adding a fuel economy guage (laptop style)

[robertisaar]

another little project. may be more fitting for the ECU tuning sections, but it's not really tuning, so i decided off topic and possibly moved later was a better idea.

here is what i have thus far:

fuel economy read out: simple enough, make a guage for tunerpro to allow a fuel-economy read out.

major things effecting fuel economy: injector flow, BPW and vehicle speed. those are influenced by other factors, but with those values alone, i should be able to calculate a fairly accurate number.

things that are going to be interesting to overcome:

vacuum changes the injector flow rate due to the FPR. now, it should be a rather linear change, but by what exact amount, i am unsure. since this is targeted more for MPFI/SFI OBD1 vehicles, i can assume 43.5psi and the corresponding flow that will come with it when there is 0 vacuum. i would like to see exact numbers in other situations if anybody has them.

this should be an interesting equation, but TP V5 can handle it...

[robertisaar]

slight update, managed to pound out some more info. have yet to get to the flow equation yet, but i am getting there.

fuel economy read out: simple enough, make a guage for tunerpro to allow a fuel-economy read out.

major things effecting fuel economy: injector flow, BPW and vehicle speed. those are influenced by other
factors, but with those values alone, i should be able to calculate a fairly accurate number.

things that are going to be interesting to overcome:

vacuum changes the injector flow rate due to the FPR. now, it should be a rather linear change, but by what
exact amount, i am unsure. since this is targeted more for MPFI/SFI OBD1 vehicles, i can assume 43.5psi and
the corresponding flow that will come with it when there is 0 vacuum. i would like to see exact numbers in
other situations if anybody has them.

fuel temperature also influences this, we can use MAT to factor for that. at 60*F, it is 6.152lbs per
gallon. gasoline expands by .069% per degF. unless the ECU automatically compensates for this. which would
make the MAT compensation table make sense... now i'm changing the table due to this new information, even
though the original may take into account both air and fuel density. i'll account for both in one table
since it seems that the table may only correlate to one. interestingly enough, the stock table was off by a
maximum of .03 compared to what i have found to be correct. oh well, what was good is now better. BHAK was off by .09, but still, that's less than 10% error. good job GM???

this should be an interesting equation, but TP V5 can handle it...

MPG = (BPW*flow rate)/MPH

equation may be reversed, i never paid attention to that part of algebra/trig

ex

55/(3.5*16.7) = .94
53/(1.2*14) = 3.15

[robertisaar]

i've pretty much finished it, just not sure about the exact difference the vacuum makes to the pressure and what i should use for something else.

here is my dilema:

1 16.7 lb/hr injector flows 2.71 Gallons per hour. if all 6 injectors are flowing 100% duty cycle, that's 16.28GPH. at 50MPH, that would create ~3MPG. 50% at 50MPH is ~6MPG, 25% is 12MPG, 12.5% is 24MPG, 6.125% is 48MPG.

now: do i use BPW or injector duty cycle for this part of the equation? i seem to be having a mental block about this, i can't make either of them sound wrong for some reason.... i want to say duty cycle, but am not sure.

[robertisaar]

thought about it more. i think duty cycle makes more sense, but then i'm going through some logs from the GP. holding a steady 53MPH, i'm getting a duty cycle of 5.09%. by what i have came up with, that means i should be achieving ~64MPG(assuming a 16.7lb/hr flow rate).... is this why GM included a "empiricaly derived mutiplier" for the $8F mask? a stock TGP number shows .527, i'm not sure how that number came to be, but when i multiply my "64MPG" by it, i see 33.7MPG. THAT seems a lot more believable.

i may have to do some code digging to see what's up with this.

[robertisaar]

i got some help from the man himself(mark mansur) and got something working, though it's not as accurate as i want. if i can get some vacuum compensated numbers, then it will be better, if not perfect.

[2.8-350TBI]

You are a good man for doing this, shhh! Dont let the big corporations see this!

[robertisaar]

You are a good man for doing this, shhh! Dont let the big corporations see this!

i highly doubt they care about me playing with my 'antique' engine management system.

i may have a decent vacuum compensation number now: supposedly the fuel pressure drops ~7psi at idle and i can extrapolate from there what i want.

the numbers i'm getting from mark's equation come out to pretty realistic numbers. interestingly, you can make a "trip computer" using the histograms function in combination with this guage. make one chart for everything, that way you get a true MPG number for the full trip. then make another to show the individual cells to show you where you are getting the best economy at.

[robertisaar]

HELL YES, check this action out! my latest drive and therefore log.
i'm happy about the ~31MPG i'm getting, even when including idling.

vertical axis of the full table is MPH, and the horizontal axis is MAP.

[bszopi]

I'd be interested in 2-3 full tank calcs and comparisons. Until then, its just funny math...

[robertisaar]

i'll be getting there soon enough, but the trip calc seems correct, at least for when the wife is driving, manages to pull an awesome 23MPG...

[Joseph Upson]

For a vacuum effect correction I would take the difference of the fuel pressure at idle and then at idle with the vacuum signal disconnected and average that pressure difference over the kpa range. What does your equation look like?

I would think it would include:

MPH, Kpa, injector duty cycle, fuel pressure vacuum correction, engine temp, rpm.

For a calibration effort if you have cruise control you can set it at various speeds and take a few logs over a stretch of a few miles at steady and calculate the fuel consumption by hand for say 60 mph for 5 minutes and the amount of fuel used based on the duty cycle average over that stretch and compare it to what the program produced.

I was hoping someone would enter a nice little subroutine to tunerpro that would calculate 0-60 and 1/4 mile time, or better yet spit out all the coordinates where detonation and lean conditions occured instead having to scan an entire datalog.

[robertisaar]

I was hoping someone would enter a nice little subroutine to tunerpro that would calculate 0-60 and 1/4 mile time, or better yet spit out all the coordinates where detonation and lean conditions occured instead having to scan an entire datalog.

mark mentioned this recently, so it may become reality in a few more releases. as for detonation and lean conditions: just make a histogram and use the 'deviation' option to track knock retard and BLM/INT.

my equations come out to this:

flow rate compensation:

(104.45-Y)/42.932

Y being MAP, it allows for the 2.2lb/hr drop that 16.7lb/hr injectors experience at full vacuum.

and the economy equation itself

X/((16.7-W)*.0000212*Y*Z)

X being MPH
W being the flow rate compensation
Y being RPM
Z being Final BPW

i tried using injector duty cycle for hours before i contacted mark and he gave me an equation using BPW, but it wasn't vacuum corrected. that's what i added into it.

why the suggestion of engine temp? the stock GM MAT correction tables does a decent job of dealing with the density gain/loss of both fuel and air at different temperatures. the ECU uses this modifier in the calculation of BPW.

[3400-95-Modified]

I have this calculation built into the software I use to scan sometimes... It was using a constant of 14.7 for the a/f, but I edited that out to make it use my wideband pid for a more accurate calculation.

[robertisaar]

why would someone throw calculated AFR in there? if the vehicle isn't running perfectly, it throws the number off... at least with a WB, it would show a realistic number.

but i still don't see the use of including it at all, it's just another variable that could be thrown off by sensor error.