# Average Mileage

1. Oct 18, 2009

### DaveC426913

How does my 'average mileage' meter work?

When I reset it, it starts averaging right away, so the first few miles have a big impact on the value. But after a hundred miles, each mile (including those first few) has a tiny impact (1/100th), right?

If my driving habits change, these changes will hardly be noticeable. After thousands of miles, virtually nothing I do will affect it.

Or is this wrong? Does it weight them differently? Does it start "invalidating" data after a certain point? What point?

2. Oct 18, 2009

### Staff: Mentor

It is almost certainly just a continuous totalizer calculation, not a decaying average. Ie, your first description, not your last.

3. Oct 18, 2009

### Argentum Vulpes

The current MPG readings work on a data base stored in the vehicles computer. The data base is formed in shop with the car on a chassis dyno. Using variables of vehicle speed, engine RPM, and the MAF sensor, the amount of fuel being consumed is measured and the data base if formed. Out on the road the car looks at the three variables, grabs the corresponding data and gives out a MPG. That is why the cars MPG and actual MPG are always off a bit.

There are a limited amount of data points that the cars trip computer could store (not the one displayed by the odometer), my parents old doge caravan would reset its trip computer every 999.9 miles. So my guess is that yours might be the same.

Last edited by a moderator: Oct 18, 2009
4. Oct 19, 2009

### Staff: Mentor

Where are you getting that from? It only takes two variables to calculate the fuel economy (fuel flow and distance) so why would it need to read three and look up the data in a table?

Car computers are required by law to collect this information and output it, so feeding it to a calculator, then a display is easy. The problem with built in trip computers isn't that they are looking at old data, it is that it is very difficult to measure fuel flow rate accurately because the flow rate is so low. I have an add-in trip computer with fuel flow calibration capability that results in highly accurate mpg calculations.

5. Oct 19, 2009

### Argentum Vulpes

I can't give you an exact link to my info because I'm going off old (2004) information of how the system works. Also a cars ECU (engine control unit) is different from its trip computer. The trip computer gets its sensor inputs from the ECU, usually via the OBD (on board diagnostic) port.

There is no way to measure fuel consumption in a car. The system has no sensors on the feed or return line to tell the computer how much fuel is being consumed, most systems have a pressure gage in the fuel system to make sure the injectors are getting feed gas. So accurate measurements of fuel consumption can only be done in the lab with a special outfitted fuel system. Also it is an extra sensor to add to the system when the computer has no realistic need for its proper running. MPG is a consumer convenience, so getting it from readings from three sensors that the vehicle needs and extrapolating it from a table is ok by the car companies standards

The reason that vehicle speed is used is for figuring out miles. Engine RPM is self explanatory on its measurement, faster it spins more fuel being consumed. And the MAF (mass air flow) tells how loaded the engine is, more air being sucked in the more fuel is being used.

For example a car on flat level ground going 60 MPH is going to be around 3K RPM and have a fairly low amount of air going into the engine. Turn on the AC the RPM and speed will stay about the same but the MAF will start seeing more air coming through it.

As for storing information I think they are only required to store information for about 6 seconds to be recalled after a crash has been detected, or any fault codes until a OBD scanner properly resets the codes.

6. Oct 19, 2009

### DaveC426913

That means I should reset it occasionally, so if my driving habits change, it won't be pulling old data, right?

7. Oct 19, 2009

### Office_Shredder

Staff Emeritus
It depends. What are your goals regarding this measurement?

8. Oct 19, 2009

### tyroman

When I start a long road trip, I always wait to re-set my average mpg until I'm on the freeway at cruise speed. This prevents my average reading being "contaminated" by all of the stop-and-go driving in town.

Except for fuel and food stops on the road, my trip average is mostly highway mileage. There is also an "instant mpg" display which can help me find my best driving speed for prevailing conditions.

My "fuel range" display is based on the current average mpg and the fuel guage reading. If my average mpg is "contaminated" with a lot of in-town driving, my "fuel range" (on the highway) will indicate less range than I really have.

9. Oct 19, 2009

### DaveC426913

Well, I'm actually asking a general question. But I'll digress for a moment.

This particular time, I wanted to test my mileage under a range of conditions while hauling my boat.
I reset my avg mi meter prior to departing on a 90km trip with my boat in tow. I got my reading.
Then I reset it again for the drive home - sans boat/trailer. This gives me average mileage readings for the same 90km trip under loaded and unload conditions.

End digression. That got me thinking about the more general use of an avarage mileage meter.

Up to this point, I hadn't reset my average mileage counter in years. That means it has been weighing every mile I drive today with the same weight as every mile I drove years ago. If my driving habits have changed (say I've moved to the city from the burbs, or I'm more cautious now that I have kids), it will now be giving me inaccurate readings. Furthermore, my current driving habits will add new data, but this new data will have virtually no effect on the previous thousands of miles of data racked up.

i.e. in general, one should reset one's average mileage computer every once in a while, to clear out old data.

10. Oct 19, 2009

### Staff: Mentor

That's what I was referring to.
Ok, I found a link that confirms that fuel flow rate is determined by measuring airflow and rpm and finding fuel flow from a lookup table - with adjustments for trim due to oxygen sensors, etc. http://www.myscantool.com/faq.html#Can ProScan monitor fuel consumption

However, I don't think that quite fits the way you described the situation before: the OBD II port still reports fuel flow rate and the fuel flow rate and speed is used to calculate MPG. What you said before sounded like the mpg itself was read off of a table. Why does this distinction matter? It matters because my OBD II tool includes a feature for calibrating the fuel flow reading the car's computer reports, using this calibrated data to more accurately calculate mpg.

Whether any factory trip computers have this same feature or not, I don't know, but I haven't seen one (I've only looked at a handful).
Clarification: a totalizer function records one data point and continuously updates it. Ie, your car odometer records the total miles traveled and continuously updates it. It doesn't data-log second-to-second speed (though some OBD II tools do).

11. Oct 19, 2009

### Staff: Mentor

I would, yeah. For informational purposes it is probably useful to measure it tank-to-tank or trip-to-trip. In particular, you can manually calibrate it by measuring it tank-to-tank and comparing it to what your gas pump and odometer tell you.
I have found that due to the fact that I now have fuel economy data staring at me while I drive, I have changed my driving habits to drive more efficiently. So yeah - you certainly can use the information in that way.

12. Oct 19, 2009

### DaveC426913

Ah OK. I've always assumed it was a long-term process. For some strange reason, it didn't occur to me to use it on a trip-by-trip basis. I say it's strange because the same computer also has an electronic odometer and a elapsed-time counter, and a distance-to-empty counter - all if which need to by reset (whether manually or automatically) on a per-use basis.

The only feature that does not get reset on a regular basis is the compass.

Thanks.

13. Oct 20, 2009

### Argentum Vulpes

Russ could you find any car out there that dose have a flow sensor or some other type of direct fuel consumed measuring device in its fuel system, and I'm not referring to the float in the tank to tell how much gas you have. I've looked for and talked to a buddy that works on cars for a living we can't find/think of one that dose. Also by your link it dose describe how MPG is found almost exactly how I described it. I didn't have the part about the O2 sensor or the air temperature sensor, because I wasn't aware that those sensors were used in that calculation. The part about the fuel trim is just talking about the dwell time for the fuel injector.

Yes it dose get its information from a table. Again from your link.

As for the storing of data I guess I wasn't clear that I was talking about the ECU. It dose data log second-to-second data but only stores that for a brief period of time, it also dose have a totalizer function but that is in relation to air fuel mixtures and is all internal to the ECU. That is why you car will run a bit on the bad side after it has lost battery power for a bit and the information it uses to adjust the air fuel mixture is lost.

So if for example you fuel pump were going toward it way to breaking down. The fuel flow or pressure would start to drop. Then the needed dwell time for the injector would start to increase to get the necessary air fuel mixture. The ECU would note this (from the O2 sensor) and automatically add milliseconds to the injector to get the mixture correct again. Next the system would also make an adjustment to the air fuel mixture table, before close loop data was looked at. If at some point by the adjustments coming from the totalizer function input about dwell time was beyond a certain accepted value an error code would be generated and the check engine light would come on.

14. Oct 21, 2009

### Staff: Mentor

I think you misread the paragraph that starts with the word "however".