Gas Gauge

[Althalus]

I dunno about anyone else, but on my first tank, the needle went slow on the first half tank (go about 35mpg) but as soon as it past the half tank mark, it started going down really fast, I ended up with only like 29mpg for the whole tank.

[Ever Victorious]

It has been said several times before on this forum "the gas gauge can move at any time for just about any reason". You can't ACCURATELY determine fuel economy off of the gas gauge.

As a ROUGH estimate, if you are getting 100 miles per 1/4 tank tick, you are getting approximately 30 mpg (+/- about 5-10%)

[BenDupre]

You can't measure mileage by looking at the gauge. It's not linear.

I don't know why engineers haven't solved this problem in the first 100 years of making cars. Your first half tank is more like your first two-thirds or three-fourths. You have to get used to knowing about owmany miles you get on a tank and track it usingyour trip meter.

Ben

[Althalus]

I don't know why engineers haven't solved this problem in the first 100 years of making cars.

Sheesh, no kidding.

[BenDupre]

Let's prognosticate:

In the year 2100, the fisometer will accurately predict the half-life of your fuel rods to plus/minus 6 months so your great-great-great grandchildren will know when the car is going to run out of gas.



Ben

[LA02MAX]

hmm..i guess my car is a miracle I usually average 100 miles per quarter tank, but on road trips i get 150, pretty much religiously....I must have a damn good fuel gauge!! I did notice the gas gauge in the versa decreasing rather rapidly after the first half on our road trip though, but we still managed to get about 180 miles out of the last "half"

[XterraVersa]

I don't know why engineers haven't solved this problem in the first 100 years of making cars.

Irregularly shaped tanks, sloshing fuel... It would take a total redesign of each fuel gauge for each model of tank, whereas the simple float gauge is a descent indicator of fuel level.

They could also tie into your cars sensors and measure fuel used & average mpg and report remaining fuel. When you fill up, you hit reset to set it back at full.

[OKVersa]

I have NO clue how it worked, being a girly girl, but we had a 1997 loaded Blazer that would give the MPG going down the road. We never once tried it to see if it was accurate, but it was fun to watch. We'd speed up and the mileage would go down; we'd slow down, and up it would go. It would tell how much gas was left in the tank, too.

[MinisterofDOOM]

My 1984 Coupe DeVille had a trip computer like that. It calculated fuel milage and both miles and gallons remaining til "empty". That car had features that Mercedes Benz is bragging about in it's new cars today.

My dad's 2004 Maxima does it as well. Shows DTE (distance to empty) and approximate fuel milage.

I'm surprised more cars don't have this feature. The Caddy's computer wasn't incredibly accurate, but it was good enough to provide a basis for planning fuel stops.Both the Maxima and Caddy computers are (were) dynamic to driving conditions. That's the biggest problem with trip computer accuracy: it can't predict whether you'll be doing 65, 85, or stop-and-go at 5 mph later down the road.

I agree about the gas gauge thing. I can't believe it hasn't been linearized by now.It actually wouldn't be hard at all to adjust them. Most gas gauges are a potentiometer or a multi position switch. The POTs would be a bit harder, but the multi-position switches would be VERY simple to correct for shape of the fuel tank. Simply increase the spacing between the switch positions a bit furter on one end to compensate.

My Cadillac's gas gauge was so touchy, it would lean opposite the car. It was oriented so that the halfway mark had the needle straight up. If, for instance, you had a half tank of gas, the needle would ALWAYS point straight up. The car had fairly floaty suspension, and if you went around a right turn, the car would roll left, and the fuel gauge would compensateby leaning right--and remaining straight up. Very touchy...and strange.

[BenDupre]

The car had fairly floaty suspension, and if you went around a right turn, the car would roll left, and the fuel gauge would compensateby leaning right--and remaining straight up. Very touchy...and strange.

That was your artificial horizon.

Ben

[Ever Victorious]

I have NO clue how it worked, being a girly girl, but we had a 1997 loaded Blazer that would give the MPG going down the road. We never once tried it to see if it was accurate, but it was fun to watch. We'd speed up and the mileage would go down; we'd slow down, and up it would go. It would tell how much gas was left in the tank, too.

MPG computers actually use relatively simple vacuum gauges, converted to a representation of mileage. They're shiny and flashy, but in general not terribly accurate.

When I was test driving cars before I bought my Versa, I considered an Escape Hybrid. It had a "miles remaining" display that would tell you how far you could go on the remaining fuel. I noticed from a short drive that it too was horribly inaccurate... the vehicle had 1/4 tank of fuel left, but was telling me I could go 300 more miles.

[fritzlovesmary]

we had a volvo xc90 and it showed instantaneous mpg, average mpg, average speed, miles to empty tank - and this thing was always pretty accurate. i often looked at the instantaneous mpg while driving to try to get better mpg - often just a very slight lift of the right foot helped. i wish the versa would show the instantaneous mpg (the new suzuki sx4 does), it is fun trying to get better mpg.fritz

[motoguy128]

Other than the first 1/4 tank being a little off, mine seems relatively accurate. The one on my '03 Corolla was spot on, and my '04 Accord was very accurate as well.

[rwanttaja]

I don't know why engineers haven't solved this problem in the first 100 years of making cars.

It's solvable, it's just more expensive.

The three components of a fuel measuring system are a fuel tank, a readout gauge, and a transducer...the unit that takes the the fuel level in the tank and converts it into an electrical signal the readout gauge can use. Typically, the transducer is a small float on the end of an arm connected to a potentiometer; a device like a volume control. As the gas level changes, the float rises and falls, causing the output of the potentiometer to change.

As it does this...it's moving in an arc. Since it moves in an arc, the output of the transducer is proportionate to the angle of the arm. But the fuel level..what you're trying to read... is the cosine of the angle, not the angle itself. The output of the transducer is indeed linear, but it's linear as to the *angle*...which isn't the same as a linear output to the fuel level.

Three ways the designers can handle this. First, they adjust the arm length and the total arc of movement to measure a fairly short section of arc...you pick the right section of arc, and the angle output is much more proportionate to the cosine. The trouble here is that you're now measuring over a fairly short range, which means any errors produce wide variations. The transducer as to have higher resolution, which makes it more expensive to manufacture.

Second, they can play with the circuitry so that the arm outputs the cosine, or the gauge itself does the conversion. But it's MUCH cheaper to make things linear. If everything is linear, you can use the same basic transducer on many fuel-tank shapes and use the same basic gauge movement on many cars.

Third, they can bite the bullet and calibrate the fuel gauge itself. You'd still have "Full" at the top and "Empty" at the bottom, but the intermediate marks would be offset up or down, depending on the calibrated output of the fuel-sensing system. The "1/2" mark, for instance, might be where you might normally expect the "3/4" mark, for example, with the rest of the markings similarly offset. I had a '70s VW Beetle and a "Thing" that had such a disproportionate gauge.

The other factor that kicks in is fuel tank shape. Fuel tanks are often a bit narrower at the bottom than they are at the top. This will throw off the gauges as well; make it look like the car is burning fuel faster as the tank gets near empty. In this case, only the third option above would help.

Yes, these days, car manufacturers could computerize the fuel gauge and just put in a "look up" calibration table to display a very accurate fuel level. But it's still much cheaper to use the same type of fuel transducer and fuel gauge that the '47 DeSoto had.

As a final note, I'll reference the Federal Aviation Administration's certification standard for small aircraft, including small commuter-like planes. 14 CFR 23.1337(1) reads, "Each fuel category indicator must be calibrated to read 'zero' during level flight when the quantity of fuel remaining in the tank is equal to the unusable fuel supply."

In other words, even most *aircraft* aren't required to have accurate fuel gauges! They're only required to read accurately when the tank is empty....the "1/2", "3/4", "Full" markings, etc., don't have to be accurate.

So don't feel so bad if your car's gauge is off. :-)

Ron

[Althalus]

I don't feel bad, it just annoying.

And I never noticed that when I was in flight school. I never let my tank get anywhere near empty though..

[BenDupre]

A modern accurate gauge would use a hall-effect non-contacting pickup and a a-d converter with digital output to the ECM which would decide how full the tank was based on a formula that takes tank-shape into account, and output a PWM signal to the gauge to display.

Of course all that's going to suck about $25-$50 worth of profit off the bottom line of the vehicle since market sets the price for this car. I'm sure in the QFD studies for this vehicle, accurate gas-gauge was never in the "house of quality." It wasn't even in on the "lawn of quality," if anywhaer it was way down the "block of quality" in the back alley behind the corner bar.

Ben

[civilone]

My WRX is similar as well. Before its needle goes below the two F lines, it would take a while, but once it goes below the middle, it goes like a falling apple.

[stevejonas1088]

my gauge was showing that i was just a hair above Empty. but when i filled up the tank, it only took 8.11 gal. i have a feeling this is going to get me into trouble later. it is my first tank of gas, so hopefully it improves. so driving around Las Vegas in 110+ degree heat, with the a/c always on max, i averaged 23 mpg... I am in awwww I previously owned a 89 Chevy K1500 4x4. and was estatic with anything over 10 mpg

[RetiredTexans]

I wouldn't be too concerned about it for the first tank or two. First use sometimes gives quircky results.

When we took our V home last year, it had a "full tank" from the dealer. When I filled up the first time, The mpg calculated at 52 mpg. I checked everything to see if I made some kind of mistake refueling or calculating. Following fuel usage was normal. The V averages about 31.5 mpg in urban driving.