Some posts recently are querying about incorrectly reporting fuel guages; i.e. always E, always F, backwards --E when tank is full, moving to F when tank is empty, etc. For folks with these problems, you might find this little tutorial useful. For the most part I’m familiar w/Toyotas, but most cars work appx like this.
The fuel level is measured with a float mechanism which rises and falls with the amount of fuel in the tank. What’s a “float mechanism” you say? Ok, that’s a fair question. Lift the top tank lid and look inside the workings of most any toilet. You’ll see a big ball shaped thing that floats on top of the water. Now flush the toilet. Watch the float fall as the water level falls. That contraption, the float and what it connects to, is what I mean by “float mechanism”. There’s something very similar to it inside your fuel tank.
The float mechanism inside the fuel tank has one important difference compared to your toilet though. Your car’s version has a variable resistor connected to it. The car’s electronics simply measures the resistance of the pot, converts it to fuel level (% full), and displays it on the gas gauge. That’s the way it works. The pot’s resistance is low (15 ohms) when the tank is full, and high (107 ohms) when the tank is empty. Recent models do the process in two steps: (1) a micro-computer (cpu) measures the resistance; and (2) the cpu computes the fuel level from the resistance measured, and outputs the result to the fuel guage for display. The conversion from ohms to % full is done in software. On earlier cars (circa 1990’s cars and earlier say), the fuel guage is pre-calibrated. It is directly connected to the pot. No cpu is used on the earlier types.
So basically all what you have to deal with when debugging a incorrect fuel gauge is understanding there’s a float mechanism in the tank, a gauge on the dashboard, and wires connecting between them. What are the three most likely things that can go wrong? (1) The float can fail. It simply may refuse to float; (2) The pot can fail. When it fails, it doesn’t measure the correct resistance. One common failure mode for a pot is it intermitantly measures much higher resistance than it should. And (3) the wiring is being jostled around when you hit potholes, etc, and can eventually chaffe, the wires break, or wires short out to chassis ground, or to another wire, and the connectors in the wiring corrode and fail to make good contact.
Ok, that’s the end of my tutorial. For an example, here’s a link to a pdf on debugging the gauges on 2004 Toyota Corollas. Pages 18 of the pdf is an overview of the wiring diagram for the fuel gauge. You can see the pot there on the left hand side (a pot is diagrammed as a wiggly line with a straight line with an arrow through it), connected between terminals 2 and 3 of the fuel gauge sending unit (the float mechanism). Brown and yellow wires connect the float mechanism in the tank via connectors ID2 and C9 to the cpu (called the combination meter unit) in the dashboard. Page 6 shows the cpu diagram. You have likely already observed that this 2004 Corolla model uses the modern two step software method, rather than the direct connection method.
If you actually wanted to start the debug process, the first thing you’d do is figure out where the connectors going to the float mechanism and the guages are. The wiring diagram for your vehicle would help you locate these components. Then you’d find the yellow and brown wires connecting in-between. The next thing you’d likely do is measure the resistance at the float. The pdf offers some debugging suggestions on how to do this.