revisit the fuel gauge and sending unit issue

-
Pawned

Pawned

N.R.A. Lifetime Member - And damn proud of it
Joined
Jun 29, 2013
Messages
1,605
Reaction score
185
Location
Valle del Sol, AZ
After ignoring it for a few years I worked on the fuel gauge problem. I am using an after market gauge but this has all the same issues as the OEM gauge.
What is interesting is the sending unit is measuring 53Ω at the sending unit and also at the wire at the back of the gauge (not plugged in). But the gauge will not register properly.
I hooked up a resistor box, I made, to the fuel gauge and it works properly with the resistor box but not the wire to the sending unit.
Thinking the problem may be in the wiring connections, I ran a dedicated conductor from the sending unit to the gauge and there was no change. problem still exists.
I am using 16 awg wire where the OEM was 18 awg. So there is no voltage drop. The reading is exactly the same at either end of the wire.

Some how the sending unit is registering a resistance to the multi meter I am using. ( a Fluke-true rms meter). I have heard of phantom voltages and have seen them, but I have never seen or even heard of a phantom resistance.

I am dumbfounded at this point. This is the 2nd or 3rd sending unit I have bought so I am not very interested to buy another. It must be something about the resistance/conductivity of the gasoline shunting the signal to ground but that is a guess at best
 
yea good luck,the few ive tried are junk.ohm out somewhat ok out of the tank,in the tank,well your quess is as good as mine as to what the fuel level is.
 
jdsduster said:
yea good luck,the few ive tried are junk.ohm out somewhat ok out of the tank,in the tank,well your quess is as good as mine as to what the fuel level is.
Click to expand...
I just had a thought, I am going to play with the grounding of the sender. will let you know in a couple days
 
You've been through multiple aftermarket fuel sending units all with the same problem?

Since you've run a dedicated wire with no change I'd check the integrity of the ground to the sending unit.

If that doesn't fix it i'd look at the gauge itself.
 
Yes. Explain where you stuck the black lead of your DMM when you measured the sender resistance. The factory's ground was a metal clip across the short rubber hose on the sender. Assuming that iffy connection works, that only takes the sender's gnd to the fuel tube which must be grounded to the car body via the mounting clips (rusty & under-coating). Gasoline is a non-conductor. Some here fab their own ground wire to the sender.
 
BillGrissom said:
Yes. Explain where you stuck the black lead of your DMM when you measured the sender resistance. The factory's ground was a metal clip across the short rubber hose on the sender. Assuming that iffy connection works, that only takes the sender's gnd to the fuel tube which must be grounded to the car body via the mounting clips (rusty & under-coating). Gasoline is a non-conductor. Some here fab their own ground wire to the sender.
Click to expand...

My sender is grounded to two points of the chassis and the gas tank has a separate ground to the chassis. Not including the ground created by the mounting the tank to the chassis.

PS I am not disputing your statement on gas being a non-conductor. BUT, The additives in gasoline may. Ethonol, detergents, additives.
Just something to think about
 
Last edited:
kiss said:
You've been through multiple aftermarket fuel sending units all with the same problem?

Since you've run a dedicated wire with no change I'd check the integrity of the ground to the sending unit.

If that doesn't fix it i'd look at the gauge itself.
Click to expand...
The wire measures 53Ωs at the sender AND at the fuel gauge. But you are correct. there is nothing left but to reexamine the grounding.

I redid the grounding. I even made it single point grounding, what does not matter at all. I thought about ground loops but that is not applicable.
Everything I did and the gauge did not change.

When I was a microwave engineer, somewhat fresh out of school, we had a term to describe this It is "FM" (******* magic)
 
Last edited:
So what does the gauge show at the 53 ohms? Without going to my references, I think you should see approx' 1/8th of a tank.
 
RedFish said:
So what does the gauge show at the 53 ohms? Without going to my references, I think you should see approx' 1/8th of a tank.
Click to expand...
It is barely above E By the miles I have driven since the last fill up there should be at least 5/8th of a tank.

The ohm setting was incorrect as the car was lifted from the rear at the time
 
So here's the deal, theres only a few things that can be wrong in your case. Let me list them out.
1. Sending unit ground. You said you have 2 ground points on the sending unit, that should be enough. I've seen several Mopar sending units work without the ground strap present. I doubt that's the issue.
2. Wire integrity. You said you ran a dedicated wire and saw no difference vs the stock wire. This also probably isn't the issue.
3. Gauge. You said it's a new aftermarket gauge, so it most likely works. Is the gauge linear or non-linear? What is the gauge calibrated for?
4. Sending unit. You said you bought several sending units with no difference. Did you ohm each of these outside of the car with the same results?
5. Sending unit position. I've seen several aftermarket sending units where the pickup tube wasn't in the right place, it was either too high (most common) or too low (you can tell that easily because it won't seal correctly with the gasket/locking ring). In addition to this issue I've seen improper movement of the float arm, such as the arm won't go near the bottom of the tank, so your tank is over half full but the gauge shows almost empty. This can be fixed by bending the arm. You would need to have the tank out of the car and look through the filler neck hole to check this. Me personally i've never gotten lucky with aftermarket sending units, each one I've installed I have had to do it with the tank out to ensure that the pickup tube reaches the bottom of the tank and that the float arm is going through proper range of motion.

Have you considered just getting an OEM sending unit and being done with it?
 
Pawned said:
Where?
Click to expand...
Mr Gorsky said:
Yeah, someone link me to the new OEM sending unit website :)
Click to expand...
To find an OEM one post in the parts wanted section on this forum. Also check ebay (that's where I got mine). For ebay searches you can look up some of the OEM part numbers. Search this forum for "OEM sending unit" and you will find some part numbers.

Call these 4 places that do rebuilds and see if they have any on the shelf:
Rebuilt Fuel Sending Unit Specialist
Home
Fuel tank sending unit rebuild service
Fuel Senders
 
Whether you are spot on correct about how much fuel is actually onboard or not, the linear sender is the problem. The linear sender has a 80-10 ohms range and will show you those end points at the workbench. It will also show 35 ohms at 1/2 range. The OEM thermal gauge needs approx' 23 ohms signal to point its half range. That 35 ohms is within needle width of 1/4 hash mark on the OEM gauge screen. You might be able to verify that with your homemade tester if you wish.
Bottom line, every ohms value between end points of a linear sender is wrong. A meter match module will correct this. In any case your 53 ohms would be well below the fuel level you submit to having. I'll go ahead and add... My 67 isn't driven often. I fire it up at least once per month and let it run 30 minutes. I might move it to wash it. I might blast off for a 20 mile or 50 mile run. More often just sits and idles. So the odometer reading is no indication of how much fuel was burned.
 
The senders are operating properly. They are providing the correct resistance value.

The last sender that I pulled is sitting around here somewhere. after I pulled it I hooked it up to the gauges on a test bench and they controlled the gauge correctly.
The problem is in the vehicle as a whole. I eventually will find the problem, my lips to Gods ears. There is some kind of FM going on here.
Maybe add some type of pull up resistor ( or some other invention not yet discovered) to strengthen the circuit from the tank to the gauge
 
Pawned said:
The senders are operating properly. They are providing the correct resistance value.

The last sender that I pulled is sitting around here somewhere. after I pulled it I hooked it up to the gauges on a test bench and they controlled the gauge correctly.
The problem is in the vehicle as a whole. I eventually will find the problem, my lips to Gods ears. There is some kind of FM going on here.
Maybe add some type of pull up resistor ( or some other invention not yet discovered) to strengthen the circuit from the tank to the gauge
Click to expand...
Then I would guess it's issue #5 that I listed in post #10. Specifically the float arm going through full range of motion. The only way to check that is to pull the tank, install the sender, and check the arm range of motion via the filler neck hole.
 
So you believe your aftermarket senders are different and more accurate than all others
You see 53 ohms at that sender and at the gauge and looking for approx' 19 ohm response from that gauge. Good luck
 
RedFish said:
So you believe your aftermarket senders are different and more accurate than all others
You see 53 ohms at that sender and at the gauge and looking for approx' 19 ohm response from that gauge. Good luck
Click to expand...
Well when I placed the float at appx half full the gauge showed 1/2 and again at E, 1/4, 3/4 and full
 
OK here I go: I ran the dedicated line from the sending unit to the gauge and then re-grounded the sender and tank. The fuel gauge is now operational. But does not register properly.
Now I need to find a sending unit that is more linear to get it to register properly.
On the existing non-linear meter, I have tried to add resistance, serial and/or parallel into the circuit and it never showed any improvement. But that is before I had a known working gauge. I am thinking ( no, I have not calculated anything yet, this is off the top of my head.)
I am thinking that I could add a 10 ohm resistor serial just before the meter. This will establish the Empty mark and it will not register lower than E. Also add some type of resistor in parallel to the sender that will at the very least prevent it from pinning the meter to the right above F.
And hopefully give some type of readings between the Empty and Full

If I am correct I will let you know. If I am not correct I will not tell you. No sense looking like a damn fool over and over:rofl:

Ed
 
The non-linearity of the replacement and aftermarket senders is well known, or it should be. The replacements are wound in a pyramid shape, while the originals are modified from that, concave on the sides where it is wound. It's a slope, expressed on a graph.
 
Maybe now that I have been able to get the gauge to operate, I should try the MeterMatch. I have tried it numerous times and never once worked properly. Maybe I will try it again when I am dumb enough to tear into this again
 
Pawned, look up metermatch, I had the same issue with my gas gauge and this pretty much solved it. This is a device that will calibrate your sending unit to your gauge even if they are missed matched.
 
Mr.FT said:
Pawned, look up metermatch, I had the same issue with my gas gauge and this pretty much solved it. This is a device that will calibrate your sending unit to your gauge even if they are missed matched.
Click to expand...
I have had one for a few years now. I have set it up at least 3 times and it never has worked once I took it off "program".
like I said, not that I redid the wiring and grounding to the gauge and the gauge is valid, I will attempt the MeterMatch again.
 
I just realized how to get the fuel gauge to read E when it is empty. I am no longer going to worry about the sender.

I am going to place a separate adjustable voltage regulator the fuel gauge separate from the rest of the dash. Then I will adjust the supply voltage of the gauge until it reads empty when it is empty. The only reading I need is an accurate empty, I can adjust my thinking to figurer out the rest of the setting desplayed
 
If you build your own voltage regulator, instead of using a fixed 7805, use a LM317 or LM350T. They are adjustable regulators that use the same TO-220 package. The data sheets can be found on Google along with circuit designs. Only a few resistors are required for the adjustment circuit. I had one in a car and it was calibrated to read correct at empty. Beechcraft has used adjustable power supplies in some of their airplanes and they calibrate them to read correct at empty. I personally like to have the gauge more accurate at empty to show if you have too little gas, because the only time you can have too much gas is if you're on fire.
 
You must log in or register to reply here.
-

Back
Top