Fuel Guage Goes Up Then Partly Fades -- 72 Swinger

-

vames

FABO Gold Member
FABO Gold Member
Joined
Aug 24, 2021
Messages
130
Reaction score
167
Location
Long Beach, CA
Hi all. Apologies for the long story.

I'm on a quest to have a working fuel gauge in my 72 Swinger-- I've read virtually every post on this site and I clearly still have a lot more more troubleshooting to do. But as I go through everything, I'm having a problem that makes little sense to me.

When I turn the key, the gauge moves up to a fair representation of where the fuel level is. But then over the next few seconds, it fades back to roughly half of what is a fair representation. See the video below -- all I did was turn the key.

The facts:
  • Previously nothing was working, but I am getting action now that I replaced the IVR with what Classic Industries claims is a solid state unit.
  • The tank has a new(ish) spectra FG69B sender. I know these aren't terribly accurate, but I haven't seen anybody suggest that it could register changing resistance when the float is at a fixed level. I have a NOS sender that I plan to put in when I'm ready to get down on the floor, but I was hoping to get everything in working order before I do so.
  • I have a ground strap from the sender to the metal fuel line. The wire from the sender to the dash checks out fine. All the connections are clean. And the gauge goes to full and stays there when I ground the sender connection (no fading).
Just trying to see if this is familiar to anyone, and (given my limited time to work on this) see if my next steps in troubleshooting seem like the right direction:

1) Check to see if its just a really bad Spectra sending unit. Test everything as-is with the NOS mopar sender (out of the tank) hooked directly to the fuel gauge pin and a ground. Move it through its motions and see what I get on the gauge. If it works, I know I only need to swap the senders. If it still fades, go to #2.

2) Measure the voltage output of the new IVR. If it wanders, then I know it's a dud. If it's a steady ~5 VDC, then move onto #3.

3) Check the quality of the ground on the fuel line that the ground strap goes onto. It seems like a longshot that it's a bad ground with resistance that stays in the 10-80 ohm range (evidenced by the gauge needle staying in bounds) but maybe something weird here? If the ground is good, then #4.

4) Could the gauge itself be toast? Has anyone seen them fail in this way? Do I have to buy a few resistors (like a 10, a 40, and an 80) and use a 5 VDC source to bench test it or is there a better way?


It's not that complicated of a system FFS! But it's also not just sitting on the bench and easily tested. Each test requires dropping the steering wheel, pulling the cluster, crawling under the tank, and all that. So I'm looking for some wisdom or knowledge of this problem just to see if I can narrow it down.

 
The new senders don't work. If you're old one is in good shape, use it. If not, get it reconditioned.
 
I was aware that the new senders not accurate. I just wasn't aware that they were actually faulty (ie. they change resistance without the float actually moving -- that is just crazy!)
 
That's awesome. I'm thinking the IVR is the fault.

If it is truly electric I see no reason for it's output to go high then settle down.

Now technically the origional mechanical ones would provide full voltage for a few seconds then start osolating on and off to regulate the votage / current.

The one you have MIGHT be mimicking that.

The Realtime Engineering IVRs claim that is what they do and even output a square wave.

How much did you pay for your IVR.

The resisters you would need are 10, 23 and 70 the gauge and sender are not liner.

With these values you should see ABOUT full, ABOUT half and ABOUT empty.

I have a graph in a number of posts that show the curve.

NOW it is entirely possible that the level shown is correct for the crappy aftermarket sender.

Many people self included report

Gallons in tank = gauge reading
Full = full
3/4 = 1/2
1/2 = 1/4 to E


Sender resistance in FSM

RTE limiter - rte


"Our limiter exactly duplicates the original limiters function by slowly switching 12V on and off.
Our limiter has a built in polyfuse which protects the limiter and your wiring from short circuits on the output of the limtier
Our limiter constantly looks for short circuits on the output of the limiter and will switch itself off if a short circuit is detected. When the short is removed, then the limiter starts working again.
Our limiter has a built in diagnostic LED that blinks when the limiter is on, helping you see that you have power to your dash and that the limiter is operating properly.
Our limiter has a warm up time at turn on, same as the original mechanical limiter. This means that your gauges will come up to the proper reading very quickly.
Our limiter is always outputting a constant 5V average, no matter what the input voltage is (Within the range of 9V to 18V). If the input voltage goes too high or too low, then the limiter shuts off"
 
Last edited:
Dana67 -- That's interesting and I never thought of that. So you're saying that the original mechanicals (and maybe this allegedly electronic one) would surge with a full 12V for a few seconds and then start oscillating to drop down to 5V? Maybe some sort of sequence to get all the bi-metal in the IVR and gauge up to temperature? I don't recall my dart's gauge ever doing that, but to be honest I've been without a working gauge for over a decade -- so long that I might have just forgotten. (I do know that many new cars peg the gauges then settle back when you start the car -- maybe they're mimicking old Mopars hah!)

At any rate, you may be right, and my #1 test of my NOS Mopar guage will let me know if that is right. Results will be forthcoming if I can just get a few hours to test it this weekend.
 
If this were my problem, I would use the resistor test (if the resistors were readily available). Instead of dropping steering column and messing under the dash, simply unplug the harness at the tank and sub in the resisters and see how the gauge reacts. Real easy, and real cheap. One end of the resister to the gauge harness, and the other end to a good ground. Go from there. I bench tested my dash assembly this way to ensure everything dash related was working before installing the dash.
 
Does the temp gauge work? I would FOR CERTAIN perform a resistor test---sub a resistor for both the temp and fuel senders and see what the gauges do. The IVR supplies both. One thing you might actually try, is to remove the (temporarily) sender wire at the left kick panel connector where the tail harness comes up. Now "rig" and extend the wire through into the engine bay (you can knock the clutch rod plug/ grommet out temporarily, run a wire and hook it to the temp gauge

FOR SURE determine if the temp gauge is "doing the same."

If not, might be loose stud nuts on the gauge, or the gauge itself, or poor connection in the path, or poor connection at the sender wiper arm (where it contacts the resistor)
 
I agree that the resister test would do alit to prove things out.

As for the gauge climbing then falling, I too have never seen that. I recently had a mechanical IVR failure and decided to go with the RT-eng IVR. the fuel gauge works just as with the mechanical IVR. also in your video the speed that it climbs seems too fast.

If you can easily reach the back side of your cluster you can clip a lead on to the correct post and read the voltage from key on to settling down.

FYI, Mechanical and the RT-Eng will give a digital multimeter a run for it's money trying to read a steady voltage.

Since to is on or off it is either 12v or 0 v and the meter will fluctuate drastically.

If you can get an analog volt meter you could watch the needle fluctuate. A test light will work as well for functionality just not for voltage.
 
I agree that the resister test would do alit to prove things out.

As for the gauge climbing then falling, I too have never seen that. I recently had a mechanical IVR failure and decided to go with the RT-eng IVR. the fuel gauge works just as with the mechanical IVR. also in your video the speed that it climbs seems too fast.

If you can easily reach the back side of your cluster you can clip a lead on to the correct post and read the voltage from key on to settling down.

FYI, Mechanical and the RT-Eng will give a digital multimeter a run for it's money trying to read a steady voltage.

Since to is on or off it is either 12v or 0 v and the meter will fluctuate drastically.

If you can get an analog volt meter you could watch the needle fluctuate. A test light will work as well for functionality just not for voltage.
A good quality Multimeter (I use a Fluke 88) can give you an "average" function.
 
My gauge does the same thin when I have the car running while pumping gas.... My engine gains on the pump....
 
fuel gauge in my 72 Swinger
Just a thought. Does the 72 have a removable IVR like 67 to 70 something darts. Or is it integrated into the fuel gauge?

If Steven attached the IVR by adapting an external then might the internal and external be competing with each other.
 
Hi all. Apologies for the long story.

I'm on a quest to have a working fuel gauge in my 72 Swinger-- I've read virtually every post on this site and I clearly still have a lot more more troubleshooting to do. But as I go through everything, I'm having a problem that makes little sense to me.

When I turn the key, the gauge moves up to a fair representation of where the fuel level is. But then over the next few seconds, it fades back to roughly half of what is a fair representation. See the video below -- all I did was turn the key.
I've seen similar behavior in temp gauge also but that wouldn't happen until the water is hot/engine up to temp. Switch off then on again and needle would go quickly to max then drop back to proper reading. Solid state regulator installed, all connections cleaned, stopped the wild needle swings.
 
After fixing some more pressing stuff, I'm back on the fuel gauge issue. I'd like to start with the "resistor test" by seeing what readings I get at the gauge with appropriate resistors placed between the sender wire and ground. So now to figure out how to do this: can someone tell me if I'm on the right track?

Since I don't have resistors on hand, my plan is to buy a pack of 10 ohm resistors. I'll put 7 of them in sequence to get 70 ohms (simulating ~empty), 2 in sequence to get 20 ohms (simulating a tad less than half) and one in line to get 10 ohms (simulating ~full). Then I'll put each of those three configurations between the sender wire (one at a time of course) and see where the needle lands.

Before I order those resistors, does anyone know if the tolerances of the resistors matter? (looks like they have the 10 ohm resistors at 1/2 Watt, 1 Watt or 5 Watt).

I may be way overthinking this, but if I remember Ohm's Law correctly, V=IR and P=IV, so P=V^2/R. For the 10 ohm reisistor, that would mean that there is about 0.5 amps of current running through the sender wire and P would be about 2.5 Watts. For a 20 ohm resistor, 1.25 Watts, and for a 70 ohm resistor, 0.36 Watts.

So if I just get 5 Watt resistors, am I good?

Sorry, this is getting a bit nerdy -- and possibly doesn't even matter.
 
Your spot on!

I would just get 5 watt for all.

As for tolerance, go for price. I bought 10 x 10 ohm 100w most were 9.8 to 10. all close enough for what your doing

If you were going to calibrate the gauge then it might make a difference.
 
Your spot on!

I would just get 5 watt for all.

As for tolerance, go for price. I bought 10 x 10 ohm 100w most were 9.8 to 10. all close enough for what your doing

If you were going to calibrate the gauge then it might make a difference.


Many thanks. I will hopefully post results this weekend.
 
If you get 5 or 10 x 1 ohm also you can mix and match to get all the happy resistances.

I was going to make a fancy decade resister box, but ended up just soldering the ends of the 10 ohms in series and the 1 ohm the same with 3 jumper wires I can make any resistance from 1 ohm to 110 ohm
I bought some 24 inch 18 guage wire with alagator clips on the ends. Works great.

Also very handy to have the wire with clips.

Bypassing ballast resister, jump starting the ignition etc

Remember this is the values right out of the 67 FSM

Empty is 73 +/- 12 ohms (61 to 85)
Full is 9.6 +/- 1 ohm (8.6 to 10.6)

I have posted a graph the I made with 3 different senders and what MY Fuel gauge wants in terms of resistance for E 1/4 1/2 3/4 and F
 
Last edited:
Good info -- Thanks! If I had only known how much the Spectras suck, I'd have never thrown out my (still working) original when I changed the tank. Live and learn. Now I've got a NOS 2421493 but before I put it in, I'm trying to make sure everything else in the system is in working order -- hence the resistor test.

I'll definitely solder the resistors together in the configurations listed above with a clip on one end for ground and a screw that fits the sender wire fitting on the other, test them with a multimeter to make sure they are close, then test the gauge. I'll then put them in a box and use them in another 20 years when I have to go through this again.
 
Lots of good info in this thread.

I looked at your video and that's quite a swing. I did all the tests mentioned here with mine and found my dash gauge to be inaccurate. However I did try a 12 volt fuel gauge hooked up to the sender and it was consistent and accurate. Mine is a problem when it gets down to 1/4 tank. So i've learned to live with it and fill up when it gets there. Hope you have better results.
 
Finally got some time to get some incremental information on this. I built 3 resistors (10, 20, and 70) and verified them with a digital meter. Results:

10 ohms -- the needle went up to 3/4 and promptly settled back down to 1/2. It proceeded to wander a bit around 1/2 +/-.

20 ohms -- the needle went up to just above 1/2 then settled back down to about 1/4 with some tight wandering afterward.

70 ohms -- the needle went just above empty and settled just below empty.

So it's clearly not just my janky spectra sender that is causing this problem. This does eliminate the sender ground as a problem since the resistors went straight to battery ground and I'm still getting this behavior. My best guess at this point is that my $15.99 IVR from Classic Industries is a dud. (I assumed the one sold by CI was solid state due to their statement "replaces the original mechanical regulator with modern electronics" - and maybe it really is, but the evidence at this stage is pointing its finger at that regulator being bad.)

Since this is a daily driver and I don't get much time to pull the dash repeatedly, on my next dash-pull I'm going to replace the IVR with an RT-Engineering solid state. I've already cleaned and verified all of the dash connections and verified conductivity for the full length of the fuel sender wire. I know I have the skills to build a robust IVR from scratch, but I just don't have the time to mess with it.

The temp gauge is not working, but I've found that problem is break in the wire somewhere between the temp sender and the gauge. I'll probably just run a new wire from the temp sender to the instrument panel, rather than trying to trace the problem with the old temp sensor wiring. Then I'll test the new RT regulator. If I'm still having the problem, then I'll know it's gauge failure and I'll try new gauges.

And some day I'll be able to see how much gas I have and what temp my engine is running.
 
Finally got some time to get some incremental information on this. I built 3 resistors (10, 20, and 70) and verified them with a digital meter. Results:

10 ohms -- the needle went up to 3/4 and promptly settled back down to 1/2. It proceeded to wander a bit around 1/2 +/-.

20 ohms -- the needle went up to just above 1/2 then settled back down to about 1/4 with some tight wandering afterward.

70 ohms -- the needle went just above empty and settled just below empty.

So it's clearly not just my janky spectra sender that is causing this problem. This does eliminate the sender ground as a problem since the resistors went straight to battery ground and I'm still getting this behavior. My best guess at this point is that my $15.99 IVR from Classic Industries is a dud. (I assumed the one sold by CI was solid state due to their statement "replaces the original mechanical regulator with modern electronics" - and maybe it really is, but the evidence at this stage is pointing its finger at that regulator being bad.)

Since this is a daily driver and I don't get much time to pull the dash repeatedly, on my next dash-pull I'm going to replace the IVR with an RT-Engineering solid state. I've already cleaned and verified all of the dash connections and verified conductivity for the full length of the fuel sender wire. I know I have the skills to build a robust IVR from scratch, but I just don't have the time to mess with it.

The temp gauge is not working, but I've found that problem is break in the wire somewhere between the temp sender and the gauge. I'll probably just run a new wire from the temp sender to the instrument panel, rather than trying to trace the problem with the old temp sensor wiring. Then I'll test the new RT regulator. If I'm still having the problem, then I'll know it's gauge failure and I'll try new gauges.

And some day I'll be able to see how much gas I have and what temp my engine is running.

The CI IVR has the same part number as a generic regulator used on these cars and available through many parts chain stores. That right there tells me that your reg is the likely culprit. The RT one looks like a much better design.
 
Hi all. Apologies for the long story.

I'm on a quest to have a working fuel gauge in my 72 Swinger-- I've read virtually every post on this site and I clearly still have a lot more more troubleshooting to do. But as I go through everything, I'm having a problem that makes little sense to me.

When I turn the key, the gauge moves up to a fair representation of where the fuel level is. But then over the next few seconds, it fades back to roughly half of what is a fair representation. See the video below -- all I did was turn the key.

The facts:
  • Previously nothing was working, but I am getting action now that I replaced the IVR with what Classic Industries claims is a solid state unit.
  • The tank has a new(ish) spectra FG69B sender. I know these aren't terribly accurate, but I haven't seen anybody suggest that it could register changing resistance when the float is at a fixed level. I have a NOS sender that I plan to put in when I'm ready to get down on the floor, but I was hoping to get everything in working order before I do so.
  • I have a ground strap from the sender to the metal fuel line. The wire from the sender to the dash checks out fine. All the connections are clean. And the gauge goes to full and stays there when I ground the sender connection (no fading).
Just trying to see if this is familiar to anyone, and (given my limited time to work on this) see if my next steps in troubleshooting seem like the right direction:

1) Check to see if its just a really bad Spectra sending unit. Test everything as-is with the NOS mopar sender (out of the tank) hooked directly to the fuel gauge pin and a ground. Move it through its motions and see what I get on the gauge. If it works, I know I only need to swap the senders. If it still fades, go to #2.

2) Measure the voltage output of the new IVR. If it wanders, then I know it's a dud. If it's a steady ~5 VDC, then move onto #3.

3) Check the quality of the ground on the fuel line that the ground strap goes onto. It seems like a longshot that it's a bad ground with resistance that stays in the 10-80 ohm range (evidenced by the gauge needle staying in bounds) but maybe something weird here? If the ground is good, then #4.

4) Could the gauge itself be toast? Has anyone seen them fail in this way? Do I have to buy a few resistors (like a 10, a 40, and an 80) and use a 5 VDC source to bench test it or is there a better way?


It's not that complicated of a system FFS! But it's also not just sitting on the bench and easily tested. Each test requires dropping the steering wheel, pulling the cluster, crawling under the tank, and all that. So I'm looking for some wisdom or knowledge of this problem just to see if I can narrow it down.


The new senders from Classic Industries are shitty. I got one from them last year and I’m replacing it again now. Unfortunately, i dont have the factory one to refurbish. From everything I’ve read on here, thats the smart bet.
 
The new senders from Classic Industries are shitty. I got one from them last year and I’m replacing it again now. Unfortunately, i dont have the factory one to refurbish. From everything I’ve read on here, thats the smart bet.

Yeah the new senders from everywhere seem like crap. Luckily I have a nice NOS unit from the 70's, but I'm trying to eliminate all the other gremlins before I put it in.
 
More progress. Last night, I put in the RT Engineering IVR. Sure enough, the weird behavior on the gas gauge went away, and the resistor test checks out great with about Full @ 10ohm, Half @ 20 ohm, and and empty at 70 ohm. This is all good news. I am a little angry that the IVR that Classic Industries is selling was a dud (probably just bad luck but its crap to buy a new part that doesn't work correctly). With that sorted out, I just have to burn through this tank of gas until its almost empty so that I can pull the garbage Spectra sender and install the NOS Mopar sender I have.

The sad news comes with my temp gauge. I confirmed that the temp sender wire was continuous from the sender itself to the instrument cluster plug. I now know the IVR is good. And yet the resistor test does not get the temp gauge to move. So in this case, I now think the actual temp gauge is busted. Not a huge deal -- I have several other temp gauges -- but damn it I don't want to pull the dash cluster out yet again :(.

I'm torn between 1) wanting to tear open the IVR from CI to see what exactly it is that they sold me 2) returning it to CI and getting my $15.99 back, or 3) chucking it into the trash, writing a bad review for that item and moving on with my life.
 
-
Back
Top