So in a attempt to do plug and play on my '56 Plodge two door wagon with Dakota front frame and fuel injection engine I used the Dakota dash gauge panel. The engine runs and the gauges seem to work except for the fuel gauge. Bit of the back story here, the Dakota tank was too wide but I found a Jeep plastic tank that fit just right and it used the same type of fuel pump as the Dakota. All wiring was moved over from the Dakota and extended where necessary.

I have a shop manual but I am so old school that if you can't fix it with a test light I'm hooped. If I ground the tank wire that goes to the gauge from the tank the gauge goes to full so does that mean that the gauge is OK? I have over a month into this and in attempt to make sure there was nothing in the tank interfering with the float, if I bench tested it in the back of the wagon OUT of the tank it worked fine, empty, half or full by manually moving the float arm on the sending unit. In frustration I ordered a new pump that came with the sender and the gauge still sits on empty. I don't know if the donor truck was like this, I just drove it into the shop. I checked the ground wires many times.

HELP !

Does the sending unit have a good ground when it is installed?

Thanks for the reply.
The pump body is plastic but there are two ground terminals in the push in block, one for the pump (which works) and one for the gauge.
I am going to run a temporary ground tomorrow to the gauge ground from a good ground source, just to check.

ttp://[/img][img][img]https://i.imgur.com/SQIxA0s.jpg[/img][/img]

Yes, When the wire from the gauge to the sending unit is grounded, and the gauge goes to full, the gauge is functioning correctly.

Next, the sending unit is a variable resister. The power comes in from the gauge (its one of the 4 wires in the connector for the fuel pump), travels down the tube to inside of the sending unit cover. Under the cover there is a piece of plastic that has wire wrapped around it, and there is a contact point beyond the winding that is the units ground wire. That wire runs back to the top and may be one of the 4 wires for the fuel pump plug, (or may be a separate wire).

The float arm has a button on it that rubs against the wrapped wire as it moves up and down with the float. Another button on the float arm makes contact with the ground wire. As the float arm moves up and down with the fuel level, the different contact point of the button on the wrapped wire changes the resistance the gauge senses.

For it all to work, the ground on the sending unit must be a good ground. I ran a separate ground wire just for the fuel gauge sending unit itself to a clean spot on the frame to use as a ground for the sending unit.

The next thing that has to work is the float arm and the button contacts. If the float arm can't swing freely over its complete range, it won't read correctly. You need to be sure the float, or the float arm isn't dragging against the side of the tank, or anything inside of the tank. After checking to be sure you have a good ground, the float free swing would be the next thing I would check considering your problem, especially since the sending unit has been moved to a different tank. Any float arm contact with anything inside the tank will cause the gauge to not work.

the last thing that has to happen (which don't believe is your problem) is that the one button on the float arm has to be in contact with the wire wrapped around the plastic, and the other button has to be in contact with the ground button on the sending unit.

None of the modern day sending units have a very robust system holding the sending units together. The Dakota seems to be among the worst. The float arm is held in position by a small pin that acts as the float arm's pivot point. That small pin protrudes through the outer removable casing on the sending unit and is located with a small bead of solder, then the housing is held against the sending unit with bent tangs on the removable casing. If the bead of solder fails, the float arm can drift away from the contact points as the float moves up and down. Because that soldered pin is also the float arm pivot point, it is constantly moving against the soldered spot. If that float arm has a bend that can roll the float arm as it pivots, the gauge reading becomes flakey. On a Dakota sending unit, this flakey gauge reading tends to be more towards the lower fuel level, as the float arm drops closer to the bottom of the tank (and gets worse as things wear.) I have managed in the past to be able to assure the float arm contact points remain in contact with the sending units wiring, but that success has been limited and doesn't seem consistent from one unit to the next. Basically I'm adding ever so slight more pressure to the float arm so the contact points remain in contact, it takes very little pressure to do that. Too much pressure causes premature failure of the windings, or complete lack of function at all. It really is a pretty poor design, but nearly all are done that way.

Thanks Gene, After a month I am getting to know just what you have said.
I think, (but I can't see enough to do anything about it), is that (as Spock would say) the only logical thing is if it works out of the tank and not in the tank than it must be tank/float interference.
The arms and float location are a bit different on the new pump. I had shortened the arm on the old pump so might try that arm on the new pump.
I also have a bore-scope camera and if the above doesn't work I could drill a small plug-able hole in the top of the tank.
I have drained and partly refilled the tank once but it's a ugly job. The Jeep tank has a small molded recess in the bottom (?) and the Dakota pump when installed is very compressed (Grasping here).
Thanks. I will post when I know anything new.

if you haven't resolved this yet, have to verified the resistance to the gauge is changing as the level changes when installed in the tank?

DO BE VERY CAREFUL while doing so if voltage is present on the circuit IE: key on. A friend was trying to adjust the float arm on a 65 corvette while key was on and another watching the gauge and it created a spark and blew the tank. Fortunately it was empty and just fumes but it still took all the skin off his hand and forearm

No, I have not resolved this yet and thanks for the warning, I appreciate it and I do understand the danger.

I still do think that using the original wiring harness was the way to go but I'm sure starting to have seconded thoughts. I got a new pump that included the sender with no change so I checked all of the many ground splices that the Dakota uses. I did all the tests that the shop manual and most of the contradictory ones that the internet listed with results that seemed to change each time.

My next step would be buying a used dash fuel gauge not knowing if that is the problem and spending more money. This is on a running but not driving car and after wasting mega time, money and stress I decided yesterday to admit defeat, that I am in over my head, swallow my pride, and give up and move on by installing a aftermarket dash mounted fuel gauge and aftermarket sender.

One last test this morning, I ran a new ground wire from the sender to the gauge (printed circuit pin) and bypassed all the Dakota grounds. The gauge went to full like when ground testing the sender circuit and I now have no ignition circuit.

I think my car has turned into ... CHRISTINE !

When you said you ran a new ground wire from the sending unit to the gauge pin and by passed all the Dakota grounds, which wire from the sending unit did you connect to, and did you only connect that wire from the sending unit to the gauge, or did you also connect it to a ground like the frame or something? Did you disconnect the original wire to that pin?

The trucks positive voltage comes from the gauge pin and goes to the sending unit, and then from the sending unit to the ground. When you connect the wire from the gauge directly to the ground, you are telling the gauge the tank is full, so that is why it goes to full. you are bypassing the sending unit. That action should have no effect on the ignition system.

I fail to see how you lost ignition unless you grounded the wrong pin, and blew a fuse. If you connected to the correct pin the most you should have lost would have been the gauge/instrument panel fuse.