Fuel tank sending unit

[Villeman]

Hello all,

 

I am a bit puzzled. Did anybody ever replace the capacitor thingy on a 76´ 280z fuel tank sending unit? Is there any replacement part? My whole sending unit it pristine, just the darn capacitor is completely rusted (guess it was in old petrol too long...) and there is virtually NO junkyard with parts here. Can I salvage an ebay unit for the capacitor?

 

See image stolen from Hybridz for reference, need the small metal tube, everything else is in perfect condition

Edited January 9, 2018 by Villeman

[wheee!]

That is not a capacitor, it is a low fuel sensor....

 

soak your sensor in CLR rust remover overnight, it will work fine again.

Edited January 9, 2018 by wheee!

[Villeman]

yes, but i guess it works like a capacitor with fuel as dielectric filling. Sorry, physicist (does anybody know who it works, I am just guessing). Problem is, it crumbled into nothingness at first touch. I honestly only have dust left. So, where to get replacements?

Edited January 9, 2018 by Villeman
spelling

[wheee!]

No fuel, no conductivity = low gas. Zero ohms (short) when fuel present, open when empty.

These fuel units are very pricey and hard to find. I restored mine because I couldn't afford/find new. Good luck!

[Villeman]

I would have guessed it´s an Impedance measurement, but maybe I am overthinking... (If I figure out how it works I can just build my own maybe).

BUT: If you are sure this is how they work, then one from another car will work the same. So I could buy a cheap ebay one and salvage the part.

 

Edited January 9, 2018 by Villeman

[Captain Obvious]

That sensor isn't a capacitor, It's an RTD. They push the current from the fuel warning lamp through it and it heats up. If it heats up enough, the resistance drops and the current gets high enough that it makes the filament on the warning lamp glow.

The trick is... When it's immersed in fuel, it won't heat up enough for that to happen. But when the fuel level is low enough and that sensor is hanging in vapor (and not immersed in liquid fuel), it heats up enough to pass higher current. High enough to light the lamp on the dash.

@Dave WM did some analysis and came up with the above.

[Villeman]

Perfect, thanks. On the other hand, thats not a concept I want to improvise

I guess nobody has the threshold values present, so back to looking for a replacement it is....

On the other hand, I have everything to build one....hmmm Wonder if its a standard PT100. Somebody have the resistance values?

Edited January 9, 2018 by Villeman

[Captain Obvious]

There is no "threshold". The current is always being pulled through the bulb filament. It's just that when the sensor RTD is cool enough, the current is so low that it won't make the bulb filament glow. The current will gradually go up as the temperature of the tank sensor increases, and at some point, you'll notice it on the bulb because the filament is glowing bright enough to see it.

In other words, it will glow dimly before it glows brightly. And it happens slowly.

It's an analog function, not step change.

[Captain Obvious]

Are you sure yours isn't working? Just because it's rusty, doesn't mean it has failed electrically.

[Villeman]

makes sense, so its parallel to the bulb and acts as a short circuit until warm enough? Meaning without it my bulb will be always on, guess I can live with that ... Improvising will be hard, if I am lucky its a standard PT100 but i guess it isn´t

 

EDIT: it is LITERALLY dust now, yes, I am sure

Edited January 9, 2018 by Villeman

[Dave WM]

in series with the bulb, cold (in liquid gas, current flows, but heat generated  is dissipated by liquid) resistance is high, out of liquid the current flowing creates heat, as it heats up the resistance of the sender drops, allowing more current to flow, in series with the bulb. as the resistance continues to drop the light begins to glow.

Always seemed odd to me to have something like this in a gas tank when it gets low enough to be out of liquid gas and if gas fumes.

 

the FSM details a test process, but it is somewhat flawed in that it has you check it with a multimeter in the resistance setting. with so low test current there is very little heat generated so the change in value is not significant when in liquid (water for the test) or air. A better test would be to hook up a 12v batter and light bulb of the type used in the indicator. In series, this would flow enough current to make the test valid.

 

Edited January 9, 2018 by Dave WM

[Villeman]

yes, that was my problem as well, but apparently this really is how they work..... now just the small task of a) finding one  or finding a replacement

 

I would have liked it to be a RTD better, Platinum wire and go, but for this to work you need a semiconductor (since in metal, resistance goes UP with the heat, so it getting warmer would reduce current flow effectively making the light go dark if in series with the bulb

Edited January 9, 2018 by Villeman