HELP! Fuel sending unit is not making any sense.

[cleandan]

New fuel sending unit with verified ohm reading between 8-73.
Verified working fuel gauge on the dash.
Verified working wiring between the gauge and the sending unit.

With sending unit out of the tank, hanging on the plug, and manually operated I can get it to read on the dash gauge; empty where empty should be, half tank where half tank should be, 3/4 tank where 3/4 tank should be, but not reach full. I can sweep smoothly between empty and 3/4 tank on the gauge as well.

I can get Ohm readings consistent with the level of the tank float and where that should be on the gauge.

I can achieve the readings with the sending unit touching the tank and hanging free from the tank.

I have verified the float is buoyant and moving freely in liquid by submersing it in a bucket of water in the position it would be in while in the tank.
I am using a poly float, not the brass float.

But, when I place the sending unit in the tank, even if just placed there, it will only read below empty on the gauge and there is about 1/3 a tank full of gas.

I bend the float arm to lower the float, effectively making it read higher on the gauge....still below empty on the gauge.

I have both visually looked inside the tank, as well as put my hand in there, to verify there is nothing in the way of the float travel through its full range of motion.

Verified dash gauge.
Verified float is floating.
Verified sending unit is sending within required gauge parameters (according to Ford literature)
Verified wiring is good.
Literally operated float by hand to verify gauge is moving with sending unit out of tank.
Bent float arm to set float lower in tank initially in an attempt to "find" 1/3 tank level on gauge with sender installed.

But the effen thing won't work once placed in the tank...just placed, not even locked with the lock ring (although it has been tested that way too)

What have you got for ideas with this one?


 

 

[WConley]

Bad ground is my guess.  Touch your voltmeter probes between the positive wire to the sender and the fuel tank.  I bet the reading is sketchy / intermittent.

[FrozenMerc]

Yep, defiantly a ground problem.  Either it is shorting directly to ground, or is open on the ground side. 

[cleandan]

The ground circuit checks good.
Using my Fluke multi-meter I get continuity from the ground leg of the two prong plug.
1) Ground post on sending unit to any metal of the sending unit checks good.
2) Ground leg on the wiring plug to the truck chassis, metal straps holding the tank in place, to the truck body, checks good.
3) Ground from battery to engine block checks good.
4) Ground from engine block to chassis checks good.
5) Ground from engine block to firewall/body checks good.

No errant voltage sending troubling signals that I can measure.

6) Pulse voltage to positive leg of the sending unit plug checks good.

I wondered if maybe I have too much rubber isolation between the tank, chassis, and straps so I installed two different jumpers from the metal tank itself to the chassis and found no differences is readings or signal to the gauge.

I'm beginning to wonder if the gauge itself has been physically damaged (bent needle, messed up spring, etc.) causing it to literally work but not read properly or accurately.

There are no grounding issues in this electrical system....one benefit of this gauge search has been to clean up and verify all grounds (which are often the cause of most electrical issues in vehicles, so I understand the suggestion)

[Thumperbird]

Fastest and most effective thing to do in situations like this might be to pull the gauge and the sending unit and verify it works as a system on the bench, then piece it back in with your own external wire harness so to speak if it works fine on the bench, go from there.  Even if you have a good ground if you have some other parasitic current path in the system it will throw things off.

[FrozenMerc]

Can you use a standard resistor of say 35 or 40 ohms to bypass the sender and see if the gauge reacts appropriately?  Re-reading your original post definitely makes it sound like a ground issue, but using a known resistor may help point you in the right direction.

[HarleyJack17]

You sure there is not a resistor down stream you are bypassing in your testing? Seems I remember old Fords did things like this.....if bench testing all parts I could see that being an issue.

[RustyCrankshaft]

My Galaxie does something similar but in the opposite direction. The gauge works and it does read, but it's biased significantly. Tested the sender and it was good, wiring is good. Replaced the sender anyway - no change. Mine stays on full far too long, 1/4" tank is nearly empty. Only thing I haven't checked yet is the CVR. I know its working, but I don't know if maybe the set point has drifted. That would possibly explain the bias. I also tried ground jumpers, etc. Everything individually is working fine, last 2 things are either the CVR has an issue or the gauge itself is the problem.

[Diogenes]

I'm not sure whether this will be of any assistance, but below is a link to a company that makes a product I purchased for my fuel gauge/sender. The gauge was never correct, but this took care of it. I've had it in my Galaxie for about 10 years--works great. Check it out, it may be what you're looking for....

https://www.technoversions.com/MeterMatch.html

[FrozenMerc]

Rusty,
  Your problem is the resistance range on the fuel sender doesn't match the resistance range of the gauge.  I have ran in to this problem before, and the easiest way to solve it is to get a potentiometer and wire it inline between the gauge and sender.  Then adjust the potentiometer until the gauge is reading correctly and finally disconnect the potentiometer and measure the resistance across it.  Replace the potentiometer with a resistor of the same value that you just measured.  This only works if the senders resistance range is lower than the gauges.  If it is the other way around, you can use the potentiometer to figure out the gauges range and order the correct sender.