I took my 2013 F150 in to the dealership to have some recall work done, specifically a transmission software update. (Modern cars, lol) As part of that they must have unhooked the battery or reset the system because all of my audio setting were gone.
Now the funny part is the gas gauge. When I dropped off the truck the gauge read just under 1/4 and the calculated DTE was at 149 miles. When I picked it up the gauge read just under 1/2 and DTE was just over 300mi. As I drove the couple miles back to work the DTE quickly dropped back to 149. I know they didnt put gas in it, but it seems like the DTE calculation was reset with the trans work and it quickly adjusted itself based on actual fuel use.
But here's the thing. The gas gauge moved too. That means that the gas gauge is driven off the calculation done by the computer, not driven directly off the signal of the fuel level sensor in the tank. Because level didnt change, just how far the truck thought you could go.
So I guess "fake" is a strong word, but it's not independent. And we know that the DTE calculations are often flawed. It's not a fuel gauge, it's a remaining mileage gauge.
Any of you run into this or have knowledge to confirm that's what's going on?
A lot of fuel gauges are pretty slow to update to avoid the effects of tank slosh. So if the vehicle is parked on a hill for a bit, the gauge will take a few minutes of driving to go back to normal. The hill will make it read high or low depending on how the vehicle was tilted relative to the sensor's location in the tank. It's a pretty significant effect in ZJs, for example. I've seen mine show easily 1/8 tank or more differently after being parked on a hill.
that's disturbing, but not unexpected. One less thing to break or need adjustment
I'm familiar with the hill effect, but this wasnt that. Dealership lot is dead flat.
I'm not sure it's ever functionally a problem, but it bothers me that I dont have an actual indicator of how much fuel is in the tank, which is what this seems like. Maybe one of the FoMoCo insiders here can tell me otherwise.
Also, the evening after having the work done part of my truck fell off in the street. It was just the fiberboard 'skid plate' but it makes me wonder what else they failed to do correctly.
The sending unit information goes to the ecu, which washes all the signals and sends everything to the instrument cluster via the bus. Clusters these days don't have a whole lotta stuff going on in there.
I think it's a real fuel fill gage. But many sensors lose reference when memory is lost- so it could have taken a little while for the sensor to figure out where 0 was and give an accurate value.
As for the display- I don't know what drives it these days (probably the body module with input from the engine module). And I'm not all that satisfied with the fuel economy and range calculations as a customer as well.
Fuel sending units in Fords still have a float that feeds info to a gauge, at least they did 10+ years ago when I was designing them. But it's all electronic, as Eric says, the sensor feeds it's input into the ECU, which decides how to interpret it.
ultraclyde said:
But here's the thing. The gas gauge moved too. That means that the gas gauge is driven off the calculation done by the computer, not driven directly off the signal of the fuel level sensor in the tank. Because level didnt change, just how far the truck thought you could go.
You have identified two things that changed, but haven't shown anything causal between them. It could be that DTE drives fuel, it could be that fuel drives DTE, or it could be that both are driven off some invisible third thing.
Fuel level senders change their outputs all the time -- if you just drive a gauge directly off of it then you'll see it go from 3/4 full to 1/4 full driving around a corner, then back to 3/4 when you're going straight. The signal needs to be damped, historically this was done with some analog electronics, but in modern cars it's done in software. The computer reads the signal from the tank and decides where to put the needle (or how many LEDs to turn on in some cases), and the damping happens by averaging out the signal that it's seeing now with the signal that it's been seeing over the last few minutes. This is the same computer that drives a "distance to empty" value by looking at the milage you've been doing and multiplying that by the damped estimate of fuel remaining, so if that computer has bad data for the last-5-minutes historical data on the level sender, then it will produce garbage output.
So the most likely hypothesis is probably that something the dealer did confused the computer about what the level signal had been for the last 5 minutes it was running. Parking it on an incline could be one answer, but another possibility is the software update. There's probably special code in the computer to cover the "I've just been turned on and don't know what happened before" case, but I could easily believe that if they've reflashed it then minor informational systems like fuel gauge calculations might be confused for a few minutes.
I would just summarize and simplify down to...
"accurate math was lost due to power loss (battery disconnect)"
Eventually, accurate math was regained after new calculations were made.
When I would drive the CTS-V on track mileage was something like 5mpg. The fuel monitor would quickly tell me that I had almost no range left and the gas guage would fall. As I drove to the hotel getting about 25mpg my range would go from something like 32 to 70 and the needle on the fuel gauge would rise. Happened more than once.
I have a similar truck to you and have had the same thing happen, a few weeks ago it read empty when o got in the truck which I never leave it low so I swung past a station and it came back up as I was pulling in. I thought someone had stolen fuel out of it since I keep pretty close track in my mileage but this was different.
Ford does what they used to call an "anti slosh circuit." It used to be very problematic, but with modern electronics they're fine.
What it does is checks the float level every second or so and averages the values to display on the gauge. That's how they prevent the gauge from responding directly to the float level and wildly dancing around with sloshing fuel. GM just uses slow gauges with friction to prevent dancing needles.
Try it sometime. Fill the tank with the key on/engine off sometime. You'll fill the tank but the gauge will say 1/2 or 3/4. Then 5 miles down the road it reads full as it continues to average the data points it reads from the float.
So in effect, when you disconnected the battery, you deleted the data points and all it had was null data. It assumed full. Then it pinged the float and the float said 1/4. So it averaged those two and displayed 5/8. Next time it pinged and saw 1/4 again and avereaged 1/4 and 5/8 and got.... you get the idea.
Curtis said:
Try it sometime. Fill the tank with the key on/engine off sometime. You'll fill the tank but the gauge will say 1/2 or 3/4. Then 5 miles down the road it reads full as it continues to average the data points it reads from the float.
First time i did this (was about -5* with a heavy wind so wanted to sit in the warm truck while the fuel filled) I got a few miles down the road and about turned around to call out the gas station for charging me for gas that never got pumped, then a few more miles and the gauges started slowly moving up to full. Was in a 03 f150.
I seem to remember this is something like how modern systems work:
The injected fuel quantity is used to monitor how much fuel is being consumed. It uses this to learn the properties of the tank float as you drive (where the endpoints are, nonlinearity due to tank shape, slosh, etc). A combination of the ECU's fuel consumption metric and the tank float position are used to estimate the remaining fuel quantity in the tank. This quantity is used to determine the DTE. The estimated fuel quantity then maps to a gauge position, which isn't necessarily linear (the gauge shows empty-ish long before you run out of fuel, right?).
While this system seems quite complicated for something so simple as fuel level (which it is), it allows for automatic compensation of things like a dented fuel tank, bent fuel arm, degraded potentiometer on the fuel arm, tank slosh, etc.
