School me: fuel gauge

Includes seating, upholstery, door trims, dashboard and heater.

Moderators: reidy, Blacky

Post Reply
User avatar
Harv
Posts: 5020
Joined: Mon Oct 08, 2007 2:00 pm
State: NSW
Location: Sydney, Australia

School me: fuel gauge

Post by Harv »

The FB/EK fuel gauges have a long history of being unreliable. For a while I tolerated it, but it's probably going to catch me out shortly. Three different vehicles, similar problem.

Grace's ute has a bog stock fuel tank. When it is full, the needle comes up nicely. When driving, the needle moves a little as the fuel sloshes. The rotten thing runs out of fuel when the gauge is slightly under half full (ask me how I know). Thought it may have been because it sat for many years before we got it, but nearly a year of running around has not freed anything up. I checked the wiring and cleaned terminals, and gauge goes full scale/zero when 12V is applied/removed. Thought it might be the float, so replaced the sender with a new Rares one.., didn't help. Gauge goes full scale/zero when the float arm is moved manually whilst outside the tank.

The wagon has a stainless aftermarket 60L tank, and aftermarket sender (both supplied by MC Dean in Queensland). When it is full, the needle comes up nicely. When driving, the needle moves a little as the fuel sloshes. With a partially empty tank, the gauge will often start by showing full, then move back to about where it should be (e.g. half full) during a journey. It then comes back up to full again a few minutes later. The rotten thing runs out of fuel when the gauge is slightly under half full, just like the ute (ask me how I know about that one too... at least it rolled into the servo). Thought it may have been because it sat for years whilst I built it, but 6 months of running around has not freed anything up. I checked the wiring and cleaned terminals, and gauge goes full scale/zero when 12V is applied/removed.

The FB originally had a nicely working gauge - read full when full, and empty when the needle was slightly past the left hand gauge dot. In the last year, it has begun behaving like the wagon. AFAIK, it is the original gauge and sender.

Appreciate views please on what the next steps are (other than bolting on a jerry can holder to all three cars).

Cheers,
Harv
327 Chev EK wagon, original EK ute for Number 1 Daughter, an FB sedan meth monster project and a BB/MD grey motored FED.
ardiesse
Posts: 1074
Joined: Thu Jan 19, 2017 9:57 am
State: NSW
Location: Sydney

Re: School me: fuel gauge

Post by ardiesse »

Harv,

You've done most of the troubleshooting yourself already with Grace's ute: the fuel gauge goes from E to F as you move the float arm. How did you arrange the earth connection when the sender was outside the fuel tank?

I think the gauge reads F when power is applied with an open circuit at the sender terminal. When the sender terminal is connected to ground, the gauge will read E. (And fear not, there isn't a direct connection from the battery to the sender terminal). The magic number, I think, is 30 ohms. With 30 ohms to ground at the sender terminal, the gauge will read F, but not go off-scale.

So, if my hunch is right, and both cars run out of fuel when the gauge reads half, there's about 15 ohms of unwanted resistance in series with the sender, most likely on the earth return. This is why I asked about the earth connection when you tested the system . . . Bad electrical contacts will be the likely culprit.

Get a multimeter and measure the voltage between the battery (-) post and the car body when the headlights are on high beam, engine not running. Anything more than 0.1 V needs to be corrected: big fat wire from car body to battery (-) clamp.

Then measure the resistance between the fuel gauge body and the battery (-) terminal. Subtract from this the figure you get when the multimeter leads are shorted together. (It obviously helps if your multimeter has a resolution of 0.1 ohms.) If the instrument panel is connected solidly to ground, then proceed.

Turn ignition on, and measure voltage at the fuel tank between sender terminal and the body of the sender. Now get a couple of jumper leads: you want to connect the multimeter (-) lead to the battery (-) terminal without the jumper leads touching the car body. Measure the voltage between the sender terminal and the battery (-). If you get a different number, you've found the problem. Put the multimeter's (+) probe on the sender body and measure the voltage to confirm. More than 0.1 V, again, indicates trouble: high-resistance contact between sender and tank, or tank and body, or both.

AC fuel gauges sometimes go "noisy" (wavering needle when the car's not moving); but it's best to check all aspects of the wiring first.

Rob
User avatar
59wagon
Posts: 645
Joined: Sun Jul 27, 2014 11:40 am
State: WA
Location: Shoalwater, WA

Re: School me: fuel gauge

Post by 59wagon »

Hey Harv,

I think Rob's checks would pin point the issue in most cases as, in my experience, the majority of electrical issues in these older cars seem to come down to an earthing fault, and with fuel gauge issues, the actual fuel gauge is rarely the problem.

With Grace's ute, you say it works fine when you move the lever with the sender removed. I assume you've connected the sender body to the car body somehow. If that's the case, and you have the same response from a new Rares sender unit, then I would think that either the sender body isn't earthing properly when installed (cruddy screws, or paint/muck between screws and body) or, you're not using a sedan sender unit rather than a ute/wagon one are you? The sedan unit has a longer cork arm as the sedan tank is deeper, so would read above empty when the ute tank is empty.

With the FB, the fact that is was working a year ago, may suggest that the sender unit is starting to fail internally - have you checked it out of the car like you did with the ute sender? You can check the variable resistance with a multimeter (0-30 ohms), or check with the arm at around 0, 25, 50, 75 and 100% and see how the gauge responds. Do this 3 or 4 times to see if it's consistent - corrosion inside the unit can give inconsistent readings. You could check the actual gauge using Grace's new Rares unit if you're doubtful about the gauge reading correctly.

And the wagon, well, it's an EK for a start :) The "wandering" you describe suggests a faulty sender unit or dodgy earth connection. I'm not familiar with the Marty Dean sender units but if they're the same basic design as the FB original, then the checks would be the same as for the FB.

Have fun mate and hope you get them sorted. Let us know how you go.

Cheers,

John
User avatar
Harv
Posts: 5020
Joined: Mon Oct 08, 2007 2:00 pm
State: NSW
Location: Sydney, Australia

Re: School me: fuel gauge

Post by Harv »

Thanks gents. I think I have my head around how they work. The diagram below I have butchered from the Workshop Manual:

Fuel gauge.jpg
Fuel gauge.jpg (556.89 KiB) Viewed 628 times

With the ignition switch on, power flows from the battery through the ignition switch (the path shown by the red arrows). Power then flows into the fuel gauge, and through the coil on the left. The coil makes a magnetic field, trying to pull the needle to the left.

Some of the power then follows the green path, flowing through the coil on the right. This coil makes a magnetic field too, trying to pull the needle to the right. The power flows out of the dash (through the two dash screws), through the body sheetmetal, through the fuel pump braided hose, through the engine, then through the battery negative lead to the battery.

Some of the power though does not follow the green path. Instead, it follows the purple path. It leaves the fuel gauge without flowing through the coil on the right. It flows to the back of the car, and into the fuel tank sender unit. It passes through a variable resistor in the sender unit, then flows through the fuel tank mounting screws, through the body sheetmetal, through the fuel pump braided hose, through the engine, then through the battery negative lead to the battery.

When the fuel tank level is full, the float arm rises and the variable resistor has lots of resistance (30Ω). This makes the purple path very hard to travel down. Being lazy, the power decides to mostly flow down the easier (lower resistance) green path. This puts lots of power through the coil on the right, and the needle is pulled hard to the right of the gauge. When the fuel tank is empty, the float arm falls and the variable resistor has very little resistance. This makes the purple path very easy to travel down, so the power decides to flow more the purple path. This puts less power through the coil on the right, and the needle starts to move to the left of the gauge.

All of this relies on a nice, clean, low-resistance path through the body sheetmetal, through the fuel pump braided hose, through the engine, then through the battery negative lead to the battery. In an ideal world, this would have no resistance at all (0Ω). There is always a bit of resistance though – copper wiring, crappy terminals, corrosion, loose screws, paint under terminals etc. The first check that Rob asked me to do was to measure just how much resistance this path had by measuring the voltage shown as “A”.

Fuel gauge measurements.jpg
Fuel gauge measurements.jpg (602.07 KiB) Viewed 628 times

This checks if there is any obstacle for power to flow from the sheetmetal back to the battery. If there is no resistance, “A” would be zero. When I measured this on the FB, I got 0.2V (it should be less than 0.1V). This is partially because I have no fuel pump braided hose, and rely on some pretty crappy paths to get power back from the sheetmetal to the engine and hence to battery (worn engine mounts, carb linkage, choke cable, radiator, rear spring u-bolts etc). Putting a strap from the battery negative to the body sheetmetal would help this.

To check just how crappy that flow path is, the second measurement Rob asked for was resistance “B”. It’s the same test, just covers more of the path. This would ideally be 0Ω.When I measured this on the FB, I got 0.4Ω. Is that a good value or not? I suspect it is a pretty good value. To check it, I ran a heavy (40A) wire from the fuel sender all they way to the battery (bright shiny copper, new terminals), and measured the resistance. It gave me 0.4Ω. This makes me think that my flow path of “the fuel tank mounting screws, through the body sheetmetal, through the fuel pump braided hose, through the engine, then through the battery negative lead to the battery” is pretty good at 0.4Ω.

What happens if this was a crappy flow path? It would make the purple path very hard to travel down all the time. Even when the tank is empty, the power would want to flow down the green path, and hence the gauge reads like it has fuel in it. If the resistance gets up to 15Ω, the gauge would think that the sender was at half way (half of 30Ω) and would show half full when the tank was really dead-empty.

The further checks that Rob suggested are measuring voltages C and D. If there is no resistance, then these are the same (and for the FB they were… 1.7V). A final check that I did was to pull out a fuel tank sender screw, clean it up, then run the lead from the fuel tank sender body direct to battery negative (perfect return path). The gauge didn’t move an inch.

So for the FB, it looks like the problem wasn’t in the earth return path. I figure I’ll test it the hard way… 5L jerrycan of fuel in the boot, run it dry, and see what it says.

(Rob: as an aside, the noise from my kitchen was apparently home-made pizza. I’ve made less mess pulling the crank out of a grey motor).

Cheers,
Harv
327 Chev EK wagon, original EK ute for Number 1 Daughter, an FB sedan meth monster project and a BB/MD grey motored FED.
belly
Posts: 4
Joined: Thu Sep 11, 2014 6:27 pm
State: QLD
Location: Brisbane

Re: School me: fuel gauge

Post by belly »

I've checked many fuel gauges and never found a faulty one. Some are slightly out of calibration. Most problems are either poor earth or faulty sender units. They are simply worn out with age. They consist of resistive wire wound into a slide wire. The float arm drives a contact which slides up and down the slidewire. This is what varies the resistance between 0-30 ohms.
If you're confident with a multimeter, disconnect the power to the sender unit, remove the sender and you can use the ohms range on the meter across the terminal and earth (the case) and move the float slowly up and down. See if the resistance varies and if it varies smoothly without jumping around. If it jumps or doesn't vary between 0-30 0hms it is faulty.
User avatar
Harv
Posts: 5020
Joined: Mon Oct 08, 2007 2:00 pm
State: NSW
Location: Sydney, Australia

Re: School me: fuel gauge

Post by Harv »

Testing still ongoing in the FB. Gauge now seems to be behaving itself - ran down to 1/8th of a tank by the gauge (jerrycan in boot), still going.

Finally laid my hands back on Grace's ute.
Voltage A is 0.1 to 0.2V (from battery to body) - similar to the FB.
Resistance B was 0 ohm (off the end of my Fluke meter) - good news.
Voltage C and D were both 10.4V.
Doesn't look like the sender is the issue. Vehicle returned back onto the road for a few more laps, will have another go once it returns. There are days when I feel like a pitcrew chief... cars go out for a few laps, come back in with something else busted :lol:

Did manage to spotto a broken temperature sender on the ute whilst fiddling. Brass lug had snapped off the top, wire dangling. Glad to have found and fixed that one.

Cheers,
Harv
327 Chev EK wagon, original EK ute for Number 1 Daughter, an FB sedan meth monster project and a BB/MD grey motored FED.
Post Reply