.
especially if you're correct about the computer manipulating the needle to agree with the computed remaining range.

I also resest the trip odometer, but since my mileage does vary from 15-22 mpg, now I've got to estimate the remaining range based on a trip odometer, speed, headwinds, number of cold starts, ambient temp, etc. ... OK, I can do this, but I'll be conservative, and I thought that's what that silicone chip was to calculate for me.

Actually, I normally drive with the computer display turned off. All I need/want is just an accurate fuel guage.

 

Damn. I have the EXACT same problem, going in this week as well.

 

Good luck; but for better or worse, let us know the outcome.

 

Re: Guys .... this is a step backwards, Agree.

 

I realize "Your mileage may vary", but I know that around town I can travel 600-700 kms before I need to fill up. This enables me to glance at the trip meter (even if the car is off) to determine when I next need to get fuel.

I don't drive around with the computer display on either - until the low fuel warning pops up. At that point I turn on the DTE setting just to keep track of where I am. This is when I began to notice this strange behaviour wrt the the trip computer calculations and the fuel gauge needle.

I was concerned enough myself initially to think that I needed to get this "problem" checked out and fixed at the dealer. However, based on my observations, I now think this is a "feature" - not a problem.

Lauren

 

.
"This is not a bug, it's a feature."

 

Re: Do you work for Microsoft? - Good one :-)

 

I don't think that distance-to-empty readings move the fuel gauge needle. If they did, driving out of and into traffic jams, or down I-90 to the Columbia river then back up again, would move the needle up and down. They don't.

Here's my guess, for which I have no evidence, why our fuel gauges sometimes read lower after restarting:

1. The A8/S8 fuel tank has two lower chambers on the left and right connected by a common upper part -- like an upside-down U. Each chamber has its own fuel level sensor, but the sensors are wired in series, so the instrument panel sees only one reading.

2. The instrument panel probably responds slowly to changes in the fuel level signal, to prevent the fuel gauge from responding to cornering, braking or accelerating. It probably averages the last few seconds' or minutes' readings to move the needle.

3. But (here the guessing starts) the instrument panel doesn't remember the past readings when you start the car, so it shows the instantaneous reading.

4. And (continuing the guessing) maybe the two fuel level sensors give an inaccurately low reading when most of the fuel is in one chamber instead of distributed evenly between the two. If so, a sharp turn into the parking lot might redistribute the fuel such that the reading upon startup is too low.

5. The return to accuracy after starting and driving may then be the result of fuel being redistributed more evenly between the tank's chambers, possibly by the complicated fuel pump system (described below).

These guesses can explain why the gauge can read lower upon restarting when there's not much gas in the tank, which seems to be what we see. Has anyone seen the gauge read lower upon restarting when that tank's more than half full (and the car's not on a hill)?

About the fuel pump system -- or frightening things one sees in the service manual:
There is one main two-stage electric fuel pump with a built-in fuel reservoir. The first stage pumps fuel from the pump's reservoir to three hydraulically operated suction pumps. The suction pumps, driven by the fuel from the main pump's first stage, gather fuel from various parts of the tank and feed it back to the main pump's reservoir. The main pump's second stage delivers fuel from the pump's reservoir to the fuel injection system. The fuel filler directly refills the main pump's reservoir, so that the system can restart after running out of gas (not that you should) and being refilled. Yee-haw! Will this all work after 300,000 miles?

 

.
Once this occurs and the gauge is reading zero, it stays there until refilled.

On two different occasions, I did try adding only two gallons of fuel just to see what would happen. Sure enough, the gauge immediately went back to reading the correct level.

I've never seen the gauge read low when restarting with more than about 1/4 tank.

Regarding the complications resulting from the two fuel sensors and the inverted U-shaped tank -- I don't make any sharp turns into the parking lot that would redistribute the fuel in the tanks. Also, I finshed the last session of the last track event with maybe 1/8th of a tank. The car certainly saw high cornering g's, but when I parked and then restarted the car an hour later, the fuel gauge did correctly show 1/8th. [and about 20 miles to empty; I averaged 8 mpg on the track.]

But ... now that you mention it, I've only had this problem occur when parking at home, and I do climb a very steep hill immediately before parking the car. Is this a clue?

 

I haven't seen the needle go back up from 0 either. But LEHIII saw it go from 1/12 to 2/12 while driving home. I wonder if this took longer than either of us has waited, or if Lauren has different instrument panel software, or if 0 is treated differently.

I also wonder if turning the key off and on a few minutes after restarting, without adding fuel, would be enough to restore the correct reading.

Do you drive slowly or idle for a few minutes before turning off the car after track sessions?

Tom