So you have a wire bolted to the sending unit, and have it run over to the truck somewhere and bolted to the truck for a ground?

 

Yes,
I have a wire attached to the ground terminal on the sending unit and it is bolted to the chassis crossmember right next to it. Clean ground, no rust or paint in the way.

Today I got a bucket of water and tested the sending unit...works perfectly in the bucket, float is working great, gauge reads accurately.

I re-installed the sending unit and the gauge still reads empty at all times. I ran a ground from the gauge to the sending unit and chassis at the same time, no change.

I am absolutely dumbfounded. There has to be some issue with the tank that I am not able to put my finger on.

 

Or maybe a damaged/broken wire in the wiring harness that doesn't flow electricity when in certain positions? Kinda grasping at straws here, this is an odd situation....

 

I completely agree, grasping for straws is about all I have left.

So since I am now the proud owner of two sending units, I installed the new one and used an angle finder to gauge what angle it sits in the tank. Its 45 degrees. I then used the second sender and held it at 45 degrees in the bucket of water and wired it. I had my wife sit in the cab and read the gauge to me. It was dead on accurate. Moved to full when I lifted the float, moved to empty when I pushed it, etc.

 

The OP bought a "gypo" aftermarket sending unit (probably made in China). Lawd only knows if it's compatible or not.

Post #9: There are 88 of the genuine Ford sending units available NOS .. I didn't list the sources, but I will if asked, but only some of them.

 

 

Did you take the three resistance readings I suggested in post #26? I'm trying to understand if one end of the sender circuit is grounded to the sender body, and thus the tank when installed. Knowing this little detail will make a huge difference for troubleshooting.

I THINK the sender circuit is not supposed to be electrically connected to the sender body, but am not positive. If completely isolated like I'm thinking, the same sender could even be installed in a plastic tank and still work fine. If

My TheoryDuJour™, subject to wind direction and my medication levels: When the sender is installed and the gauge is showing empty, the needle is actually slightly below E. 12 ohms to ground from the sender is designed to put the needle on E. If there was an unwanted path to ground (between the sender circuit and the sender body), you'd have closer to 0 ohms. That should drive the needle slightly below E.

Try the gauge with the sender out of the tank and see exactly where the needle sits. It should be on E. Then install the sender and repeat. Expect to see the needle slightly lower. If so, that confirms there's an unwanted ground path through the sender body and tank.

Did you get both senders from the same source? Could be the same fault with both. Take the readings suggested in post #26 and report back. Maybe the problem is really two bad senders. Don't just rush out and buy a third just yet. Report the resistance values first, and if warranted, get a new sender from a different source.

 

Okay,
I performed the tests in post #26. See below. Also, when the sender is out of the tank and on "empty", the needle is just above the E on the gauge. When installed, the needle is pointed exactly at the middle of the E on the gauge. The two senders look just alike. They are most likely both from the same factory, even if they came in different boxes. I cannot find where I purchased the first one. It was in 2014-2015 and I don't have the receipt, although I am certain it was from rock auto or amazon. The second one was from Amazon and is a Spectra.

Correct, the meter showed right at 1.7 OHM

I connected one lead to the body and one to the ground terminal. I still got 1.7-2.0 ohms

I connected one lead to the body and one to the terminal that goes to the gauge. The reading was 77 ohms empty and 12 full.

This functioned the same as above. 77 ohms empty and 12 full.

 

 

 

I did try it with the ground disconnected, but that was a while ago. Maybe I didn't isolate that variable at the time. I will try that again tonight.

My shop has LED lights, and I wouldn't think they would be playing this much hell with the sending unit, but at this point...who knows???

The way the sender is powered is from the gauge. The gauge is powered with switched 12v from the fuse box. There is no charger on the battery,etc.

 

I have a strange story to tell, its not over yet. Where i worked several years ago, i had a gm van. Electric pump went out. The facility mech changed the whole assembly out. Right after i got it back i noticed the fuel guage didnt work. I was just grateful to get it back and did not say anything. A couple of fill ups later it started working and kept working

Just recently had to change the fuel pump-sending unit in my tahoe. Put it all back together and the fuel guage did not work. On a hunch i have been driving it, using the trip odometer to keep track of the fuel. The guage didnt read, the one day it started reading half with a very full tank. I was going to give up on it till the other day my wife noticed it was reading full with a very full tank. While keeping track of the mileage i have been watching it slowly drop. Still not sure its working yet, but i can keep hoping. I cant explain it except i get the feeling when the new float is first submerged in fuel, its not floating. Fuel is a little different than water ,

 

Funny you mentioned that. I was just telling my wife last night that I am going to go fill the tank all the way up, drive at least one full tank through it and hope for a miracle, haha. Maybe this is a sign that I should do just that!

 

So I gets to thinking some more...

Obviously something is changing when the sender is installed. But what?

Try this. For now, let's take the gauge completely out of the equation. Let's stick to just a resistance reading from the sender's insulated terminal (the one to the gauge) and the ground point on the frame. Keep the sender's ground wire (to the frame) connected. In other words, I want the ohmmeter to see what the gauge would normally see.

Take a resistance reading with the sender not installed, but with the ground wire connected. Next, put the sender in the tank so it's making metal to metal contact, but do NOT install the locking ring yet. If the resistance changes just from the sender body touching the tank, then we know it's an electrical issue.

If no change yet, install and tighten the locking ring while watching the meter. If you now get a change, that would indicate a mechanical issue with the sender. Tightening the locking ring is warping something and somehow affecting the resistance of the sender.