Cluster analog fuel gauge - accurate or buffered?

[g300d]

After finally hopefully solving why the 4.6 is eating left fuel floats senders (replaced the charcoal filter), I am checking to see if the fuel gauge works properly.

Filled the tank and now in the middle of a long and spirited drive.

OBC test 6 now shows the tank has 57.5 liters left. Checking online the fuel tank capacity is 93 liters, which comes out to about 62% fuel left.

Analog gauge is just a bit under the 3/4 line.

OBC test 6 shows both senders giving data and fluctuating normally, also at Phase 1.

Also did the OBC test 21 reset and gauge pointer position did not change.

So I'm wondering if the analog gauge is buffered or there is still something wrong with the fuel measuring system.

[andrewwynn]

Test six does show both buffered and unbuffered values.

The laws of physics prevent accurate fuel reading at the top end of the scale: the floats hit the top of the tank about ten liters short of full.

I'm not sure how the computer guesses how much fuel is above 86 L but the needle and the digital values never go above 86L when the tank holds 96L. This is exactly why the needle is pegged at full for the first 100km or so the needle won't move.

[Qsilver7]

...and yes, the gauge is buffered which is why after replacing a fuel level sensor...you're supposed to reset them by doing TEST 21 (see info below...TEST 21 mentions the software's "damping function").

[80stech]

Originally Posted by andrewwynn

Test six does show both buffered and unbuffered values.

The laws of physics prevent accurate fuel reading at the top end of the scale: the floats hit the top of the tank about ten liters short of full.

I'm not sure how the computer guesses how much fuel is above 86 L but the needle and the digital values never go above 86L when the tank holds 96L. This is exactly why the needle is pegged at full for the first 100km or so the needle won't move.

I suspect the buffering of the fuel sloshing, previous history and measuring the fuel that is being used is what tells the computer how much is above the level that the sender float can move.

[andrewwynn]

But it's completely guess work. The computer has to assume the tank was filled not 5L short. it will make an adjustment once the floats are floating I just find it interesting for the first 100km or so there is no "sloshing" as far as the fuel gauge is concerned. The floats are pinned to the top of the tank. You can observe this with test six it will show no change I forget if it's 82L or 86L but there is no wiggle in the number for about 10L of consumption.


Sent from my iPhone using Tapatalk Pro

[wpoll]

It's 82L

My diesel gauge is marked in litres, although the top level is marked "Max" rather than "93L" or "82L".

And it very accurate. If it's saying 40L and I fill up, the car takes 53L. Bang on.

[andrewwynn]

40 + 53 is 93 at least in my hemisphere :-)

Pull up the hidden menu when full and note that it reports about 10L shy of the total. I'm curious how close the number is on the low side: if you drive to zero DTE, what quantity to fill?




Sent from my iPhone using Tapatalk Pro

[wpoll]

Originally Posted by andrewwynn

40 + 53 is 93 at least in my hemisphere :-)

Pull up the hidden menu when full and note that it reports about 10L shy of the total. I'm curious how close the number is on the low side: if you drive to zero DTE, what quantity to fill?




Sent from my iPhone using Tapatalk Pro

Exactly - the tank really does hold 93 litres - I've checked it.

And yes, the menu reads as 82L - and doesn't drop for the first 11 litres of consumption. And neither does the gauge.

[80stech]

The computer knows exactly how much fuel is being used from "injector on time" and knows exactly how fast you are moving, so with a little averaging from previous history it can do a pretty good job at giving you distance to empty even without the sender moving and then average that in as the rest of the tank is used.

[andrewwynn]

Distance to empty requires knowing the amount of fuel. Something the computer can not know until there is less than 82L of fuel.

I think the program assumes a full tank and uses math to estimate burn rate and subtracts from 93L until it can get a reading at 82L where it will make a quick correction.

Say for example you fill to 87L, the car will start with 93L in it's calculation and get 490 miles range and right before the cross over from 82L it would show 466 because it thinks there is 88L but as soon as the threshold is crossed and it measures 82L the math is corrected and you get 434 miles DTE.

It's just a curiosity of mine. It must be something like this. I can't think of any other possibility.

I don't think I've had the DTE up exactly when I crossed under 82L remaining but I have seen pretty rapid swings in the DTE plenty of times. Usually on the low side of the tank when going from City to highway driving but also when on a full highway drive.