear when you're riding cross country? Maybe FADEC + manual prop control is the right compromise for most folks.

 

[L.Adamson]

Nah, it's more like putting electronic injection and electronic ignition on a Harley... except you don't have to, because they come with that from the factory with now. OTOH, if the FADEC includes prop control, then you might be on to something... but even then I'd compare it to CVT rather than an old slushbox. Still, how often do you change gear when you're riding cross country? Maybe FADEC + manual prop control is the right compromise for most folks.

Yes, "true" FADEC means one lever control. Push to go fast, and pull to go slow.
Kind of like combining a Harley with a modern scooter!

I prefer mags, carbs, blue knobs, and air cooling for airplanes...
Motorcycles actually do better with liquid cooling!

P.S. ---- Ever seen all the electronic wiring to a Subie engine installed in an airplane?

L.Adamson

 

[rv6ejguy]

P.S. ---- Ever seen all the electronic wiring to a Subie engine installed in an airplane?

L.Adamson

By ditching the factory ECU, you get rid of the 1.25 inch diameter, 15 lb. OEM wiring harness and replace it with a .400 inch, 1 lb. harness. A lot less wires required than on your average glass display. Very simple indeed with a dedicated sheathed cable for each sensor.

 

[osxuser]

I agree with the possibility of higher reliability, but I've found that with cars, you have to FULLY understand the system to be able to quickly and accurately diagnose problems, even with a computer telling you what is wrong. On aircraft, the chances of a complete diagnostic system similar to OBD II on cars is highly unlikely right away. Even with that, on a particular system, parts of the system can react in unexpected ways to a failure to mask the true problem. I know this first hand dealing with a VW's ME7 system on my 1.8t GTi, even with OBD II, it's a challenge.

No such problem on a non-fadec setup, because each system is completely independent of the other. I'm a big fan of EI (single!) and EFI by themselves, but when they all talk to each other, strange things can happen. I'm not saying that it won't be reliable, but on a carb/Mech FI and Mag engine, there is a VERY finate number of symptoms leading to some pretty easy conclusions.

We'll see when IE2 comes out how the system reacts to the inevidable failures. I've also been very impressed with a certain TSIO-550 in a Egg lookin airplane flying around northern nevada really really fast. I know what a EFI/EI system can do .

Stephen,

You raised an important issue, but I agree with only part of your conclusion.

You are absolutely correct that FADEC presents a different paradigm to the powerplant mechanic. Today, no doubt there are many A&P's in the field who have never seen a FADEC, and would be rather lost the first time they encountered one. That'll change over time. But you are correct that it does currently present a potential obstacle for the aircraft owner/operator.

About added complexity for the A&P though, I have to disagree. It may appear complex now since it's still unfamiliar, but really FADEC makes life simpler for the A&P as well (although it may have a negative impact on the A&P's income). There's nothing to twiddle with. No set screws, no parts that wear out all the time, no fine adjustments to be made. The system is self-calibrating. Just like modern cars, gone are the days of the "tune-up". The system has a finite number of components (ECU's, sensors, injectors...) and the A&P doesn't go inside them. The A&P doesn't need to understand the internal design of an ECU. When on a rare occasion a component in the system fails, you replace it. It is generally not difficult to isolate the failed component, and in most cases the ECU will actually tell you.

So I guess I'm agreeing and disagreeing. Yes, A&P's will need to update their skills, and that will take time. There may be some hiccups along the way. But it's a step in the right direction.

P.S. Looking at your user name, I presume you use Mac OS X? To draw an analogy, imagine taking a computer technician who has worked on DOS systems for 20 years, and put him in front of a Mac OS X machine for the first time. I'm sure he'll be lost at first. But if he has the willingness to learn, he can become proficient with the new system very quickly, and will probably then recognize how it is vastly superior.

 

[rv6ejguy]

Pretty well all ECUs these days have built in sensor diagnostics. We simply have ERR messages in the 4 gauge modes which show which sensor is bad. The pilot can easily view this info in flight. Defaults help the ECU continue to run the engine in the event of shorted or open sensors in most cases.

 

[Andy_RR]

...you have to FULLY understand the system to be able to quickly and accurately diagnose problems, even with a computer telling you what is wrong.

You have to fully understand engines to figure out any problem, whether with electronic control, mechanical FI or with mags and carburettors. Many a wild goose chase has been had with carbs and mags.

On aircraft, the chances of a complete diagnostic system similar to OBD II on cars is highly unlikely right away.

Whoever releases a FADEC into the market and doesn't adopt automotive OBD technology is nuts and deserves to fail. So much development has gone into this stuff there is no reason to re-invent the wheel. Not that I'm suggesting the ideas won't need some adapting, but there is room for that in ISO15765 and earlier incarnations.

Even with that, on a particular system, parts of the system can react in unexpected ways to a failure to mask the true problem.

Again, if this happens it's either a poor understanding of the diagnosis problem or someone hasn't designed it properly. A full system FMEA is necessary in the design process (several, actually) and can catch all the permutations of failures and deal with them appropriately.

No such problem on a non-fadec setup, because each system is completely independent of the other. I'm a big fan of EI (single!) and EFI by themselves, but when they all talk to each other, strange things can happen.

...but on a carb/Mech FI and Mag engine, there is a VERY finate number of symptoms leading to some pretty easy conclusions.

See my argument above. "Simple" independent systems provide no guarantees of a lack of complex and bewildering interactions! The garden path awaits the poorly educated and ill-informed.

 

[sonny junell]

Of course you want FADEC.

 

[rv6ejguy]

There are simpler diagnostics available than OBD. No need to even plug in a reader or laptop. I don't see the need for complicated diagnostics on a simple system. I've seen some ambiguous faults logged in OBD that have not been helpful in finding problems and even the OEM specific trained mechanics have not solved it the first time.

We use the following standard sensors- CHT, IAT, MAP, crank and possibly a wideband O2 for automatic targeted mixture control when/ if unleaded avgas ever comes. Really all that is left is injector and coil failures which any AME should be able to check those out in about 10 minutes using traditional methods (VOM, eyes and brain).

You can have knock sensing too and use the TPS only for acceleration enrichment so these are non-critical.

If you start going to timed injection with individual cylinder fuel trim and ignition timing, complexity of code increases and so does the cost and the chance for bugs and unknown responses to failures. While this might eek out a couple percent more power and about the same for fuel burn, is it needed on a Lycoming running at relatively constant power settings for most of its life? Lycoming will release IE2 soon with all the bells and whistles with the price to go along with it. I don't doubt the FADEC part will add a minimum of $6K to the engine price and maybe much more.

If you go full blown like this, you might want to use OBD like diagnostics because you may need them. With complexity, many more things can go wrong, clouding clarity. While many poo-poo simplicity on this field, the two most widely flight proven non OEM EMSs in the experimental world both fit this mold. If you don't need it, leave it out. If it ain't there, it can't fail.

Finally, while FMEA is a useful tool, nothing beats actual, real life flight hours (lots of them). Both should be employed to best effect.

So, while this is all being debated, lots of experimentals are already flying with this stuff, both on traditional Lycos and Contis as well as automotive conversions. The latest thing is simply to apply existing, proven hardware and software to the popular 4 banger Lycomings in an easy to install, complete kit for those interested.

I've said this before, resistance to change is entrenched in this field but when a few take the leap, they generally see the light. Certainly Mike Dacey of Reno fame (as Stephen mentioned) just about switched back to servos and mags but crew chief Bruce Patton stuck with it, programmed the problems out and they wowed almost everyone in Sport Class, even the Lyco techs working on high dollar Nemesis with how well it ran. With the ECU managing the engine at 75 inches, Mike managed the piloting part, showing fin to many other higher dollar efforts struggling with conventional engine controls. Mike said he simply could not have gone that fast without the EMS last year. With this being said, Nemesis had conventional controls but totally scienced out by the Textron boffins before the races. The stock Conti system has many more technical issues to solve at these fuel flow rates and is less adaptable than the Lycoming setup.

 

[Kevin Horton]

A full system FMEA is necessary in the design process (several, actually) and can catch all the permutations of failures and deal with them appropriately.

FMEAs are only as good as the imagination of the people who create them. They rarely catch all possible failures. I am aware of one very common type-certificated turbo-fan engine, from a very large engine manufacturer, that had several major engine malfunctions caused by a sensor sending noisy data. The type of unexpected noise in the data tricked the FADEC into misinterpreting the data, and it took the wrong action. Fortunately, all these engines are installed on multi-engine aircraft, and the problem only ever affected one engine in a given flight.

Another turbo-fan engine, from a different very large engine manufacturer, had a fall back simpler control mode that was used if the normal control mode lost critical sensor data. But, the designer got a bit too clever about attempting to ensure no change in thrust when the system automatically switched to the back up control mode. During one ground test, a sensor was deliberately failed to trigger a failure of the main control mode. The engine accelerated to the red line because of the sensor failure, and the pilot got concerned and slammed the thrust lever to idle at the exact moment that the FADEC switched to the back up control mode. The logic to prevent a thrust change on the switch over biased the control so that the engine remained at red line, no matter what the pilot did with the thrust lever. Investigation showed that under other scenarios the thrust could have been locked at idle, no matter what the pilot did with the thrust lever. Needless to say, the logic was revised.

In principle, I like the idea of a FADEC. But, given my knowledge of problems with FADECs from very large companies with huge numbers of engineers, and very large test organizations, I am not in a hurry to put a FADEC from a small company in a single-engined aircraft.

In a FADEC, simple is better than complex, as it makes it more likely that the FADEC's internal logic will always work as expected. Complex systems can have all kind's of unexpected interactions between different aspects of the internal logic.

 

[L.Adamson]

In a FADEC, simple is better than complex, as it makes it more likely that the FADEC's internal logic will always work as expected. Complex systems can have all kind's of unexpected interactions between different aspects of the internal logic.

Problem is; I consider the "red" knob as simple. I've always kept the red knob in mind, as I've always been around high altitude airports. I see no reason to exchange a cheap and simple red knob, for an expensive & possible problematic computer board.

Throughout this thread, there are insinuations that the red knob is somehow complex, ugly, for the un-educated, and adds greatly to pilot load at the worst times. That is nonsense; and I think that anybody who thinks so.................must just be too use to sea level airports.

L.Adamson

 

[Ron Lee]

That red knob is useful to pull way back after startup to minimize plug fouling. I suspect that this option is not possible with a FADEC system.

 

[RichB]

I think the arguments are that the red knob is less refined than the FADEC, and that the complexity is only moved from the FADEC into the pilot. The arguments for and against both have merit, so like tricycle vs conventional gear, it's probably one for the "never ending debate" forum.

 

[Captain Sacto]

Good question

I'm one of those who would not currently consider a FADEC, although I'm planning electronic ignition.

I just don't want to lose the ability to fiddle with things when the engine goes sour, and I'm flying over terrain.

From the responses, however, I might be in the minority.

 

[roee]

That red knob is useful to pull way back after startup to minimize plug fouling. I suspect that this option is not possible with a FADEC system.

With traditional FI, you have to do this manually with the red knob as you described, or it will run too rich and foul the plugs at low power settings. This is a limitation of traditional FI. A FADEC on the other hand will always automatically seek the optimal mixture for your current power setting -- without you having to do anything -- so plug fouling is a thing of the past.

 

[roee]

Big companies vs. small companies

FMEAs are only as good as the imagination of the people who create them. They rarely catch all possible failures. I am aware of one very common type-certificated turbo-fan engine, from a very large engine manufacturer, that had several major engine malfunctions caused by a sensor sending noisy data. The type of unexpected noise in the data tricked the FADEC into misinterpreting the data, and it took the wrong action. Fortunately, all these engines are installed on multi-engine aircraft, and the problem only ever affected one engine in a given flight.

Another turbo-fan engine, from a different very large engine manufacturer, had a fall back simpler control mode that was used if the normal control mode lost critical sensor data. But, the designer got a bit too clever about attempting to ensure no change in thrust when the system automatically switched to the back up control mode. During one ground test, a sensor was deliberately failed to trigger a failure of the main control mode. The engine accelerated to the red line because of the sensor failure, and the pilot got concerned and slammed the thrust lever to idle at the exact moment that the FADEC switched to the back up control mode. The logic to prevent a thrust change on the switch over biased the control so that the engine remained at red line, no matter what the pilot did with the thrust lever. Investigation showed that under other scenarios the thrust could have been locked at idle, no matter what the pilot did with the thrust lever. Needless to say, the logic was revised.

In principle, I like the idea of a FADEC. But, given my knowledge of problems with FADECs from very large companies with huge numbers of engineers, and very large test organizations, I am not in a hurry to put a FADEC from a small company in a single-engined aircraft.

In a FADEC, simple is better than complex, as it makes it more likely that the FADEC's internal logic will always work as expected. Complex systems can have all kind's of unexpected interactions between different aspects of the internal logic.

Kevin, great insights about complexity and the limitations of design verification! But I have to point out that in my experience (electrical engineer) I've found that large organizations with many engineers are actually far more likely than an individual or a small team of engineers to produce designs that are unnecessarily complicated, not self-coherent, and prone to those types of subtle interactions and gotchas. There are reasons why large organizations do poorly in that respect, but that's a whole other discussion... Of course in either case we'll presume that we have competent engineers. But if for example you challenged Scaled Composites and Boeing to a design competition, I'm pretty sure I know which one would produce the more innovative, yet ultimately simpler and more reliably design.

 

[L.Adamson]

With traditional FI, you have to do this manually with the red knob as you described, or it will run too rich and foul the plugs at low power settings. This is a limitation of traditional FI. A FADEC on the other hand will always automatically seek the optimal mixture for your current power setting -- without you having to do anything -- so plug fouling is a thing of the past.

That "red" knob is like the blue constant speed prop. It just takes a few hours to get use to both. Then plug fouling is pretty much a thing of the past..........anyway. But if one wants to spend thousands of dollars to get rid of the red, and more to get rid of the blue, then so be it. FADEC might be the right thing for "student pilots"..

On the other hand; I might give in a small bit. I can think of two fatal accidents off hand, where pilots being use to sea level airports, may have not leaned for high altitude, because of being rushed, & subsequently lost available power when the aircraft was over loaded weight wise. One of these took place just down the road from home.

So once again, it comes down to computer failure, computer programming failure, versus pilot failure........ or a combo of pilot/computer programming failure, such as the twin Diamond Star, where both engines quit because of an initial dead battery that was jumped to start.

L.Adamson --- would rather spend the money on synthetic vision

 

[Ron Lee]

It does not take hours to learn to operate the mixture knob. The accidents mentioned were preventable with a quick discussion of leaning to best mixture during the run-up. Simply it is lean to max RPM during the run-up. I do it before every flight.

Flying at altitude requires additional education/training just as I took a course in mountain flying before I ventured west from Meadow Lake.

 

[rv6ejguy]

In our case, we are able to draw on experience in the automotive market and get many more field hours (millions) on hardware and software before release into aviation markets. This is a huge benefit to thoroughly cleanse any bugs or design problems out that aviation only companies cannot duplicate due to the tiny market size.

In most cases, software and hardware is identical between both markets. The aviation destined units use different default settings in the maps to negate most sensor failures and some higher temperature rated components in the crank sensors.

Interestingly through the aviation market experience, we learned a few things we later incorporated into all other units to improve them.

Finally for those who still want the red knob, we have a manual mixture control as well on the system should you want to fiddle without re-programming. This allows a final backup for the pilot to over ride in case of an odd sensor failure mode, hopefully making it possible to get the aircraft back to an airport.

I personally don't think FADECs will replace carbs and mechanical injection on Lycomings anytime soon in large numbers but they do offer an alternative to those interested.

 

[Rene Bubberman]

Knob fiddling

Personally, I would not want to ditch the blue knob. I like to have manual pitch/prop control because it gives me grip on drag, take-off performance, etc.
As for the red knob: the Fadec tweaks your mixture to optimise the fuel burn, also when you are watching the scenery, looking out for traffic or reading your maps. It does this constantly, directly and automatically and, with all due respect, better than every Human Knob Fiddler. Then there is this BIG difference: for doing this it has a "red knob" for every cylinder individually so every cylinder has it's own perfect mixture independant of it's condition, tolerances and relative temperature.
Therefore a Fadec system claims less fuel burn but also less wear and tear, so in the end a longer operational life and less costs on maintenance or overhaul.
I've only 110 hours on our Fadec, so I guess that is not enough to give a final verdict. So far it feels good and the engine is the smoothest I've ever flown behind. In the end, installation of the system was pretty straightforward and hooking it up to an electronic engine monitor was far simpler than most analog systems.
TCM (present owner of the Aerosance Fadec) confirms, via Mahlon Russell, it stopped working for the Experimental market, but spares and support stay available. I hope someone will jump in this gap; that's what the experimental world needs, isn't it. We'll see.

 

[TomVal]

Now TCM has stopped producing for the experimental market, I hope there will remain people or companies available should we have serious problems.

Rene,

I just got off the phone with Teledyne Continental Motors. They will continue to support existing experimental Aeorsance FADEC engines.

TCM FADEC contact: Tersa Ballard 866 804-9905

Email: [email protected]

Parts Support: Mattituck
Parts Support Contact: Maylon Russell 800 624-6680 ext 305

Data Files Download Analysis: Send to the email address above. They will need the following files:

"edi" files or "bin" files plus the "fault" files

If you are downloading the files from the "EDI" flashcard, DO NOT download the "system" files from the flashcard. Without this file on the card, the card will be unable to download future data. If you inadvertently erase this file from the flashcard, TCM can re-install the file on the card.

Regards,

 

[Rene Bubberman]

Thanks Tom!

This is useful info for us Aerosance-orphans, I'll pass it through.

All the best,

 

[Captain Jack]

FADEC system

I would like to give some information regarding FADEC systems and how well they work. I have installed the Eagle EMS system by precision airmotive on my RV8 and have about 150 hours on it so far. Initially, as with anything, I had to work at getting it configured perfectly but this didn't take long. I have to say that the folks at Precision Airmotive were very interested in making sure that I was happy with the system and would work with me at anytime. The benefits that I have seen so far would make me want to put this system on any airplane that I would buy in the future. I get better fuel economy and most definitely better starting in both cold and hot environments, and yes, I have been to Arizona. I am so comfortable now stopping anywhere during the heat of the day and doing a hot restart within 20 minutes in only 2 prop turns. I never have to play with any controls to do it. It was rather expensive compared to conventional systems but the satisfaction of never having to adjust fuel mixture again is wonderful. It comes with a lean knob since the system is set at the factory to best power (12.5 to 1) and allows you to lean it past peak if you want. I never go beyond best economy (14.7 to 1) though. The nice thing is that it always stays at that mixture no matter what the altitude or power setting. If you have any specific questions about the system I would be happy to give you my findings about them. Let me know.