I've been running my Edge for almost two months now. I finally decided to compare the AVE MPG calculated to actual by using the manual method.

To my surprise, The gauge was reading an ave of 14.3 MPG but at the time of fill up the manual process came out to 13.2

What's going on or wrong?

 

Nothing "wrong" other then it's just not that accurate. Edge says they will be releasing an update to make the MPG reading more accurate. That was several months ago and still nothing been done.

 

It is two different calculations. The average shown by Edge is the average over the last certain number of start ups(maybe 100?) where the manual is the calculation over the last tank. If you manually averaged all the fillups over the same 2 month period, they might be closer. However, as stated, there is a known problem with the method Edge uses, so might not be closer.

 

It takes approximately 4 tanks of gas to come close to the manual calculation method.
This means avg econ cannot be reset in between fill ups.

 

how accurate is the manual method anyways?

It all depends on when that pump clicks off to determine how much gas you've put in and how full the tank is. I'm sure can vary can't it?

 

I've had the same problem. It's like bluejay said though. If you manual average one tank and it was all highway miles it will not be the same as the Edge unless the driving was the same for the previous tank or two. If you do a mix of driving like I do it will never be dead on but it is usually close, within a gallon or two.

 

If you go to the same station, the same pump at the same time of day and fill at the same rate, it will be very close. There can always be a slight variance, but not much. I top mine off and fill very slowly so as not to get to much air. One other thing, if you will fill up before you have used a half tank, you will get much better mileage.

 

So why is there a reset on the Ave MPG then?

 

I know that the way the gryphons record avg. mpg is by taking the IECON readings. It takes the readings at about one a second and throws them into a formula that simulates the average. Overall it takes about 100,000 readings to be a good estimate and this takes around 4 tanks to achieve. I'd assume the Edge is the same way.

Wish I'd known that, where did you hear that? I'd think the extra weight of about 13 gallons wouldn't help, but then again isn't that only around 60#s? I know over here the Dept. of Agriculture keeps the gas pumps pretty accurate. I often fill by a bill like $20 because I'm lazy and if the price is the same the amount of gas is pretty close.

Seems self explanatory but what if your leaving for a big trip and you have been getting 14 mpg city you'll want to reset to get a bit more accurate reading for the trip. It's not that you can't it's that you shouldn't if you want the most accuracy.

 

Is there anyway to calibrate the mileage on it?

On our GTs/PMTs we give you the ability to adjust the average so that it will match hand calculated numbers

 

There is no end user adjustments, it can only be adjusted in the firmware.

Edge realizes that they have this issue.

The only end user adjustment is not to reset on every fill up.

 

No adjustment for that on the Edge. It requires a programming revision.

 

What 88 and Bluejay said.

After about 4 tanks the average will be closer. The problem is that averages need numbers to be accurate and it takes time to build the numbers. In 4 tanks many people will have a huge variety of "driving" from sitting at stop lights to cruising down the highway at peak mileage. This is an average that may not properly represent what the driver considers is "normal driving."

When you say "adjust" do you mean you can go in and raise/lower a few to make it match hand calculated results and then it incorporates that into the average? That skews the results but I'd imagine it does make the shorter term results to appear more accurate. Many people complain that the averages are reliably 1 or 2 above so being able to fix the average is a good thing for most people. Only problem I see is when the average does get accurate the results will be skewed and the driver wont know it. How does (and often do) your programmers gather the data for mpg? Also what is the maximum sample size? Ex. These take the instant econ. once per second and builds an average untill 100K. A shorter span in recording to maybe 5 times a second would make the average more accurate over the the time of one tank. But obviously that would make it less accurate over the long term like the current 4 tank time.