I could use some help with an experiment...

[camrycev6]

MikeHTally,

I would say the 6-10% deviation could be attributed to driving styles, environmental conditions, and any mods folks might have. I think the conclusion is evident with the data that trip computer is consistently high when compared to the actual MPG. I was most disappointed that it did not improve despite driving several hundred more miles without a reset. It simply cannot be used with any reliability, or at best, subtract a semi-random 2-4 MPG to get a good "idea" of what you actually got.

 

[davegutz]

I see the same over-estimate of mpg. My own theory is that the system is calibrated for 100% gasoline whereas all the gas I buy is at least 10% ethanol. Since ethanol has only half the energy density of gasoline then the mass flow meter is going to be off but so will the mileage. I'm not sure now this nets out but it's a large enough factor to consider. We would need to know how the system measures flow: mass flow or volume then how it calculates mpg. This is question for Mazda, me thinks.

 

[john blutarski]

Jesus, I didn't even know we had a MPG readout on the trip computer. I've always calculated it based on miles driven / gallons to fill up.

 

[camrycev6]

I see the same over-estimate of mpg. My own theory is that the system is calibrated for 100% gasoline whereas all the gas I buy is at least 10% ethanol. Since ethanol has only half the energy density of gasoline then the mass flow meter is going to be off but so will the mileage. I'm not sure now this nets out but it's a large enough factor to consider. We would need to know how the system measures flow: mass flow or volume then how it calculates mpg. This is question for Mazda, me thinks.

What ever system they use... they need a new one!

 

[m5rMS3]

Even though it doesn't give you completely accurate readings, the readings do seem to be fairly consistent. The main reason why I monitor my fuel consumption is because it is a great early warning system if anything is going wrong with your car. If you can establish the norms on the computer then you will know something is up if that drops significantly with no other major changes (driving style, temperature...)

 

[camrycev6]

Even though it doesn't give you completely accurate readings, the readings do seem to be fairly consistent. The main reason why I monitor my fuel consumption is because it is a great early warning system if anything is going wrong with your car. If you can establish the norms on the computer then you will know something is up if that drops significantly with no other major changes (driving style, temperature...)

I will give you the readings have a relatively high degree of precision, but a poor level of accuracy. For me, that just doesn't work well for me. If perhaps, the precision was extremely high, I could at least adjust by the deviation from the known actual value. (Based on the miles / gallons calculation.) Unforunately, the precision is still not high enough to be useful, unless as you mentioned, some radical change takes place in the readings. I would submit, if that was the case, you would have other obvious signs of a problem(s) as well, making the computer trip readings superfluous.