The heavier spring is the one made of thicker wire, period. Loop size is not involved. As I recall, the heavier spring in the distributor on the W2 engine had the large loop.
The Valiant's sender ground is literally a fuse tap held onto the pickup's outlet pipe by a power-steering return hose clamp. Cheesy as hell, yet the gauge works perfectly.
First and foremost, double-check your ground connection. You say it's good and I believe you. Check it anyway. Now check your connections at the bulkhead disconnect. If nothing looks burned or melted, simply push on all the wires to ensure they're fully seated in the insulator. Do this from
both sides of the firewall. Check gauge operation before proceeding. No change? Proceed.
As usual, I'm making this as detailed as possible for lurkers/future readers who might find this while having the same issue. That doesn't give you permission to skip anything unless indicated in the text. Read everything anyhow; don't skip around or try to skim this. It's a process. Do you want it to work or not? Despite having auxiliaries, all my
factory gauges work too along with the oil light (using a 20PSI switch). I've been down this road.
Step 1: Does your
temperature gauge work? I don't
care that it's not the issue, you need to know for sure. If it works properly, move on to
Step 2. If it's connected and doesn't seem to function, skip directly to
Step 5. If it's not connected, ground its wire (should be violet on your car) direct to battery (-) to see if it works. The gauge should slowly swing well into overheat range, very close to the "H" hashmark. This will
not happen instantaneously; give the gauge and limiter some time to reach operating temperature (yes, that's a thing--they're thermal gauges). Gauge pegged as expected? Proceed to the next step. No gauge movement? Skip directly to
Step 5.
Step 2: Check the sender:
With the key off, connect your DMM between the tank ground and the sender's output stud. It should read between 8.6-85Ω depending on how much fuel is in the tank. More fuel = lower reading. This test
requires a DMM; at 50Ω+ on a 12V circuit a test lamp might not even light. If it tests open or outside that range, the sender's faulty. If the reading seems plausible, we've now eliminated that as a potential problem.
Step 3: Since
know it's grounded, the next check is pretty simple: Connect the tank ground and sender output stud with a jumper wire. Turn the key to the RUN position. The gauge should peg at or beyond the "F" mark, with the needle no
less than 1/32"
below the "F" hashmark. The nature of the thermal gauges means you'll get a slow needle swing up to "F",
not a lovely snap right to it. Give it time. If the gauge doesn't peg, proceed. If it does, there is
absolutely no reason the fuel gauge shouldn't work. Triple-check your ground.
Step 4: Check continuity between the sender wire and the rear body wiring disconnect (on my Valiant, this is in the driver's-side kick panel area, likely the same on your car). You're looking for the dark blue wire on Cavity 2. Check continuity on the
dash side of the connector to ensure the connection is good between the sender and the dash wiring. Verifying continuity betwixt this connection and the gauge can be done by jumping this terminal to ground, which should peg the gauge. No? Either the wire is cut/broken between here and the cluster (not likely) or we're into the cluster itself.
Step 5: Now we'll check the cluster limiter. Connect the negative lead of your DMM or test light to battery (-), and the positive lead to the factory
temperature sender wire (violet) with it disconnected from the sender itself. Turn the key to the RUN position. Once again accounting for warm-up, the DMM should show a pulsing ("throbbing" if you're
@Jester) 12V on the sender wire as the cluster limiter cycles. Using a test light, a good circuit will cause the light to flash slowly. If you have pulsing voltage/flashing, the limiter is good. If there's nothing (again, give it a mnute), you're open somewhere in the limiter circuit or cluster itself. If you have a constant solid battery voltage or constant illumination of the test light, the limiter's stuck. I doubt the latter is the problem; stuck limiters usually result in pegged temp and fuel readings.
If, at this point, everything checks out as it should (limiter indicates good, tank sender checks out, gauge pegs when shorted), you've either got a bad gauge or a burned trace on the cluster PCB.
We've avoided disassembling shit as long as we can. Disconnect the battery because it's time to pull the cluster for testing. This requires lowering the column, so loosen (but don't fully remove) the two bolts on the column anchor plate and remove the nuts holding the column to the dash to lower the column. That should get you enough clearance for the cluster to come out, if only barely. Don't bother leaving anything connected such as the ammeter wires; it's much easier to do your testing with the cluster fully out.
Step 6: Cluster diagnostics
a) The first order of business is to look for any burned traces on the PCB. These are usually painfully obvious, but using the DMM to check continuity between each pin on the connector and its termination point is a good idea; they do develop hard-to-see cracks. Any burns or breaks can be repaired with a jumper wire carefully soldered at the base of the pin and the trace's termination, or by replacing the PCB. I chose the former; $150 is ludicrous. Any component (gauge, bulb, etc.) at the end of a burned trace is almost certainly fried.
b) With the limiter removed, connect a
new AA/C/D battery (anything 1.5V--do NOT use a 9V battery; you'll burn the gauge) across the studs of the fuel gauge. The (-) side of the battery connects to the stud marked "S". The needle should swing (very slowly, remember?) to about 1/4-1/3 tank. We're not too concerned with calibration so much as
movement. If the needle doesn't move, the gauge is dead. Check your temp gauge while you're there, or be like me and do a bunch of redundant work later after not bothering to test it. While it doesn't apply here, the same test works for oil-pressure gauges. All three should read approximately 1/4-1/3 scale on a 1.5V battery.
DO NOT ATTEMPT TO TEST THE AMMETER in this manner. At best, you'll burn your fingers; at worst the battery will explode. Leave the ammeter alone. If it's exhibited problems in the cluster, replace it.
c) If there are no burned/broken traces and the gauges check out, make sure all the connections are clean and the nuts are tight--don't go crazy here, they gauges are fragile. You can test the limiter if you'd like (see below) but I'd suggest replacing it with a solid-state unit such as those sold by
RT Engineering. The two styles they make cover most muscle-era Mopars; the non-Rallye '67-'76 A-body cluster uses their part number
IVR-4. I put one in the Valiant and I have one for the Challenger. It's cheap insurance ($50). You can also make your own using a 7805 linear regulator and some other components, but those run very hot (never good) and do not provide any short/open-circuit protection for your increasingly-expensive gauges.
- c1) To test this style of limiter, connect the case's ground terminal to the negative side (-) of a 12V source. Connect the insulated terminal (in PCB material) closest to the ground to the positive side (+) of the same source. Now connect one side of a 12V light bulb to the ground terminal and the other side to the terminal furthest from the ground. The light bulb should flash on and off as the limiter cycles. No or constant light means the limiter's junk.
d) Replace
every single freakin' bulb in the cluster with a new one. Changing bulbs with the cluster installed
sucks. Do it now. I do
not recommend LEDs unless you can find dimmable ones. Hint: They're
expensive, unlike anything offered on Amazon.
e) When you reconnect the cluster, make sure all the terminals in the insulator are fully inserted and seated on their respecitve pins. Check continuity between the connector wires and their termination points on the PCB.
7) Reinstall gauge cluster & steering column. Reconnect power. Start engine, and witness the power of your fully-armed and operational battl... er, instrument cluster. Feels pretty good, huh?
If, after all this, you can't get a working gauge you'll need to confer with a demonologist or Catholic bishop well-versed in exorcisms, because your car is clearly posessed. Of course, that's not the case because you found the problem somewhere during this procedure. If you didn't, you fiddle-fucked around, skipped step(s), or whatever and somehow missed the problem.