Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

automotive fuel sender unit / fuel gauge

Status
Not open for further replies.

drage

New Member
Hi there,

I am reasonably good at electronics and working on cars, but i have a fault with my vauxhall corsa.

The fuel gauge is bouncing when driving along.

I have shorted two wires together, and this makes the gauge go right up to full, and then i disconnect this jumper wire the gauge falls to nothing.

Once connected the wiring must be supply a resistance to the fuel gauge as it moves the needle, it is only once driving that this fault occurs.

I have read on the internet this seems to be a reoccuring fault with this model of vehicle.

I have also read changing the sender unit doesn't always cure the fault.

So I am wondering is there a circuit i can make, or a circuit that already exists that would:
Take the resistance over a period of seconds / minutes, and then put out an average resistance rather than a constantly changing one?

This is something which i am guessing would have to be computed by a microcontroller or some such?

Has anyone got any advice?

Cheers
Paul
 
I thought it was being computed and averaged already by the engine management system for the onboard trip computer stuff...perhaps not in your case. Check all the connections, wiring et all from the sender to the dashboard at least, one of em will most likely prove to be dodgy.

Modern cars are very poorly wired in my opinion, it really is a case of where what's cheaper, is not necessarily better now. I think in the last 10 years or so virtually every electrical fault I have looked for in a car for a friend, has turned out to be either bad wiring, or dodgy connectors. When I say bad wiring, I mean actually broken internally, sometimes due to corrosion, sometimes due to not being properly dressed. Only a few times has it actually been a faulty transducer, sender or sensor at fault. While modular wiring is great in it's concept, the rough environment that a car presents is not the place where I would consider it a great idea, the manufacturers see it as a revenue gravy train for their dealers service departments...
 
yeah i will do the wiring checks but have a feeling the wiring is going to be ok, like i said when i put a wire between two pins on the sender connector it makes the dial gauge go to max, then when i take it off it drops, and i imagine if i put a resistor between the two it would read X amount of fuel..

i actually dont think this vehicle has a computer averaging the signal at all. it's a shame really.
 
The fuel tank itself must have a ground wire on it . It is a common problem for this connection to get corroded or loosened..
 
drage said:
i actually dont think this vehicle has a computer averaging the signal at all. it's a shame really.

I have a Corsa, and a Haynes manual.

According to the manual, there's just two wires from the sender, a ground wire and a single wire to the instrument panel. The instrument panel has the fuel gauge and a circuit to light the warning lamp.

I'd bet that you have a poor earth connection somewhere by the tank.
 
that being the case, i.e his corsa not having the fancy trinkets on it...then his fault could well lie within the dashboard panel as well as the body earth for the tank...
If it is an analog gauge on the panel, it is possible, I have seen it before, but not on a Vauxhall, that the winding for the meter movement had one end broken off...but still touching the wiring post on the meter. The winding terminations had no slack in them, and just being bumped about snapped it clean at the wiring post. So if the wiring shows up nothing, maybe a careful inspection of the actual meter movement, and perhaps the shunt resistor that goes with it might be in order...
 
cheers, never thought about the tanks earth, this sender has 4 wires, two for the low pressure fuel pump and two for fuel gauge but will investigate further..

ta
 
I have exactly the same problem.
I will check the earth to the tank first, but provided this is OK, would it not be feasible to connect a small electrolytic capacitor to the output of the sender to smooth any fluctuations in the output voltage to the dashboard?
 
Does the meter move somewhat slowly back and forth as if responding to the fuel sloshing, or is it a more erratic movement? That may tell you whether the movement is normal or the result of an intermittent connection.
 
Some gauges are fluid filled. This dampens any sloshing. Thus I suspect the gage. Also note if the gage moves a bit at standstill.

I've been at both ends of that circuit once.
 
On mine, the gauge swings high/low fairly slowly, as if responding to the fuel sloshing in the the tank. When the car has been stationary for a bit (and the fuel has stopped sloshing) the gauge gives a steady readout.

Presumably there should be either mechanical or electrical 'smoothing' somewhere when has stopped working (note taken about fluid damping in the dial itself) - I just wondered whether a capacitor across the sender output would be worth a go? I would just try it, but am a little nervous of soldering near the fuel tank!
 
On mine, the gauge swings high/low fairly slowly, as if responding to the fuel sloshing in the the tank. When the car has been stationary for a bit (and the fuel has stopped sloshing) the gauge gives a steady readout.

Presumably there should be either mechanical or electrical 'smoothing' somewhere when has stopped working (note taken about fluid damping in the dial itself) - I just wondered whether a capacitor across the sender output would be worth a go? I would just try it, but am a little nervous of soldering near the fuel tank!

The damping should be at the gauge end; not the at the sender. There is no need to even look the fuel tank. Because of the low sender resistance, it would take a HUGE capacitor to provide RC damping, maybe a ~1 FARAD 5V super capacitor.

Most modern gauges are stepping motors controlled by the ECU. If you have an analog gauge, it is usually a milli-ammeter, which has some damping of the moving armature caused by back emf. I have also seen gauges which are driven by a bi-metal heater, i.e. the needle is moved by the bi-metal heating being controlled by the sender resistance. In that case, the damping is provided by the thermal time constant.
 
Last edited:
You are looking for an "anti slosh" module. It masages the bouncing sender signal to keep the gauge from jumping around.
Many cars had them, two for sure are mid 80's Ford Taurus and mid 80's Jaguars.
 
I have also seen gauges which are driven by a bi-metal heater, i.e. the needle is moved by the bi-metal heating being controlled by the sender resistance. In that case, the damping is provided by the thermal time constant.
That type of gauge would see to have it's accuracy significantly affected by ambient temperature variation (unless it's operating temperature is way above ambient).
 
I have shorted two wires together, and this makes the gauge go right up to full, and then i disconnect this jumper wire the gauge falls to nothing.

Once connected the wiring must be supply a resistance to the fuel gauge as it moves the needle, it is only once driving that this fault occurs.

This is totally a guess based on a similar problem with a car not even close to what you have. The symptoms were the same though. Start the car sitting flat and the gauge was fine. As soon as you moved and started driving the gauge was all over between full and empty. Make a hard right and you had a full tank, make a hard left and you had an empty tank. Stop on a flat surface and eventually the gauge would read stable and accurate.

The fuel sending unit was a variable resistor. If I remember close to right a full tank was about 25 maybe 30 ohms and an empty tank was maybe 245 ohms give or take. Finally drained and dropped the tank (that was fun) and removed the sending unit. Worked the float and everything worked just fine. The problem was in the tank.

The float assembly that drove the variable resistor was supposed to be in a screened type baffle assembly which dampened the sloshing fuel for the sensor. Unfortunately it became loose and undone from its internal mounting in the tank. There was no dampening for the fuel and the gauge bounced when moving and bouncing along. Wasn't as bad on nice flat open roads but hell in the city. Tilting the tank and it became dialoged from wherever it was stuck in there and you could see it. Trip to junk yard and another tank the problem was solved.

You may want to place a variable resistor across the two sending wires, I have no clue, maybe 500 Ohms and vary the pot. If the gauge responds nice and smooth you have a problem with a baffle or something in the tank. In my case they did not use electrical/electronic dampening it was just a baffle in there. Anyway, something to consider or try. :)

Ron
 
Ron - I think that may be exactly the problem i'm having. Sometimes when going round a corner or over a bump there is a sound that seems to be coming from the fuel tank - I could imagine it being a plastic baffle floating around in the fuel, it certainly fits what i'm hearing. Pity it's so awkward to fix!

Mike ML - Thanks for the feedback re fitting a capacitor to the sender, I didn't realize the resistance of the sender was so low. I think that the gauges on the dash are actually controlled by stepper motors from their behavior, which rules out doing something similar on that end. Looks like i'll have to put up with it - or change the fuel tank, which seems like a lot of trouble for a fairly minor (albeit annoying) problem!
 
If you can snatch a pot then you can try as I suggested. Not saying I have this right but it sure is familiar. :)

Yeah, fuel tanks as to remove and replace are never friendly. :)

Ron
 
I agree with the baffle thing. Dropping the tank is not a fun job. If you do it make sure it's near empty, An extra person helps too The knock, I think, confirms it. I think it's nailed. Been there, done that.

BTW, in old "Ford T-birds" the sender was inside the trunk.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top