Technical Computer average fuel consumption Ca

[Ralf S.]

My computer has always been a bit economical (do you see what I did there? ) when reporting the fuel consumption.

I have to take its word for it, on the "Instant"... but "Average" is something I can measure myself.

Assuming that the mileage it thinks the car is travelling is correct, the computer has always/since I've had it, come in around 8% too optimistic. For example, it might tell me I've managed 55mpg... but I calculate my actual consumption as 51mpg.

Yes... I'm sure I'm calculating the mpg correctly... and the odometer gives me the same distance figures on my regular/usual journies that other cars do (variance of less than 1 mile on a 120 mile trip), so I think that bit is okay. The car has standard 15" steel wheels and 195/65x15 tyres so no variance caused there either.

I'm guessing that the likely culprit is the computer itself (or how it measures fuel use) since it must count or measure how often the injectors fire and how much fuel they deliver (unless it relies on the fuel level in the tank/similar to the "Range until Empty" doodah).

So... is there any way to tweak the computer, so it becomes more accurate?


Ralf S.

 

[sussexa]

short answer yes.
as far as im aware there are many calculations and algorithms in use but if you ever check your speedo to a gps device youl see its off also, so its already flawed in any mpg calc.
your right about it using calcs from the fuel rail pressure and injector open times in micro seconds , it has stored data for solenoid opening times and closing times but with anything complicated errors can be compounded.
so say an injector in your case maybe slow opening by a few microseconds and early closing by a few micro seconds theres been less fuel injected than the ecu thinks, add this to any injector wear which may increase less injected and more leak back and partially blocked orifices again less injected than ecu thinks.
then add fuel rail pressure sensor within spec but reading low again less injected than thought.
then youve got mechanical losses of drive train losses like engine and gearbox friction(oil grade ,quality friction additives), cv joints wheel bearings brake calipers etc all if not as new lead the car to need more fuel(also can be variances on fuel type,brand,lubricity, cetane, even winter diesel especially this time of year) to be burnt to get it x miles than the ecu expects its going to take.
one way is to mod the injector duration map, ive done this on my jtd16v as i have none std injectors fitted so have lost ecu injector calibration data and tuned map to give somewhere near what mpg computer shows.not the best but possible.
also im running larger tyres giving near spot on mph vs gps but as you can see from these fuelly notes comp act to comp mpg different weather different fuels, different additives, and most of all type of driving(commute close to real now but slow long distance (holiday travel especially single trips)now way off) all result in varying mpg from actual tank to tank(havnt modded map across this page)

its probably easier to just know its 8% optamistic and leave it at that.

 

[Ralf S.]

You might be on to something (including the "or just deduct 8%" ).

If I had a slightly bigger (rolling radius) tyre that would reduce the mileage recorded by the odometer.. so assuming that this raising the gearing doesn't cause a reduction in fuel economy because of needing more throttle, then the mpg would be "less miles" divided by the same amount of fuel.. so that would reduce the computer's calculations.

I'm not sure how big an "8%" bigger tyre will have to be... but I'll get my calculator out. I'm planning some 17" wheels for next year, so that might be a good time to tweak it a bit, even if I can't compensate 100%. I'm not all that bothered after all... just curious if there was an easy way to do it inside the "brains" rather than physically.


Ralf S.