I could use some help with an experiment...

[ericrapp]

Does it make any difference that the fuel tank is not empty when the computer says 0 miles left for 5-10 miles? I guess not just threw that out there

 

[MikeHTally]

The main thing for this calculation is to fill the tank absolutely full. Pump cutoff isn't accurate enough. The downside of that is that filling the tank that full isn't good for the evaporative emissions system. I've been filling "to the brim" for several tanks now. Guess I'll do it a few more times (just for the group (lol2) ). Maybe I can remember to reset the computer, too. Normally I don't pay any attention to that particular gadget.

 

[ericrapp]

oh s*** Mike you have to reset

 

[MikeHTally]

I know. I had to run an errand last night, so I filled 'er up ($4.03 at Sam's). The 'puter is reset. We'll see what I get. It'll take about a week for it to need appreciable gas.

 

[clos561]

my dash hawk says 38mpg average LOL i can get 27-28ish with very conservative driving. i havent had my gas readings since i installed my radio. jsut have it set on miles left on tank

 

[camrycev6]

I will be looking forward to your results gentlemen.

 

[thajones]

I have been doing this same thing - ever since my Scanguage hooked up to my old car, I've tried to measure MPG (the old fashioned way) and compared to the Scanguage with the variance it allows you to correct each fillup.

Same thing with the Mazda. My route and traffic are pretty much consistent week to week. Unlike what many say here, my computer will recalculate on as little as adding 2 gallons of gas!

In the 4 months since I've gotten it, I've disconnected the battery twice (for mods) and watched the calculations change, slowly but surely. It started like 4-5mpg off, whittled down, and now, after about 6 weeks, I can safely say it's only 1-2 mpg off at most. I use only a select few gas stations depending on if I'm staying here in Newport News or driving to Maryland (which I do monthly), and while it may not be absolutely accurate down to the mile, when the computer tells me I have 50 miles left, it's pretty close when I add it all up at the next fillup and calculate manually.

It worked well enough for me to determine that my car gets better mileage with the AC on and windows up at highway speeds halfway through the trip (I know, they all say it should, but this is the first car where I did see that. My 3 Nissan prior to this preferred windows down). It made a difference of 4mpg windows up versus halfway down.

I will probably steal the scanguage from my girl (she drives the Sentra now) and, using it's variance adjustment, try for a month and just see how accurate it can be

 

[MikeHTally]

my dash hawk says 38mpg average LOL i can get 27-28ish with very conservative driving. i havent had my gas readings since i installed my radio. jsut have it set on miles left on tank

I was wondering how the DH could be that far off, but it's just getting the same info the computer displays on the dash.

 

[MikeHTally]

... It worked well enough for me to determine that my car gets better mileage with the AC on and windows up at highway speeds halfway through the trip (I know, they all say it should, but this is the first car where I did see that. My 3 Nissan prior to this preferred windows down). It made a difference of 4mpg windows up versus halfway down.

"Mythbusters" had the same results, I believe.

 

[camrycev6]

Modern AC systems are pretty darn efficient. I run my AC pretty much year round. (I may turn it off in the Winter...just depends on the humidity levels.)

 

[tiagotiago]

Is 92 octane (the only stuff I can get) useful for your experiment?

I have archived data calculated from the last 5 full tanks and counting, tough no correlation to what the trip computer says (it's been reading aprox 25 MPG, a far cry from the 17mpg aprox that I'm really getting, though I haven't reset the computer's average at each tankfull).

I run through a tank of gas each week, so I could start this friday/saturday if you'd like.

 

[ericrapp]

Is 92 octane (the only stuff I can get) useful for your experiment?

I have archived data calculated from the last 5 full tanks and counting, tough no correlation to what the trip computer says (it's been reading aprox 25 MPG, a far cry from the 17mpg aprox that I'm really getting, though I haven't reset the computer's average at each tankfull).

I run through a tank of gas each week, so I could start this friday/saturday if you'd like.

sir, gas quality will vary alot from station to station despite the octane rating. My mileage is dropping now since the prices have dropped up here. I think something is up. Seriously I have lost 5 tenths in two weeks. And i always use 93. Something is amiss.

 

[camrycev6]

Is 92 octane (the only stuff I can get) useful for your experiment?

I have archived data calculated from the last 5 full tanks and counting, tough no correlation to what the trip computer says (it's been reading aprox 25 MPG, a far cry from the 17mpg aprox that I'm really getting, though I haven't reset the computer's average at each tankfull).

I run through a tank of gas each week, so I could start this friday/saturday if you'd like.

Yes...that would be fine. As long you use the same gas each time.

Also...I think it is very interesting that you didn't reset your trip computer and the average did not improve. If you did 5 full tanks, that should be close to 1400 miles and if the computer is still way off, that puts a real big hole in the hypothesis that it gets more accurate the more you drive. (Without resetting it of course.) After I finish the first part of this experiment, I am going to run several tanks without resetting it, to see if the accuracy improves.

Thanks for the data!

 

[thajones]

I gotta say, I didn't see more accuracy with doing partial fillups - I'd break out the calculator when I did a partial, and measured miles to empty based on what I added vs it's "average."

It almost seemed like it would calculate what filled the tank and weigh how many miles it got last time and slowly bring it down. After reset, fillups said 475 miles or some nonsense like that. Now it says about 400 miles on fillup.

 

[camrycev6]

I think the most popular hypothesis right now is that the more miles you drive (obviously without resetting the computer) the more accurate it gets. I think several have already agreed (and witnessed) that a single tank (apprx. 300 miles) is not enough to provide accurate results. The computer still is 2-3 miles off the actual MPG. I still want more data to verify this, but I am also going to run multiple tanks to see if the computer does get more accurate, and at what point are we <1 mile off the actual.

 

[tiagotiago]

I think the most popular hypothesis right now is that the more miles you drive (obviously without resetting the computer) the more accurate it gets. I think several have already agreed (and witnessed) that a single tank (apprx. 300 miles) is not enough to provide accurate results. The computer still is 2-3 miles off the actual MPG. I still want more data to verify this, but I am also going to run multiple tanks to see if the computer does get more accurate, and at what point are we <1 mile off the actual.

Yes I remember reading this theory.

Now, this is what I remember it reading, this was some time ago since I've been on the road a bit and obviously the trip computer's average mpg increased a bit. I might be remembering wrong, or made the liters per 100 kms to mpg conversion wrong, so don't take this number too seriously.

Also since my computer reads in kilometers and litres, I'll include those numbers, as well as my conversions, just in case I screw the conversion up.

 

[camrycev6]

No problem... We can just use the 1.609 miles / Km and 3.785 l / gallon. We will be good to go! Thanks for helping out!

 

[FrequentFlyer]

My wife's MDX is always 2-3mpg optimistic also, so it's not just the MS3. Unless they installed a flow meter in the fuel line and it calculated actual MPG, it's not going to be super accurate. I'm sure it's just basing its calculations off injector duty cycle, which is probably another reason why it's not very accurate.

 

[camrycev6]

I hear what you are saying FrequentFlyer, but I want to see if that is a constant 2-3 difference fluctates over distance. It it get more reliable as you drive, I want to know approximately how long it takes. (In miles) If not, and the difference is almost always 2-3 (or whatever it works out to be), then I could use that as a guidline no matter how far I drive.

 

[camrycev6]

Ok... Here is my first update... and it is interesting.

I filled up with 13.001 gallons, and I got 21.7 MPG.

The trip computer reported 25.1... that is pretty bad. We are talking about a 3.4 MPG difference!