I have a 1986 F-150 4x4 with a 5.0 FI engine. I needed to replace the in tank fuel pump/sending unit and tank. This is a 19 gallon single tank side mounted setup. Yes this is one of the 2 fuel pump systems. I bought a new fuel pump unit and tank from an online parts supplier. The tank is a perfect fit and the new fuel pump works fine...
The problem is the sending unit. First of all the new one is wired the opposite of the old one. With the float all the way down one the new unit, the fuel gas shows full. When the float is about 1/2 way up or any position further up, it reads empty. I spoke to the supplier about this and they insist this is the right unit. I called the local Ford dealer and their parts person verified that it was the right one for the truck. I took resistance readings from both the old unit and the new unit. Old unit: 11 ohms up, 90 ohms down. New unit: 170 ohms up, 20 ohms down. With the float about mid way the new unit is about 93 ohms. That would explain why the reads empty with the float only at the midpoint. I can swap my variable reading connection point to the opposite side of the rheostat to get the readings going the right direction. The supplier also sent me a replacement pump in case the first one was bad but I still have the exact same problem.
I even spoke to Tech Support at Federal Mogul who makes the Carter pump and they were clueless as to what the problem was.
Any ideas that might help me keep from pulling the rest of hair out overt this?

ChuckT

 

Sounds to me like Federal Mogul has built the rheostat backwards. At low resistance (11-20 ohms) the guage should read full (float at the top of the tank), as you've mentioned, and as the resistance increases (float decends towards the tank bottom) the guage should move toward empty. At least that's the way it works on new fuel pump and guage sending unit I just installed in my '87 F-150 I-6 300.

I've concluded that the reason for the 11-20 ohm reading when the float is all the way up (aka, a full tank) is to keep from "pegging" the dash guage real hard if the value were zero (a dead short to ground when the tank is full).

If you can't convince the supplier/Federal Mogul the unit is made backwards, here's a possible option. If you have two tanks with a switch on the dash to flip between the two, it could be a relatively simple fix. If you only have one tank, however, this fix will still work, but you'll have to figure out a way to get into the wire going to the fuel gauge (perhaps at the back of the guage itself). Here's the fix:

1. Go ahead and swap the variable reading connection point to the opposite side of the rheostat to get the readings going the right direction, as you said you can do.

2. One side of the rheostat is grounded to the supporting member inside the tank, and there's a single wire running to the tank connector (at least that's how it's wired on my unit). There's two wires running to the pump itself, and each of those run to two pins on the connector. The connector itself has four male pins, and the fourth one is either unused or serves as a ground (mine is unused, and the ground is actually through the mounting of the tank).

What you can do is connect a 170-ohm 1/4 or 1/2 watt resistor between the single wire from the rheostat to ground. DON'T do this inside the tank, because the resistor probably won't hold up to the gasoline. If you have a single fuel tank, you could do this where the wire from the rheostat attaches to the fuel guage in the dash. If you have two tanks and a switch on the dash, do it at the back of this switch. The switch is a "double pole, double throw" type, that is, it will have six connections on the back of it - two columns of three connections each. The two center connections are the "commons" that are switched between the tanks. One column of connections provides the 12-volts to the pump in the respective tank (front or back) and the other column switches the single wire that goes to the guage. You'd need to put the 170ohm to 250 ohm resistor between the guage switched side and a good ground. If you connect a voltmeter (set to read 12v) between the "common" in one column of connections and ground, then turn on your key, you should measure 12v for about 2-3 seconds (before the ECM turns off the fuel pump relay). It'll be the other column of connections you'll need to deal with, then, since that's the one connecting the fuel level sensor in the tank with the dash guage. Run a resistance test between the appropriate female pin of the connection at the tank and each of the switched pins on the tank selector switch connector - whichever one shows zero resistance (or very close to it) is the one you want. Connect one end of the 170-ohm resistor to that wire and the other end to a good ground. Be mindful of vibrations while driving, so connect it in such a way that the resistor leads don't eventually break in two.

Here's why it will work: Two resistors of identical ohm rating, when hooked in parallel, will give 1/2 the resistance of the value of one resistor. Since the rheostat resistance at EMPTY is 170 ohms, connecting another 170-ohm resistor in parallel will reduce the overall resistance to 85 ohms. This is close to the 90 ohms of your old unit, and should be close enough. You could always plug the unit into the tank connector, and have a buddy watch the guage while you operate the rheostat to verify it works OK before you do the final installation.

When the new rheostat is in the FULL position, you state it measures 20 ohms. Since the new 170-ohm resistor will be permanently attached to the circuit, the final resistance value will be what's known as "the reciprical of the sum of the recipricals", or:

1 divided by the sum of 1/20 + 1/170 or 1 divided by 0.05 + 0.0059, which equals
1 divided by 0.0559, or 17.89 ohms. This should be close enough, since your old
unit measured 11 ohms in the FULL position (the guage just might not be
"fully pegged beyond FULL", is all.

If you do a "pre-final installation check" with a buddy watching the guage as you operate the rheostat and you find the guage does NOT go all the way to EMPTY at maximum resistance of the rheostat, a 200-ohm resistor would give you:

1 divided by the sum of 1/170 + 1/200 = 1/0.01.9 or 92 ohms at EMPTY, and
1 divided by the sum of 1/20 + 1/200 = 1/1.055 or 18.2 ohms at FULL

A 250 ohm resistor would give you:

1 divided by the sum of 1/170 + 1/250 = 1/0.00988 or 101 ohms at EMPTY, and
1 divided by the sum of 1/20 + 1/250 = 1/0.0504 = 19.8 ohms at FULL

You can buy single resistors at Radio Shack, or any electonics parts house, for about 79 cents apiece. Sometimes, however, they can't give you exactly the resistance you want because not every value is manufactured (plus there's a + - 10% tolerance, usually). But they should have something in the range between 170 ohms and 250 ohms. Resistors available today will have three bands of color around the body (with the first band closer to one end), signifying the resistance value. For the ones I've figured here, the color bands would be:

170 ohms: Band #1: Brown Band #2: Violet Band #3: Brown
200 ohms: Band #1: Red Band #2: Black Band #3: Brown
250 ohms: Band #1: Red Band #2: Green Band #3: Brown

Good luck! Please post back as to how this solution works, if you decide to use it.

 

HiTraderjoe28
That was one of the things that I tried.I aligator clipped a 180 ohm resistor across those 2 points and my new readings were 100 ohms up and 19 ohms down which would put me in the right range. Unfortunately it did not change the way the dash gauge worked. It was still reading empty at the 1/2 way of float travel. I also ran a wire right from the negative side of the battery to the pump to ensure that I had a good ground. This was all done with the pump out of the tank. The old sending unit works fine this way so I am assuming that the gas itself doesn't affect the reading. I would just swap out sending units but the old one is a different style and it is also in sad shape. I don't think I would count on it holding up too long.

 

First, you're right in assuming that the gas itself doesn't affect the reading.

Something is really screwball here. You say the readings from the old unit were 11 and 90 ohms at the extreme positions, and the resistor-modified extreme readings on the new unit are 19 and 100. Close enough so that the float at the halfway position shouldn't be showing EMPTY on the gauge.

Exactly where did you alligator-clip the 180 ohm resistor, ie, across what 2 points? You say "this was all done with the pump out of the tank" and you "ran a wire right from the battery to the pump to ensure you had a good ground". The wire to the pump may not have provided a ground to the added resistor, if you just alligator-clipped between the rheostat wire and the mounting metal for the whole assembly while it was out of the tank because the pump itself is not grounded to the mounting plate - that's why it has 2 wires going to it. You can verify this lack of pump ground with a resistance check from each of the 2 wires individually and the mounting plate.

You need a good CHASSIS ground on the resistor end opposite the connection to the gauge wire connection, since pump itelf is not grounded to the mounting plate when the whole unit is out of the tank. Try connecting the grounded end of the resistor to the wire you ran from the battery negative terminal. You may have measured 100 ohms at the EMPTY position with your meter, but the gauge itself is NOT seeing it, for whatever reason.

If you short the tank connector pin for the level indicator wire to a chassis ground (with the key on), the gauge should slam over to read FULL (zero ohms). Disconnected (infinite ohms), it should show EMPTY. Around 60 ohms it should show roughly half full, assuming all of your resistance measurements have been correct. The float-operated resistance rheostat is roughly linear across it's operating range.
.