Of course, I'm cursing myself that this is a blurry image, but I got 92.5 mpg for the first .5 miles on a fresh tank betweek the gas station and work this morning! If only it could stay that high...

 

hey, sometimes you can. i think it must be from the car being warm when it's filled. i have a few shots of 80+ on the 8 miles between where i fill and home.

 

Are you guys serious?

It's just a sampling rate, you just happen to be sampling at the beginning and most likely on a terrain conducive to high mileage.

Over time your sample is much larger and therefore more acurate.

 

For just 0.5 mi? I've done better than that. I've gone more than 2 miles in EV-only mode, so if I did that at the start of a tank, I'd have almost ∞ (infinity MPG!!!!)*.

Good luck and keep trying... but I'd say you really need a good 100 miles before can say it was due to something other than terrain (downhill) or strong tailwinds, etc. My best 100-mile stretch was about 62 mpg.



* "almost" because I'd need some gas to get to 10 mph... the lowest speed where you can get into EV mode.

 

aw, don't pop our bubble. actually, the 8 miles has got some decent hills. and i thought holding 80 for that long was pretty good. i have filled in other locations and not gotten that kind of early mileage........it does seem to happen when you fill up an engine that's already warm. if i go at lunch from work, i never see anything like that number.

wish the high winds around here would die down. i am maintaining 60.9 for 356 miles, but it's a struggle.

 

If you want to be thrilled by big numbers... hit reset while cruising down a very long hill. The same thing happens then. Your sample size goes back to 0... your sample rate will be hitting large numbers... divide the sum of that by your sample size and you get a big number. Has very little to do with filling up at a gas station with a warm engine. A cold engine cruising down a long hill after hitting the reset button will produce big numbers as well.

If you don't believe this, try not resetting your gauge after a fuel stop. Your mpg will track your average accordingly. You will not see an increase in your short trip back home.

The larger the sample size, the harder it is to shift your mpg one way or the other.

 

i totally agree and know just what you mean. my gauge has been sitting at 60.9 for most of this tank, even though i have some decent stretches where it the display lights go into the 80 range (coasting, or running with a draft), but the actual numbers don't change.......

 

livvie, you must be an engineer. Or maybe a statistician. But your explanation is very clear.

The best I've done is 54 mpg on my work-home route. It's my 3rd tank, so I have a lot to learn. It seems like my tanks start out very poor and get better the longer I drive. All my gas stations must be at the bottom of hills.

Cheers,
Kathie

 

I'm not a statistician... but an engineer of sorts.

I actually don't know the sample rate for the HCH or the max number of the sample size.

The larger the sample size the better but at some point it can work against you. Let's assume the sample size for the HCH has no limit.

Let's say that you drove 100,000 miles and never hit the reset button. Let's also say that your car is sampling 100 times per mile. This translates to 10 million samples (100000miles samples X 100 samples/mile) . Let's say you managed to drive the entire 100,000 mile at 30 mpg. Meaning each time the computer took a sample, it was at 30 mpg.

So your average is 30 mpg... from a sample size of 10 million.

If all of sudden you started driving at 1000 mpg (incredible numbers). For the next 100 miles (that's only 10 thousand new samples out of your 10 million already accounted for)... your average would still show 30mpg on your reading. Which is not reflective of your current mpg reading.

So you have to find a happy medium. So maybe a maximum of 5000 - 10000 samples total. And a sample rate of 10-50 per mile.

With the smaller sample size in the above scenario... your 1000mpg numbers will be factored in much quicker and your reading would immediately start to increase from the 30 average.

So essentially your oldest samples get dropped off (once you hit the MAX) as new samples come in.

Confused?

 

Your explainations about are quite good. I think of it as a FIFO register.

But I wonder why they use sampling at all. Would it not be more accurate/economical to measure the volume of gasoline consumed and the distance travelled? Assuming they could accurately measure the volume of gasoline consumed, the processing and memory requirements would be much less.

How do they calculate the instantaneous fuel consumption? There must be some type of flow meter or weight sensor in the vehicle. It would be nice to learn more about the system.

Cheers,
Kathie