CK5
Register an account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members.

Gas 'dissapearing' from 2004 Tundra....

TheBeast_88K5

Banned
Joined
Aug 30, 2010
Posts
12,274
Reaction score
18
Location
[Far away.]
Okay....

This is REALLY starting to piss me off!!!! :angry1:


I stop the truck with a 1/4 tank in it,
If I fire the truck up, and get it to operating temp.... I'm almost on 'E'....

wtf!? :dunno:

This happens ALL the time, now!
I noticed that the fuel pump seems to 'click', after shut off.
Could that have something to do with it?

Am I just insane?
Could the guage be wonky?

I'm at a loss here... :doah:


But $20 in gas dissappearing in about 3 minutes, is getting REALLY old....
 
Last edited:
locking gas cap?
or are you saying you watch it lose a 1/4 at a time?

at first, I thought it was gas theft,
but it's ONLY when it has 1/4 tank left.

Never any other amount.
3/4 tank, no problem.
1/2 tank, no problem.

1/4 tank-> 'E', in about 3 minutes.

I have yet to find a fuel leak, either.

Also, it's only on startup, with a 1/4 tank in it.
Possibly shutdown, if it's a fuel pump issue.

If I'm driving the gas goes normally.
 
This just a guage thing, the sender is a resistor and the higher the input voltage the higher the output.
With a full battery and the engine running, the guage would show a different number from when the engine is off or just started and the battery charge is still low, once the battery is up to full charge and the voltage is back up, the gauge will show less fuel.
When it's more full, it's no showing as much difference because all senders have more sensitivity at the end of the range.
at first, I thought it was gas theft,
but it's ONLY when it has 1/4 tank left.

Never any other amount.
3/4 tank, no problem.
1/2 tank, no problem.

1/4 tank-> 'E', in about 3 minutes.

I have yet to find a fuel leak, either.

Also, it's only on startup, with a 1/4 tank in it.
Possibly shutdown, if it's a fuel pump issue.

If I'm driving the gas goes normally.
 
This just a guage thing, the sender is a resistor and the higher the input voltage the higher the output.
With a full battery and the engine running, the guage would show a different number from when the engine is off or just started and the battery charge is still low, once the battery is up to full charge and the voltage is back up, the gauge will show less fuel.
When it's more full, it's no showing as much difference because all senders have more sensitivity at the end of the range.

So, if I'm reading this corectly:

I actually have less than a 1/4 tank;
But, the guage doesn't understand that fact until the truck is started?

I guess that kinda makes sense.... :dunno:
 
Nope, not exactly. Most gauges these days are not voltage sensitive.
They used to be.
Back in the days of the thermal gauges, you had to have a voltage regulator just for the gauges so they were accurate.

A lot of them ran on 6 volts so they would read correctly even with a weak battery.

Nowadays, its a ratio thing between two coils. One of the coils pulls the needle towards E, and its hooked from the battery voltage and ground.
The other coil pulls it toward F, and it is between the same voltage and the tank sending unit.
As the voltage changes, the amount of pull from both coils changes, but in the exact same amount since they are both fed by the same point.

The old gauges were moving against a spring, and the more voltage the more movement.

Your problem is hysteresis.

Your gauges can respond pretty much instantly when the sending unit moves. But that is a bad thing.
If your gauge was hooked to the sending unit directly, it would be swinging from full to empty all the time you were on a rough road as the gas sloshes back and forth in the tank.

So there is a small electronic unit in the system, usually inside the gauge cluster. Some have a small circuit board mounted to the back side of the cluster.

There are a couple of different ways of doing things in that system depending oh who made it.
First, it averages all the varying voltages that come in as you drive and slosh.

This takes care of most of the problem, but if you drive a few miles on a slope, or up a steep grade, the gauge will soon read wrong.

So, it adds in hysteresis, also known as dead-band. It will not pass along a change until it exceeds a certain amount.

Unless you are running a top fuel dragster, you will never suck fuel fast enough for any of this to matter.
The gradual use of gas will satisfy the averaging part, and you will not notice.
But, the dead-band is noticeable.

When you pull up and stop, the amount of gas in the tank is slightly lower than what is reading, but not enough for the unit to cause the needle to move.

When you kill power to the system, it resets the unit. When you turn it back on, it reads the sending unit and sets the needle accordingly.

Different companies do it different ways. I have seen some setups where the needle will gradually move down if you sit still with the key on for a few minutes.
Those only do the averaging and not the dead band.

But it looks bad when you see the gas gauge move down with the engine just idling, so I suspect that most companies don't do that anymore.
 
..I have a hysterical guage in the 'burb...:pimp:
 
Thought you would like that one.

The gauges they use now have some damping in them themselves. The shaft the needle is mounted on has some friction. If you move it with the power off, you can feel the resistance.
So it tends to ignore small changes.

But some cars in the past, such as Saab used actual D’Arsonval movements. The same jeweled movements in voltmeters.
They are very sensitive and move quickly. A friend of mine had a short in the panel of his father's old Saab, and we rewired it.
Whatever they used for damping got left out.
Just driving down a smooth highway, the gas gauge was all over the place.

When he came to a stop, it would slam back and forth between full and empty as the gas sloshed back and forth in the tank.

We were teenagers, and he was a banker, so we had no idea what was wrong. We finally rigged a switch to turn the gas gauge on when he wanted to read it.
It was just a normally open pushbutton.

Every so often, when he had been sitting still for a few minutes, he would press the button and read how much gas he had.
 
Is the fuel pump clicking or is that the float hitting the bottom? Strang gauge behaviour doesn't mean you're losing any gas. Check your fuel consumption. Do you have a trip computer that tells range and fuel used?

I don't know why we're talking about meter movements and damping, when cars have used stepper motors for gauges since the 90's. Yes, they look analog, but they are digital devices. The cluster receives either a voltage or a CAN message and then adjusts the needles. So the damping/hysteresis is easily handled in firmware. That being said, I have two such vehicles with good sending units that have wonky gas gauges. The trip computer gives good info, but the gauge just does it's own thing.
 
Not sure when they went to steppers, did not know they had. My '89 Ford, and my factory electrical wire diagnosis and diagram manual for 1990 GM C/Ks is still the two coil system.

Had to be after that.
But, the original problem still remains. Widely varying inputs to the system from the gas tank.
Whether the problem is dealt with by hardware, firmware, or even software, the results are the same.
A gauge that does not respond to a rapidly input, and will change when first powered up.
I saw a perfect example of it this morning. I was in my ''98 M-class, facing up a hill with the motor running.
I had between a half and a quarter of a tank of gas, and that was where the needle was reading.

I was helping my doctor friend get his Kubota tractor loaded up on his trailer.
I don't like to run much under a half tank, and glanced down at the gauge as I was getting out.
Then, I switched off the engine and went back to help him hook up the rope.

As soon as I got back in the car and cranked it up, the gas gauge read empty.
I had not lost any gas, the system had just reread the sending unit.
When I got through, and was on the way home, it was still reading a almost empty tanik.

I got out to the gate with the low fuel light on, and the needle down on E.

Switched off while opening the gate, and when I cranked back up, the gauges swung back to between half and empty, like before.
 
The gauge in my dodge sweeps fine (on E when empty and on F when I fill it till it clicks off) but it absolutely without a doubt drops faster once it hits the 1/3 of a tank mark. Just the way it is and I know it so I account for it.
I think that's all you're dealing with here...
 
My tundra is the opposite, it pulls off full almost right away, 5 miles or so, then E lasts forever, I can drive almost 40-50 miles on Empty. Mine is a 07 model. I rely on the mileage on every tank...I never put a partial tank in it, so I reset the mileage every time.
 
fordum said:
Your problem is hysteresis.

Your gauges can respond pretty much instantly when the sending unit moves. But that is a bad thing.
If your gauge was hooked to the sending unit directly, it would be swinging from full to empty all the time you were on a rough road as the gas sloshes back and forth in the tank.

So there is a small electronic unit in the system, usually inside the gauge cluster. Some have a small circuit board mounted to the back side of the cluster.

There are a couple of different ways of doing things in that system depending oh who made it.
First, it averages all the varying voltages that come in as you drive and slosh.

This takes care of most of the problem, but if you drive a few miles on a slope, or up a steep grade, the gauge will soon read wrong.

So, it adds in hysteresis, also known as dead-band. It will not pass along a change until it exceeds a certain amount.

This right here is reality, and it is never more apparent than when you have something like a Cadillac with no fuel gauge. The Caddy shows the number of gallons in the tank, and it is really common to turn it off when it shows say 5 gallons in the tank, and turn it back on a few minutes later and it reads 2-3. Regardless of how it is handled, shutting the vehicle off causes the system to reset and read the real 'current' fuel level.

My Toyota will work in both directions, when I am between 1/3rd and 3/4 full and I shut it off, it will always show higher when I turn it back on. However below 1/3rd and it drops like a stone and usually reads lower when I turn the key back on.
 
My tundra is the opposite, it pulls off full almost right away, 5 miles or so, then E lasts forever, I can drive almost 40-50 miles on Empty. Mine is a 07 model. I rely on the mileage on every tank...I never put a partial tank in it, so I reset the mileage every time.
With a full tank in our 2004, it stays on the "F" forever! :eek1:


Half, and below, it drops like a rock. :haha:

It's also a 4.7L-V8 :doah:
 
One more comment.
The damping and other things I have mentioned is true, and is the main cause for a different reading when you turn it off and on.

BUT, for the most part, it is NOT the cause for the needle to move at different rates depending on how much is in the tank.
The reason for that is in the sender and the shape of the tank.

The sending unit is basically a potentiometer. Same as the volume control on a radio.
At least the older radios.......
If you were to go to a electronics store and ask for a certain value potentiometer, hereafter to be known as a pot, because I am tired of typing the whole word, you would be asked what taper you wanted.
Pots come in different tapers. As you turn the shaft, or move the float in a gas tank, the resistance changes.
With a pot, you can have a linear taper, where it changes the same amount per degree of rotation from lock to lock, a log taper where the change is a log function, or an audio taper, similar to a log taper where the change is a function of the ear's audio response.

You would expect the sending unit to be linear. But that is hard to do, and costs money to make it accurate.
So, most sending units are "good enough".
But, there are other problems.
Think about how the arm moves on a sending unit.

When it is sticking out at right angles, a 1/4 inch drop in the level of the fuel, moves the float 1/4 inch.
Which moves the wiper on the resistor wire a certain distance. But, when it is hanging down at an angle, that same drop moves it a different distance, and moves the wiper a different amount.

So, if you want the output of the sending unit to be linear, then the resistor cannot be linear because the movement of the arm is not constant per amount of fuel movement.
Which makes designing the resistance wire even more interesting. If you look at the guts of a sending unit, the wraps of wire might be even all across the sweep of the wiper, or they may be closer together at different parts.

That is an attempt to make the output more correct.

But, even so, you would figure in these days of super computers, it would be a simple thing to design a wrap pattern, perhaps with different thicknesses of wire as you go across so as to make an accurate gauge.

And you could.
Except along comes the third problem.
Tank shape and size.
Check out the picture at the bottom. Its an extreme example, but as you can see, the level will change much faster towards the bottom per amount taken out than at the top.
Of course few gas tanks look like that, but they can have some strange shapes as car makers try to get more gas into less space.

They have all kinds of depressions, folds. Few of them are perfectly flat and square which would give a nice linear drop of fuel level per gallon used.

Add all of this together, and its a wonder we know how much gas is in our tanks at all.

Because, while mass production makes things cheaper, it also magnifies small amounts of cost.
So, while it would be quite possible to design a sending unit that will perfectly match the depth and shape of a gas tank, they would have to design and make different ones for every different tank out there.

You think getting the right bearing is hard? Imagine if there were different sending units for each model car and truck........

Even if you had two nice rectangular tanks which gave nice constant drops of fuel level, you would still have problems if they held different amounts.

The sending unit responds differently as the arm moves down, so if one tank was deeper than the other..........

So, does this mean you have to put up with a gauge that moves faster or slower as the tank runs out?
If it did so from the factory, you probably do. But, if it has gotten worse either gradually or suddenly, it might help to replace the sending unit.

They do wear out.
Or, if it reads normally until a certain point, when it just drops to empty even with gas left, then the last part of the resistor wire has broken and replacing it will fix the problem.

I know this is long, drawn out, and boring, but I was waiting on a phone call, and could not proceed on a project until I got it.

Got it now, so I'm out of here........

fuelta.jpg
 

Latest Posts

Top Bottom