possible fuel gage improvement?
 

I believe that cars of yore used a wire wound rheostat inside the fuel tank as the variable resistor for supplying a signal to the fuel gage. This would give a rather coarse resolution but certainly work for the on-dash gage.

I am wondering if this is still what is used for the fuel gage 'sending unit' or if something with much better ability to finely resolve the level is used, such as a regular carbon pot.

Just wondering if it would be worth the trouble to tap into the fuel gage's signal wire and feed it to a voltmeter, which can resolve to hundredths of a volt, for the purpose of being able to know very accurately your fuel level, such as when parked on your level garage floor.

I dont think it would be worth it. There are just too many variables that give a less than perfect reading.

For starters, the shape of most gas tanks is not a rectangular box, but an odd shape designed to make the best use of space in the floorpan. That right there would make the reading be inaccurate at different points on the sender, even if the sender was perfect.

Then there is the level of vehicle, and whether you are accelerating or decelerating usually makes a difference in the reading even with the same amount of fuel in the tank. Also there is the internal baffle around the sender, and Ive noticed that the fuel will read higher if Ive stopped and restarted during a long drive, than if I just drove it straight. Stops and restarts will refill the baffle and give higher readings for awhile, then when the fuel really runs low the gauge drops much faster.

Then there is the geometry of the float arm itself. Assuming all else was perfect, the only way to get a true reading would be at the point when the float arm is horizontal. As it goes higher or lower, some of the float swing is horizontal motion, so the float (and gauge) moves more for the same drop in fuel level. The point where the gauge moves the slowest would be the point that the float arm is attached to the sender. It might be different for all cars.

Some newer cars might have better float systems anyway. I am mainly familiar with my old GMs.

I hope to add "fuel remaining" to the mpguino at some point (mpg gauge), that is why tank size is one of the parameters.

I was not expecting for it to read a linear value, i.e. n Volts per Gallon. I figured you could start with an empty tank, take your reading in volts then add a half gallon, let it settle, take a reading, and so on. This would give you a table of volts vs. gallons. It would not make any difference what the shape of the tank is or the geometry of the float arm.

I am just trying to determine, accurately, gallons used over short distances rather than waiting two weeks till I fill the tank.

My anticipated use of this would not be for while driving, but rather for when the vehicle is sitting in the garage, still and level.

My only question is whether or not the sending unit of current day cars has the infinite resolution required as opposed to the old days where I believe they used a wire wound rheostat which obviously has very rough and poor resolution.

Can you describe the intended use a bit more? Is it running in the garage or sitting still, yet somehow needing a refill every two weeks?

If you get the mpguino dialed in for your vehicle it will give you instant, as well as how much it thinks you used for an arbitrary amount of time. I don't know where the other fuel is going in your situation.

You could also weigh the vehicle I suppose.

Also maybe you can tee into the fuel line and add a sight gauge:
http://opengauge.org/cycle/fuelgauge.JPG

I guess I have not made myself very clear. I am only trying to figure out how much fuel is in my tank, say every morning before I take off driving, or every night when I get home. The dash board gas gage already does this except it is a very crude method of telling 'how much' fuel is in the tank. I want to measure my remaining fuel very accurately.

According to the electrical diagram of my vehicle (Chilton's manual) The fuel gage sending unit in the tank is nothing more than a variable resistor. However, I am trying to learn if that variable resistor, in newer cars, is a wire wound pot (poor resolution) or if it uses some other type of variable resistor which might have very high resolution.

The sight gauge should work except I don't want to go to that much effort. It would not be very easily doable on my Sonoma pickup.

I want to measure my gas tank within a fraction of a gallon.

I could just find the wire coming from my sending unit and discover for myself if the variable resistor has the resolution I need but I thought I would ask on this forum first. Maybe someone already knows and can tell me, since it would be a bit of work finding the wire, tapping into it, and then doing the required procedure to determine the info I am seeking.

Kinda crazy but why not use some kind of pressure sensor to measure the weight of the tank? As said earlier most tanks are odd ball in shape or internally baffled around the pump.

But in answer to your question, if your sending unit is a Bosch, its a coarse pot, 5v reference signal. Resistance dependent on fuel level. Resistance is inversely proportional to fuel level. I don't believe most cars use a fine resolution sensor, just something that would give them +/- 0.5 gallon resolution.

Thanks! That is just what I wanted to know but unfortunately not what I hoped for the answer to be ;-)

Strain guages mounted on the straps, or maybe just on the bottom of the tank. The tank bottom or support straps will deflect more with added gas. You will only measure weight, not volume. You will have to remove the gas cap to de-pressureize the tank before you measure. Naturally using gauges in all 4 legs of the bridge will help negate temperature effects.
Strain Gages, Accessories and Instrumentation

Hello Gregte, Yes you can read how much fuel is in the tank with the a multimeter. I used to use one in my old Sunbird when my fuel gauge stopped working, I would measure the resistance in ohms, I'm not sure now what the readings were now but if you tap into the wires you will soon find out your full mark and when its empty. Mine was easy to get at I just kept a multimeter in the trunk, cause the sending unit was accessible. hope I helped.:thumbup:

Yes, I already know you can do this. My question is concerning the resolution of the pot used in the sending unit. In a couple posts above Red has indicated that current day cars are similar to older cars in that regard and therefore the low res component is not suitable for my purpose.

Maybe you could go to the parts store and ask to see some from newer cars. You might be able to tell by looking at them. I know on mine you can clearly see the conductive lines on the sender. Maybe some newer ones might have a solid resistive sheet.

I know that Ford tend to use segmented senders as part of the fuel delivery module that have something like 50 segments to read from the top of the tank to *near* the bottom. The readout is dependant on which segments the float places the sender (as there are I believe 2 or 3 contacts on the wiper arm that sends the signal), so the output is an average voltage across the segments. It also works on a voltage, not resistance input. The instrument cluster control module then interprets this voltage and applies logic filtering to damp out changes in level when cornering (slow fill constant). If the ignition is turned off and then on again and it sees a different tank level between off and on, it then applies a fast fill constant and the guage goes directly to the new reading (as it determines that the tank has been filled up).

New systems are pretty complex!!

I just want a gas gauge that sits on the top line when full and the bottom line when empty.

Right now at full, it's somewhere around halfway through F and at really empty (but not totally) its somewhere around halfway through E. There are 2 ticks between top line and 3/4, 4 ticks between 3/4 and 1/2, 4 ticks between 1/2 and 1/4, and 2 ticks between 1/4 and bottom line.

How much is left in the tank? I can usually guess to within a 1/2 gallon, but I've gone over 60 miles after the "low fuel" light turns on. If I sit in the car with the power on but engine off, I can watch the fuel gauge slowly drop and slowly rise again when I start driving. You'd think it'd be easy to read on level ground with the car at a stand-still.

I once read that the automotive industry tried a gage that read full when full, not 1/8 inch over full on the meter, and of course started dropping right away as you consumed gas, instead of staying over the full mark for the first 50 miles. But, people didn't like it. They liked the 'feeling' of still having a full tank even after driving a good amount.

This can be seen everywhere you look, the idea of us people trying to fool our brains. It is a reality that we human beings like to pretend things are different than they really are, even though we know better. It does not stop at childhood.

I read the same thing once before. That car designers design the gas gauge to make you feel like you can drive forever before it comes off F, then when it gets low, they want it to make you feel lucky that you could make it to the gas station when it was on E.

I finally decided to bit the bullet and spend the time locating and tapping into my fuel gage wire and in the end it does indeed appear to have a wire wound pot for the sender. I put a jack under the vehicle to slowly tilt it to change the level in the fuel tank and thus the reading on my voltmeter and it was quite coarse and a bit irregular. So the question is answered for certain.
Thanks for all the suggestions!