You might try Cummins-Onan. They are the only small engine FI setup I've seen around in large numbers. At one time they had a conversion for early Onan RV units. Been a while, though.
Assuming you are talking about a spark ignition engine, you can use the Megasquirt ECU which seems to have been used for pretty much everything. You won't get a kit as such but there is alot of practical advice and guidance on the website
It will help to know what sort of application that we are talking about here.
My stock 2007 Honda CBR125R has a fuel injected 12 horsepower single cylinder engine. There are quite a number of other fuel injected single cylinder motorcycle engines. But without even the faintest hint of what the application is ... the rest of us have no idea whether you have need for the gearbox that is inseparable from that style of engine, whether the throttle is manually operated or has to be under some sort of governed automatic control (generator, snowblower, etc), and a ton of other factors.
I'm buiding my version of the "Edison 2" my conservative calculations tell me i need a 15HP net. I'm looking for the most fuel efficient engine possible in that range. Most motorcycle, ATV, snowmobile, engines are built and tuned for power not economy. But i have considered an ATV engine.
Those marine engines are actually carbureted, and the vertical shaft arrangement and lack of starting motor might be inconvenient for the original poster's application. For the application that the original poster is talking about, a gearbox will be needed anyhow, although the motorcycle gearboxes don't have reverse, so that will have to be done by some other way. Honda makes several different fuel injected small scooters and motorcycles, and there is a new (2011) CBR150i fuel injected engine that makes 19.5 horsepower, but you'll have to get one from Thailand. The motorcycle-type geared transmissions are more efficient than the scooter-type CVT transmissions.
The CBR125 engine is a single overhead cam hemi head, the CBR150 engine is a DOHC 4-valve pentroof. Both have rolling-element bottom ends. The DOHC arrangement opens up the option of you tinkering with the cam timing of intake and exhaust independently to optimize it for your needs.
My CBR125 uses 3 litres per 100 km with me beating on it mercilessly.
Bike engine including all the electronics, chain drive to a rear axle (you will need to figure out how to make a differential happen with a chain drive), use the motorcycle's 6-speed sequential transmission as is, have a battery driven electric motor engage somehow for reverse. Maybe swing the whole motor and reverse gearbox with a sprocket attached to it into the path of the chain when reversing and swing it completely out of engagement otherwise to avoid causing extra drag.
Your vehicle had better be very light. The CBR125 only weighs about 260 lbs. Acceleration and hill climbing is okay up to about 80 km/h and can only be described as "extremely slow" above that. In a substantially heavier vehicle, you're not going to have acceleration or hill climbing ability that anyone would describe as acceptable.
19Delta - I am a bit surprised you are only aiming to use an available EFI engine. As the main point of the contest seems to be fuel efficiency I would be inclined to use a 250cc engine modified with a CR of 18:1 or so which would give a strong Atkinson Cycle effect. Limit the compression pressure with LIVC. The LIVC performed either by a specially-ground cam or possibly by using the variable LIVC system of the helical cam which reduces pumping losses.
I have seen this cam run and its idle and part-load economy are remarkable - combined with a very high CR it would have about as high a TE as an SI engine could manage.
Speaking of SI engines - the rules seem to allow diesels - surely this would be a simpler approach?
Possibly again the CR could be raised to 30:1 or so and the compression pressure limited by a fixed amount of LIVC.
I had a fuel injected 20HP Kawasaki V-2 gasoline engine on my 1990 John Deere 285 tractor. It used a single twin jet port fuel injector, and a rudimentary controller, to feed fuel to both intakes. It was open loop, and was meant to replace the carbs to eliminate maintenance. I doubt if it was any more efficient than the carbs. However, this controller was simple enough to reprogram. You could adapt that to a 20hp single without too much difficulty, assuming the parts are still available through Deere.
Kawasaki was one of the engines i considered. They have a single cylinder water cooled and they do have twins that are fuel injected. I'm also looking at a two cylinder kohler diesel that has good fuel consumption, weight is a bit heavy though.
Now i need to find a engine dynamometer in Houston or San Antonio. Has anyone ever made a dyno? How did that turn out? Was it very accurate?
In the early stages of Transonic's gestation, we made our own dynos for single cylinder development engines. We utilized a motor/generator of the appropriate capacity, say 20hp. Some batteries for motoring, a variac, and a resistor load bank for sinking generated power, an ammeter and voltmeter to measure power, and viola!
Simple, and relatively inexpensive. You can built the whole setup for less than $10k.
