MPG onboard calculator

[Dale_K]

I thought the system increments a running total of the amount of time the fuel injectors have been commanded open to measure the fuel used. It ought to be pretty accurate given the precision of the injectors. I don't think there is any estimation involved.

[VTXrider]

Originally Posted by Chevy_cowboy

Unless you follow someone that kicked off the pump and then drained every drop out of the hose before hanging it back up. Not many do that, but some do.

So you can do the same and it still evens out.

To the OP, I have no clue if it is an estimate or what, I always figured that the computer could measure what comes out of the injectors (or fuel pump) and calculate against distance traveled. Either way there can still be error in the measurement of fuel or estimation. I have tried to do this same tracking but didn't care enough to follow through. When I did it my calculated miles tended to me 1-3% better than the DIC. I figure there could easily be enough variation between how "full" I filled the tank to account for the difference between 1% and 3% as this is basically 1/3 of a gallon. I just assumed the average over time was about right.

[FunkedOut]

the computer knows without a shadow of a doubt how long it has pulsed each injector since the last time you reset the average MPG entry.
it also knows without a doubt the wheel spin count.

now, you have to trust in the GM engineering crew to model the injector performance accurately (fuel flow vs pulse width), and the diamter of the wheels to be accurate.

the value used for miles traveled is the same for your spreadsheet as it is for the ECU.
all of the variance lies between injector modeling and how consistent you can fill the vehicle each time. does it always auto-shut-off at exactly 18 gallons? probably not. this fill-up's slight overfill is your next fill-up's improved mileage.

I'd trust GM's injector modeling over a number of different fuel pumps' auto-shut-off consistency.

[Castrosua]

I did think of that. How full is full, either being done at the first stop, and topping it off. But overall the variation should go to 0. (limits, I do teach calc lol)

And dont joke about that. I nearly lost it when the dealership (with my Crown Vic) had my car for 8 weeks and they added a gallon at a time, and messed up that and the next calculation.

Of course that paled compared to when I found out why my car was in that long, that they dropped it off a lift.

And I keep records of every maintenance action, like "Engine whining under load" and corrective action "Could not duplicate"

Actually writeups like that help a LOT long term if something does got wrong. You can spot when something may have happened and maybe trigger the cause. Less important now when the cars are new, but 10 years from now, big help

[Castrosua]

Originally Posted by VTXrider

So you can do the same and it still evens out.

To the OP, I have no clue if it is an estimate or what, I always figured that the computer could measure what comes out of the injectors (or fuel pump) and calculate against distance traveled. Either way there can still be error in the measurement of fuel or estimation. I have tried to do this same tracking but didn't care enough to follow through. When I did it my calculated miles tended to me 1-3% better than the DIC. I figure there could easily be enough variation between how "full" I filled the tank to account for the difference between 1% and 3% as this is basically 1/3 of a gallon. I just assumed the average over time was about right.

I did that on my Cv (oddly Im talking about that car a lot on a Camaro forum lol, but hes and yes hes a boy car, hes hurt and I cant wait to fix him back up.) The filler neck is on the left, and has a locking cap, so I always drained the hose and shook it like I was in front of a urinal

The Camaro being on the right, and me having my injury means Ive gotten lazy, although if I see one more dumb clerk just let the cap dangle against the quarter panel, Im just going to have to put a locking cap so I have to get out and do it myself

[Chevy_cowboy]

Originally Posted by VTXrider

So you can do the same and it still evens out.

Sure, but how would you know what the last guy did? lol

[ChrisBlair]

Originally Posted by Squillo

That's because when you're idling you're still burning fuel. How does hand calculating change this?

It is true that every car gets the same MPG while sitting still: zero mpg.

What I'd like to know is just how the rated MPG is determined for 'city' driving: how many stops and for how long, for any period of time. Way too subjective for my taste. I can drive through Boston at 4 am and get 17 mpg, and then at 4 pm on the same route, get 8 mpg.

[CrazyRED]

I think variance is more likely that your car is sitting at different angles when filling it up. Even if you fill the neck all the way to the top if you're on some type of incline in any direction yaw, pitch, or roll you're gonna have a bubble of air in the tank somewhere. If you're within 5% I'd say the computers doing a pretty darn good job. Even with your highest variance you're still less than 3/4 of a gallon different.

(edit) Actually the way you calculate it would account for any of those types of variances because of the total amount added. Either way, I still still the difference in the real time tracking and your calculations are pretty close and well within my acceptable tolerances. Then again, I never take it easy on the gas and really don't care lol

[OrangeVert]

Originally Posted by ChrisBlair

It is true that every car gets the same MPG while sitting still: zero mpg.

What I'd like to know is just how the rated MPG is determined for 'city' driving: how many stops and for how long, for any period of time. Way too subjective for my taste. I can drive through Boston at 4 am and get 17 mpg, and then at 4 pm on the same route, get 8 mpg.

+1 down to the Cape or on the way home to Nashua, Chris. That traffic is brutal! I thought it was just distance / fuel level drop, but could be wrong.

Also, don't really care about mileage too much but being an anal retentive engineer type with OCD, always want to know exactly what my car is doing!

[Backspin9]

The MPG readings I get on my 2015 V6 have always been within .1 or .2 of my manual calculations. That's pretty darn close. I spot check it manually once in a while but don't really obsess over it. The car only has 1600 miles on it and the MPG are slowly getting better. I typically get 275 to 310 miles per tank doing mostly suburban driving and only use about 1/2 tank a week. I really need to get it out on the highway a little more.

[Castrosua]

I dont obsess over it (i compulse over it?). I just like knowing these little things about my cars, and this is a peculiarity mostly because none of my other cars have sophiticated computers.

My previous newer car still had analog odometer.

[Scalded Dog]

Castrosua, I'm not QUITE as bad as you are, but on the same end of the sorta OCD spectrum regarding mileage calculation. I find my onboard calculator to be within 2% - 3% of actual, and it swings up or down, averaging (long- term) pretty close to accurate.

I'd love to know HOW the car calculates it, though: Does it actually measure fuel flowing through the lines? Does it estimate based upon throttle positions, angles of ascent/ descent, GPS data? Does it weight the fuel tank and compare it to miles driven?

[Castrosua]

Originally Posted by Scalded Dog

Castrosua, I'm not QUITE as bad as you are, but on the same end of the sorta OCD spectrum regarding mileage calculation. I find my onboard calculator to be within 2% - 3% of actual, and it swings up or down, averaging (long- term) pretty close to accurate.

I'd love to know HOW the car calculates it, though: Does it actually measure fuel flowing through the lines? Does it estimate based upon throttle positions, angles of ascent/ descent, GPS data? Does it weight the fuel tank and compare it to miles driven?

Exactly. Im curious too. I would think it would be maybe based on fuel pump? although that wouldnt be so accurate, although would account for the variation since pumps are in the tank now so you have all that fuel that may be under different pressures based on how you drove right up until shutting down.

Is it milieage from the computer, with I think might be GPS based too. I was looking at my odometer since I like to see with my own eyes rollovers (espcially those 100k ones that go back to 00000) but i was at 11999 then within 2 tenths of a mile I went to 12001.

[sledgehammer]

Originally Posted by Castrosua

Exactly. Im curious too. I would think it would be maybe based on fuel pump? although that wouldnt be so accurate, although would account for the variation since pumps are in the tank now so you have all that fuel that may be under different pressures based on how you drove right up until shutting down.

Is it milieage from the computer, with I think might be GPS based too. I was looking at my odometer since I like to see with my own eyes rollovers (espcially those 100k ones that go back to 00000) but i was at 11999 then within 2 tenths of a mile I went to 12001.

I'm not sure what GM does but I would think it's in line with the other manufacturers. When I went to BMW service school years ago mileage was calculated by the computer counting the number and dwell of injector pulses. By doing that it has a very accurate tally of fuel used it then divides by miles driven and presto MPG. Miles driven are calculated from wheel revolutions not the odometer. The odo and speedo can and usually are off by a small amount reading high. This is due to federal regulations on speedometers (and odometers) that must read accurate or higher than actual speed and never lower so most manufacturers adjust them to read faster. BMW used to do +1 MPH.

But this was all over 20 years ago (with Bosch systems) who knows what they are doing now.

YMMV