How accurate is GM's mpg computer?

[Blue_Goose]

Just wondering..the DIC usually is a bit higher than if I do the math myself.....so accurate is the computer computing your mpg?


Goose

 

[Blue_Goose]

I mean so HOW accurate is the computer computing your mpg?

One of these days I'll be able to type



Goose

 

[Titan]

My parent's GM computer is about 2 mpg too high. My father doesn't care...he brags about his mpg as if it is accurate. My Subaru computer is anywhere between dead on when I do the math up to 2 mpg higher than when I do the math. This is another example of why a couple mpg may be meaningless, and why I often pop into a conversation where someone says they got 0.5 mpg better by changing a fluid. It takes more than a couple of fill-ups to have a chance at establishing a trend...even after 20 fill-ups, you still have SO MANY variables....

 

[Blue_Goose]

Well is it really higher or is doing the math not all that accurate giving all the variables?


Goose

 

[dwendt44]

The computer, IIRC, calculates fuel mileage by subtracting the fuel returned to the tank from the fuel sent to the engine. Doesn't account for evaporation or other losses.
The one-way fuel injection system is different.
Still fairly close to actual.
Personal calculations are only 'close' anyway. It's almost impossible to fill the tank to the exact same level every time.

 

[JTK]

Yep, these are a one-way/returnless system with a pulse width modulated fuel pump. My Cobalt is the same as goose. I can get 37mpg on multiple hywy runs, yet my total miles driven / gallons pumped calc shows ~32mpg. I'm sure the DIC and manual calc would be much closer on very long hwyw runs where you burn thru an entire tank in one run.

Joel

 

[Titan]

Quote:

---

Well is it really higher or is doing the math not all that accurate giving all the variables?


Goose

---

It really overstates the MPG...because it NEVER understates it, and, after 35,489 miles since September, 2006, if the computer were accurate, it would have shown up by now.

Generally, the computer is exactly 1.6 mpg too high.

 

[rriddle3]

Mine shows 2%-3% higher than computing it myself.

 

[Shaman]

My 97 GTPs was pretty close until I got a DHP PCM. After that it was about 2 mpg too high.

 

[crw]

The first thing we all need to do is find out exactly how fast we're going when the speedometer (and hence, the odometer) says a certain value, by calculating it by time and distance, or by using a GPS. In my Jetta for example, when the speedometer says 79, I'm actually traveling at 75. So even if you calculate it by hand, it's important to know how accurate your odometer is, or is not.

 

[crw]

I suppose the most accurate computer would be a precisely metered system measuring fuel on the one hand, along with an "always on" GPS system on the other. But always remember that your car doesn't know how fast it is going, but only how fast it's going based on the mechanical bits that make up the odometer reading. I would imagine that some of them are pretty accurate, and hopefully none of them read "low".

 

[Texas Aggie]

You've also got to take into account the accuracy of the gas pumps. There was a spreadsheet recently sent out of inaccurate gas pumps in the Dallas area. There were quite a few stations that pumped more or less than the actual quantity measured.

 

[Scott P]

The one in my GP was pretty stinkin' close the vast amjority of the time.

 

[Jason2007]

my silverado is usually 1 mile high on its calculation

 

[mstrjon32]

Sorry about the formatting, but here are my records of the GM Trip computer's calculations vs. my calculations for fuel economy. This is data for my 2006 Monte Carlo powered by a 3900. On average, I find that my calculations and those of the trip computer are within 1.93% of each other.


Avg. MPG Comp. MPG Percent Error
#DIV/0! #DIV/0!
18.2 17.08 6.58%
19.3 18.75 2.94%
18.5 17.99 2.82%
22.9 21.85 4.81%
19.5 18.97 2.82%
19.2 18.97 1.23%
26.0 25.09 3.63%
27.2 27.26 -0.22%
26.2 26.37 -0.65%
25.1 24.98 0.48%
23.3 22.25 4.73%
25.9 25.78 0.45%
26.9 26.99 -0.34%
13.6 12.61 7.87% (I drove the car up Pike's Peak, here)
25.3 25.00 1.20%
26.0 25.81 0.74%
25.1 24.38 2.97%
25.9 26.20 -1.14%
26.5 26.50 0.01%
23.9 24.16 -1.09%
24.7 24.19 2.10%
24.8 24.20 2.50%
23.2 22.71 2.14%
24.9 25.15 -1.01%
21.8 20.82 4.70%
23.5 23.39 0.47%
21.2 20.91 1.40%

 

[Blue_Goose]

Which one is you calculation? The one on the right? That is pretty close either way....It isn't that close on my HHR


Heh Pikes Peak...we drove up that 2 weeks ago in a rental Mitsubishi Outlander.....I wonder what kind of mpg we got there lol


Goose

 

[mstrjon32]

Sorry, I suppose that is a touch ambiguous. The "Avg MPG" is what the computer indicated, the "Comp MPG" is what I calculated by dividing trip miles by gallons used.

 

[LoneRanger]

DIC on my 2007 HHR 2.2L ecotec w/ auto trans was showing 30.6 mpg avg last week. Actual computed out to 29.1 using same pump at same station as prior fillup (probably still margin for error). This was suburban driving, not interstate.

 

[tmorris1]

My question is how does anyone know that their calculations are accurate? Unless you physically measured each gallon and refilled it to the exact same spot after driving, it won't be accurate. If you refill with a pump until the pump shuts off, I bet there is a gallon or more of varability.

 

[mstrjon32]

Well, based on the fact that most of the calculations that I come up with are within 3% of the trip computer, I'd say that either I'm pretty close; or both me and trip computer are grossly miscalculating the mileage. Also, even if I am off by a gallon, my calculations are consistent enough that it shouldn't matter.