[rwhitaker]

I'm looking for someone to talk me out of my suspicion that my gauge is bad. I sure hope I'm wrong.

The problem: Gauge is on E but I know I have almost a half tank of gas.

Here's what I've done so far and the results:

Disconnected tan wire coming from gauge to sending unit . Gauge goes to F.

Grounded the sending wire to the frame. Gauge goes to E.

Removed power/sending connection from back of gauge. Tan wire read 38 ohms of resistance -- reasonable as 45 would have indicated half full. Reconnected gauge.

Jumped a ground direct from battery to back of gauge housing as well as left post. No movement in gauge.

Removed sending/pickup assembly from tank. Tested 90 ohms at full, 0 ohms at empty.

What have I missed?

Any help would be appreciated.

 

[rwhitaker]

Update:

My fuel gauge is not the problem. I removed the (brand new stainless) sending/pickup unit from the tank and hooked it up near the dash. When connected and grounded the gauge followed the movement of the float.

I then put the sending unit back in the tank. Instead of connecting it to the sending wire (which has consistently resulted in an Empty status though I have almost a half tank of gas). This time I ran a jumper wire from the stud on the unit to the sending connection under the dash -- which had proven to work properly above. The gauge read Empty, reflecting a fully grounded circuit.

I'm stumped. What would cause a new sending unit to work perfectly at the dash, then ground out when installed in the tank? The only difference I see is the presence of gasoline.

The only thing I know to do now is buy a new sending unit and hope it solves the problem.

Has anyone ever run into this before?

 

[Jim K]

Greetings RW,
I haven't seen a reply post to your issue but maybe I can help.
If you have already solved your issue then you can ignore the rest of this..

First off, do you have a wiring diagram for your vehicle? Electrical issues can be a pain so one can never have enough info..
You can find one at Pontiac - Le Mans - GTO & Tempest - ClassicCarWiring
or do a PDF down load here.
1967 Pontiac Service Manual
Assuming your 67 is the car with the issue. The link above has diagrams for several models of Pontiac's, so you'll just have to search around to find it for the tempest which should be identical to your GTO.

The fuel tank sending unit has to be grounded, but that ground has to be through the resistor. If it grounds out before that then the gauge see a direct ground which sends it to empty. such as the drawing below (taken from the first gen site). I would test your sending unit, with an ohm meter between the connection for the tan wire and the body of the sending unit.

Hope this helps..
Jim K.

 

[rwhitaker]

Hi Jim,

Thanks for the links and the info. I'll get good use out of those!

I think you're right. Today I jumped a wire from the stud on the sending unit to the connection under the dash -- twice. Once with the sending unit grounded through the ground screw (result: Empty reading, as usual). And again with the ground screw out (result: open circuit and Full reading).

That tells me the ground is coming through the strap, not somewhere else like the outside of the unit. And I think it's telling me the unit is not limiting (through resistance) the ground to the gauge which is why I get a constant Empty reading when it is correctly wired. So I think I need a (second) new sending unit, which I ordered today.

But one thing I still don't get. The sending unit works perfectly connected under the dash, but does not work in the tank (using same jumper wires for each test). Oh well, not the first mystery I've had on this 53 year old classic.

Fingers crossed the second new unit works!

 

[rwhitaker]

btw, Jim K, Great looking '71!!

 

[Jetzster]

Most likely the float is not floating? If it doesn’t rise With liquid around it then you’d have a constant ground from the unit correct? Perhaps it’s defective that way, if you can set just the float part in a sink and see if rises ,
did your meter have a small or near zero ohm reading before you add the gas back and after you add gas stay the same? You should be seeing it go up as you add each 5 gal so Back, it’s either not floating or it’s internal strap is shorted out , gas should not affect its brushwinding...I’ve had good luck with Dorman senders if ya got to try another..

 

[Jim K]

Let us know if the new sender works RW!.. By the way.. I really like that color on your 67.. Beautiful condition too!

 

[rwhitaker]

Update:


The new sending unit behaved exactly like the other new one -- works great at the dash, doesn't work in the tank.

Jim K, I followed your recommendation and measured ohms at the stud -- 32. Ohms on the face of the sending unit were also 32. That doesn't seem right. But the insulators look good on both units. (The two units are from different manufacturers.)

Volts coming from the sending wire from the gauge - 8 volts pretty consistently when the ignition is on.

From testing, I know the ground screw to the body is providing a good ground.

Also, as I said, I was able to make both units work correctly at the dash by connecting a jumper from the gauge sending wire to the sender stud, and grounding the unit. I was able to recreate the false E result at the dash ONLY by attaching the jumper to the body of the sending unit. Doesn't that support that the same reading (32 ohms) at the stud and the body are at the heart of the problem? But that would suggest the same defect came from two new sending units from different manufacturers. Very unlikely!

So it appears the two new sending units work fine until they are installed in the gas tank. What would cause the gauge to read E when the sending unit is installed in a tank that is half full?

Any help would be much appreciated.

 

[Jim K]

Hey RW,
wow this one is a bit strange. Okay, the sending unit has been eliminated as the problem. The tanks is half full.. I am presuming the tank is not mounted up to the under side of the car since you've been installing and removing sending units? If it is laying loose, one thing you can try is to leave the sending unit out of the tank, connect the wire coming from the dash, connect all your grounds to get you back to battery negative and work the float manually while someone watches the gauge. Similar to your testing at the dash but using all of the vehicles wiring. This is going to sound like a dumb question, but you do have all your O-rings and insulators installed at the mounting of the sending unit correct?

 

[rwhitaker]

Thanks for responding so quickly, Jim K.

The tank is strapped under the car. I removed it, cleaned it, and re installed it before replacing the sending unit. The '67 sender hole is easily accessible from under the car on the side of the tank. Since the car was on jack stands during testing, I removed them and lower the car to normal stance to make sure I wasn't getting some strange readings because of the downhill angle. Same E result.

I wondered about the o-ring too. But I have removed the sending unit and replaced it several times. Seems only one way to install the o-ring to create a seal -- in contact with the tank, then the sender flange, then the locking ring.

I did buy new insulation to go between the tank and the trunk pan. And I had to use some spacers under the straps to get the filler tube positioned properly in the license plate opening. I used cork gasket material (about 1/8" thickness) for the spacers between the tank and the straps. I realize this could also angle the tank slightly, but 8 gallons showing up as empty??

I also did the same under-dash test under the car with the sending unit uninstalled:
Jumped wire from sender stud direct to gauge sender wire at dash
Grounded the sender unit (two ways, first to another jumper direct to battery negative, then to ground screw -- both indicated good grounds)

Results:
Gauge worked perfectly moving directly with the float.

Installed the same sending unit it the tank, gauge reading E.

Now, for your recommendation, Jim K, I did the same test as above, except I used the car's tan sending wire coming out near the tank. Results: E reading at the gauge no matter how the float was raised or lowered. So there appears to be a fault in that wire. But the sender should work with a jumper wire replacing the tan wire.

Any thoughts?

I'm stumped for now.

 

[rwhitaker]

If the resistor connector on the sending unit was somehow coming in contact with the rim of the tank opening, would that create a constant E reading on the gauge? I don't see how that could be the problem and the locking ring still go in and lock, but I'm grasping at straws at this point. The stud is on the left side of the sender flange, not near the center.

 

[Jetzster]

Isolate the tan wire, ground one end of it,
Check with a meter for a solid ground from the other end, that wire usually goes thru one block connector Between the gauge and the sender, and that connector might be causing an issue ,

Also do you see + 8v coming from gauge tested Under the dash like you see at the sender wire tested under the car or is it different?

 

[Jim K]

RW,
If the sending unit shows Empty all the time with the tan wire hooked up regardless of how you work the float manually but yet it works fine replacing the tan wire with a jumper then the fault is with the tan wire coming to the sending unit. Electricity follows the least path of resistance to ground. If you have a "partial" short in the tan wire, it might still show 8 volts to ground through an OHM meter which has an extremely high internal resistance, but it's enough to jump to ground because the sending unit puts a resistance in the circuit. Not sure how the wire is routed on a 67 but I'd look for any points of contact along the wire. Worse case, you may have to get a rear harness and replace the one you have. Especially if that harness is original.

 

[rwhitaker]

Solved!

The float arm and frame of the new sending units were at a slightly different angle than my old unit. It was hardly noticeable but enough to position the float above the gasoline at less than a half tank. I bent in the float arm and the gauge is now registering 1/4 tank - less than actual but ok for now. Better than registering over.

I guess the moral here is, with aftermarket parts, just because it "fits" doesn't mean it will work properly. Adjustments and trial and error may be required for it to work like original part.

Now on to finding my short in my sending wire from the dash.

Thanks to all who took the time to offer suggestions. I really appreciate the comments!

 

[Jim K]

Glad you got that part of the problem solved RW! I've noticed myself that there are times when aftermarket parts don't quite work/fit like original. Tried to replace the rear bumper on my 71. Found out the angle of the bolt hole mounts were very different than original. Ended up having to shim up the mounts where they bolt to the body to get the bumper to bolt up right.