school me on fuel injection systems (standalone user here)

[the4ork]

ive had a standalone engine management system in my car for a few years now, and have done all the tuning myself.

www.034efi.com <- im using the stage Ic

anywho, its wastespark, sequential injection... the works

ive seen SDS installed on RV's before...

034 is a great system, but its really meant for automotive systems and specifically for bosch sensors....

i doubt there is alot of user programmable systems out there, but i have no idea

what are some of the better engine management systems out there people are running on the 360/540's?

[rv6ejguy]

I'm not aware of any other programmable EMSs being used on Lycoming engines. Most FADECs offered are pre-programmed for certified engines.

DG's Lancair racer with the Conti had a Microtech EMS fitted in 2007 but it was removed this year with the team reverting to the troublesome mechanical system. http://www.microtechusa.com/

Nemesis had EFI type fuel rails installed I assume for future use of the IE2 FADEC in the wings from Lycoming maybe for next year but they were running mechanical injection this year as well. IE2 should be available next year for some models of Lycomings.

Precision Airmotive offers the Eagle EMS: http://www.precisionairmotive.com/ for experimentals.

Aerosance has offered a system for Contis for some time but it has never caught on well and some of the sites with info on it now are dead links: http://www.tcmlink.com/aerosance/

The Thunder Mustang has used Motec for many years: http://www.motec.com/home

One of our vendors is currently working on a kit for 320/360 Lycomings using SDS hardware. Flight testing should be commencing early January in California. We've fitted SDS to a handful of Lycomings and Continentals over the years and they work well, including Mike Dacey's 375 mph Questair Venture at Reno this year.

No reason why the 034 should not work in an aircraft, just watch what sensor failures might do to engine operation.

[rv7charlie]

Tracy Crook at Real World Solutions can set up his EFI for a Lyc, if you're willing to install bungs for automotive injectors.

http://www.rotaryaviation.com/eficont.html

Charlie

[nucleus]

Originally Posted by rv7charlie

Tracy Crook at Real World Solutions can set up his EFI for a Lyc, if you're willing to install bungs for automotive injectors.

http://www.rotaryaviation.com/eficont.html

Charlie

I like Tracy's box because it is actually two systems in one, redundant. It also has a littlle mixture knob.

Hans

[RichB]

Originally Posted by nucleus

I like Tracy's box because it is actually two systems in one, redundant. It also has a littlle mixture knob.

Hans

Redundancy is a good selling point, but 2 duplicate controllers doesn't (IMHO) add much. There's a toggle switch to flip controllers and injectors, but do they share sensors, code implementations, hardware revisions, and so on? I recall reading somewhere about Shuttle flight systems using multiple controllers, the software having the ability to disable an errant controller (eg, 2 out of 3 agree, the 3rd gets disabled) and an entirely separate system with different hardware, different codebase, made by a different team, available for failover if there's a deadlock situation (eg 1 of the 2 left disagrees).

For the DIYers, you could arrange something really redundant fairly easily - consider a simple case with a whizz-bang ECU squirting fuel and firing the plugs. Add an engine monitor that picks up each cylinder EGT (and others ideally) and if it picks up out-of-range values, it switches over (with pilot confirmation) to an alternative ECU from a different vendor: As an extreme example, a Motec M600 with sequential injection + ignition, and handover to a MicroSquirt with batch fired fuel and an e-mag for the ignition. The latter set would run off your backup alt & electrical bus. Cost and weight overhead should be fairly minimal. You just need to monitor the engine vitals yourself... until MGL or Vertical Power or Someone Else steps up with something to help there...

BTW, be thorough when you evaluate an EFI system (or any system for that matter). Check that they're using at least automotive or military grade components. Check that they validate their software and hardware against known standards, and that they adhere to a known development process. This isn't a dig at any vendor in particular... just think of it as analogous to selecting the right materials and having a quality process for the design and manufacture of a mechanical component.

[rv6ejguy]

Originally Posted by RichB

BTW, be thorough when you evaluate an EFI system (or any system for that matter). Check that they're using at least automotive or military grade components. Check that they validate their software and hardware against known standards, and that they adhere to a known development process. This isn't a dig at any vendor in particular... just think of it as analogous to selecting the right materials and having a quality process for the design and manufacture of a mechanical component.

Can you point me to recognized hardware and software standards for EFI design? I'd be very interested.

In 14 years of aviation fuel injection experience we see a lot more engine shutdowns caused by sensor anomalies and inappropriate code designed to protect the engine, especially in the case of OEM units. Hence my point about knowing how the system will react in the case of various sensor failures. If in doubt, simulate at full power during ground testing by disconnecting or shorting sensors one at a time. You don't want to be flying a system which goes full rich if an engine temp or IAT sensor fails open for instance.

There have been some good articles this year in Contact! magazine on EFI failure modes, especially relating to OEM systems.

[RichB]

Originally Posted by rv6ejguy

Can you point me to recognized hardware and software standards for EFI design? I'd be very interested.

I can't point you to anything specific to EFI, but there's a handful of standards for developing automotive embedded systems, and a few for aviation embedded systems. When I wrote that post I was thinking specifically of folks using automotive or military grade components rather than commercial and industrial grade, and validating the code against state diagrams and tables etc, as well as coding to standards like MISRA C rather than being "clever" with the code... and following a formal software engineering development process (SLDC / V-model / Formal Method?) and documenting it - eg would you be comfortable with a large customer auditing your work? My EE-foo is relatively weak, so I'll gloss over the hardware angle...

In an ideal world, we'd go back and review against IEC 61508, DO-178B, DO-254, ARP4761, ARP4754, MIL-STD-882D, etc... but that's not going to be cheap, and perhaps not practical. However selecting the right components / connectors / materials, running static analysis tools over the code, and getting out the pen and paper to document everything is pretty easy (if uninspiring!).

Originally Posted by rv6ejguy

In 14 years of aviation fuel injection experience we see a lot more engine shutdowns caused by sensor anomalies and inappropriate code designed to protect the engine, especially in the case of OEM units. Hence my point about knowing how the system will react in the case of various sensor failures. If in doubt, simulate at full power during ground testing by disconnecting or shorting sensors one at a time. You don't want to be flying a system which goes full rich if an engine temp or IAT sensor fails open for instance.

That is very good advice, and a good strategy for testing. "Limp Home" mode for a car (eg, you have a bad sensor, a check engine light, so it falls back to a default map) is probably not what you want in the air. It's also the reason that parallel and different systems make some amount of sense - you mitigate sensor failure and design/developer (or product selection!) error. It's a good question to ask your vendor - what happens if (say) the airflow sensor reads incorrectly? Can the system even determine that it does? Get them to show you on a test bed. They don't have one? Run.

Originally Posted by rv6ejguy

There have been some good articles this year in Contact! magazine on EFI failure modes, especially relating to OEM systems.

I think I heard something about that. Will take a look... Thanks.

[kbehrent]

I have the Eagle EMS system from Precision which includes both electronic fuel injection and electronic ignition. It has completely redundant systems including dual computers and sensors and utilizes it's own battery backup. A really nice/clean setup and all components are made to install on the engine side of the firewall.

Aerosport is the most experienced engine builder supporting this system and they have done a lot of dyno work with the system on 320s, 360s, and 390s.