Gauges come on when I turn on the parking lights or headlights (1969)

[MaximRecoil]

Without the engine running and with the key in the off position, if you turn on the parking lights or headlights, after a few seconds, the needles on the gas gauge and the oil pressure gauge start to rise. I don't know about the other two gauges, because I have the ammeter bypassed, and the temperature gauge doesn't work anyway.

Leaving the parking lights or headlights on, if you then turn the engine on, or the key to on or accessory, the needles will rise even more, more than they should. For example, the gas gauge will show far more gas than is actually in the tank, and it will also fluctuate, even though the car is sitting still. Turning the lights off makes things go back to normal with the gauges.

I recently had the instrument cluster out to work on the vacuum switch, and I may have plugged some wiring back in wrong, specifically on the headlight switch, because it is missing the plastic plug on the harness, so the wires have to be plugged onto the terminals individually. The way I have it is as follows (starting from the bottom of the switch):

1. Green wire
2. Pink wire and gray wire
3. Black wire and black wire
4. Skipped
5. Black wire with white stripe

The switch makes the lights work as they should.

[A383Wing]

I'm thinking you lost a ground somewhere, like maybe to the cluster itself....I lost the engine to firewall ground one year on my car...every time I stepped on the brake, the fuel gauge needle would go up

Bryan

[lukedukem]

You said that your missing the plastic plug on the harness, so the wires have to be plugged onto the terminals individually, so maybe the terminals got pushed together and are touching causing the short. So when you turn on the switch then it is giving some voltage to the rest of the circuit board
Do the dash lights come on too, those three for the gauges. I'm no expert

Luke

[MaximRecoil]

I'm thinking you lost a ground somewhere, like maybe to the cluster itself....I lost the engine to firewall ground one year on my car...every time I stepped on the brake, the fuel gauge needle would go up

Bryan

The cluster is grounded fine. In addition to grounding through the screws that hold it in place, I also made an extra ground wire for it a couple of years ago; a 14 gauge wire with ring terminals on both ends, secured to one of the screws on the cluster on one end and secured to the car chassis on the other end. There is perfect continuity (as tested with a meter) between the metal frame of the cluster and ground, even when the cluster is pulled out some.

The engine to firewall ground is fine too; I redid the grounds in the engine compartment a couple of years ago using new 4 gauge wire with copper lugs, sanded to bare metal where the lugs bolted to the engine block and the chassis, and I tested it all with a meter. No problem with grounds in the engine compartment at all.

You said that your missing the plastic plug on the harness, so the wires have to be plugged onto the terminals individually, so maybe the terminals got pushed together and are touching causing the short. So when you turn on the switch then it is giving some voltage to the rest of the circuit board
Do the dash lights come on too, those three for the gauges. I'm no expert

Luke

The terminals are not touching each other. The blade terminals on the switch are not bent, and the wires from the harness all have quick-disconnect terminals and are pushed on securely. The dash lights come on when the switch is in the parking lights or headlights position, as they are supposed to.

When I put the cluster back in, I didn't plug in the two wires that go to the windshield washer switch, because there is no windshield washer in the car anyway. I don't know if that could have anything to do with it or not.

[A383Wing]

still sounds like you lost a ground somewhere and power is back-feeding though the cluster and through the gauges

Bryan

[MaximRecoil]

Well, it seems to have fixed itself. Yesterday morning I turned on the parking lights, waited, and the gauges didn't do anything. Then I turned on the headlights and waited, and the gauges still didn't do anything. I hadn't done anything to the car since the day before when the gauges were coming on with the lights, I hadn't even wiggled or banged on anything, I hadn't even started the car.

Then I decided to plug in the two brown wires that go to the windshield washer switch, just for good measure.

If it was caused by a bad ground, what ground could it be? The cluster itself is grounded solidly/perfectly. As I mentioned before, in addition to being grounded through the screws that secure it to the dash, a couple of years ago I added a ground wire with ring terminals that is screwed to the back of the metal frame of the cluster and bolted to a factory hole in the steel framework inside the dash. 

[A383Wing]

loose pins on cluster circuit boards can cause this issue as well. Ground is through one or 2 of the pins....right now it seems like connection is good..but it ain't gonna fix itself

Bryan

[jlatessa]

If you ever have the cluster out again, you should solder the pins and the female
connectors for the voltage limiter as they're only mechanically clinched.

Good insurance....Joe

[A383Wing]

I soldered all the pins in my 66 Dart and the 70 Charger...some were loose, some weren't... I just did all of 'em at once

Bryan

[MaximRecoil]

This gauge/headlight issue keeps coming back.

A couple of weeks ago I was driving in the rain with my headlights on and the temperature gauge needle started moving all the way to the right. Thinking the car was overheating, I pulled over and lifted the hood; everything seemed fine. Then I remembered this gauge/headlight issue, turned off the headlights, and the temperature needle went down to normal. I turned the headlights back on and the needle started rising again.

At the time I had a solid state 5V regulator in there, the kind you build from wires, a capacitor, a heat sink, and the 5V regulator "chip", and I thought maybe that had something to do with it. I've been wanting an RTE limiter for a while anyway, so a few days ago I removed the homemade regulator setup and plugged in the RTE limiter. It seemed to work good; I tested it with both the headlights on and off, and the gauges read the same both ways. However, it wasn't raining when I tested it.

Today it was raining hard, and I turned on the lights and all the gauges went dead. I shut off the lights and the gauges instantly came to life again. I turned the lights back on and the gauges went dead, and so on. I suspect this is a case of the built-in protection function of the RTE limiter kicking in, so the gauges shut off instead of heading to the right.

So it seems to have something to do with the rain. Some water does leak onto my front floorboards if it rains long and/or hard enough, so I suppose some water could be getting on the back of the instrument cluster. But what do the headlights have to do with anything? If water is shorting something out, you'd think it would be a problem whether the headlights are on or not. How does the headlight circuitry affect or tie into the circuitry for the gauges?

Also, rain just seems to make the issue worse; the issue is still there when it is dry, it just isn't very noticeable. When the homemade regulator setup was in there, turning the headlights on when it wasn't raining would cause the fuel and temperature needles to go up slightly. With the RTE limiter, turning on the headlights doesn't have any effect on the gauges when it isn't raining.

By the way, all the pins and the 3 regulator connector slots are soldered, and they have been since before I first started this thread a year ago, and the cluster is well-grounded.

[MaximRecoil]

I got this fixed, for now. It was still screwed up the next day, even though the rain had stopped the previous night and it had been sunny all day, so I took the cluster out, resoldered one of those connector pins for the gauges that seemed loose, the one closest to the voltage limiter (the old solder joint probably cracked from me plugging in and unplugging that connector several times since I first soldered it), added a lot more solder to the female connector slots for the voltage limiter, and I placed a dedicated ground wire directly under the lower left grounding screw on the PCB.

I'll have to wait until it rains again to see if it stays fixed. If it doesn't, I'm going to redesign it. That 5-pin connector for the gauges is a terrible design; just plugging it in and having the plug and weight of the harness always hanging from it puts a lot of stress on it, and those pins only have a pressed in connection from the factory. Sure, you can add solder, but there are no actual pads for solder, so you have to scrape away solder mask, but the pins are going into very narrow traces, so it doesn't give you much to solder to regardless. Plus it is only a single-sided PCB, so you can't solder on the underside for reinforcement. 

If I end up redesigning it, I'm going to ignore those pins, and find some good points to solder wires to, 5 wires, one to take the place of each pin (I could just use ring terminals on 3 of the wires; clamp them down with the nuts that screw onto the posts of the gauges), and then terminate the 5 wires with a Molex connector. Then I'll splice the other half of the Molex connector into the wiring harness, in parallel with the original plastic connector (I could just cut the original connector off and reterminate the wires with the Molex connector, but I'd rather leave it there). Plug the Molex connectors together and I'll have a long-term connection that isn't constantly under stress, or ever under any significant stress at all, for that matter: