I think I broke something . . . but it doesn't make sense (66 Newport)

[fury fan]

Is voltage supplied to the idiot lights in the same way as it is sent to the gauges?

Idiot lights should get 12v constant, the 5v regulator is to slow-down the response of the gauges. Otherwise the gas gauge needle would wiggle around when going around turns, stopping at a red light, etc.

 

[Michael E Mason]

So, last weekend while I had the panel that lives under the steering wheel removed so I could install my tachometer, I thought I'd try installing one of my LED bulbs in the instrument cluster. I reached up to unscrew the nearest light socket, pulled it out, put the LED in, screwed the socket back in and turned on the dash lights to see what it looked like. Nothing. I thought the orientation might be off so I pulled it out, reversed it, plugged it back in, turned the dash lights on again. Still nothing. Perplexed, I put the original bulb back in and went back to installing the tach.

After getting the tach installed, I was ready to start the engine. Turned the key to "on" and noticed that the OIL light failed to light up. Turns out this is the bulb I had been swapping out. Started the engine, checked the tach, everything seemed in order except no OIL light with engine stopped and key ON. I went under the hood and checked for voltage at the sending unit to see if it was just a bad bulb. Then I was really perplexed because there was no voltage getting to the sending unit anymore. I posted a question on this forum, futzed around with the test light, couldn't find the correct wire at the bulkhead because a PO had painted absolutely everything with a thick coat of black paint. Decided it wasn't worth continuing on with at the moment, and mentally postponed the examination until summer.

THEN I noticed that my fuel gauge wasn't doing anything anymore either. It wasn't reading correctly before, but at least the needle moved. Now, the needle is pegged on E and doesn't budge no matter what.

Today I had the car up on a lift and I checked for voltage getting to the fuel tank sending unit - nothing. I cleaned off the ground strap real good, also wire brushed the metal fuel line and noticed that the sending unit looks to have been replaced in the last 5 years or so. Replaced the ground strap and checked again - still nothing coming to the tank.

SUMMARY - I changed out and replaced back to original the OIL light in the dash and now the OIL light doesn't work anymore and neither does my fuel gauge. No voltage is getting to either sending unit. What could I have done?

TIA

When I pulled out my cluster to see why my fuel gauge worked intermittently and my dash lights weren't working, I found three of the pins that the wiring harness plug onto broken loose from the circuit board. So while I had it out, I soldered the three pins back on and soldered every loose connection on the circuit board and also used emery cloth to gently sand all the connection points for the bulbs. When I was done, everything worked perfectly. I also developed a short in the rear brake lights later on that turned out to be the wiring harness laying on a sharp piece of metal behind the left kick panel. Hope all this info helps!

 

[rags]

Idiot lights should get 12v constant

and the oil and temp switches complete the circuits to ground. the switches can be disconnected and the circuits can be tested with a test light to ground. then the test light would complete the circuit to ground. two bulbs in series. both would light depending on their wattage.

 

[darth_linux]

When I was evaluating LEDs in a 65-66 cluster (here), I found that the bulb sockets were VERY finicky about how much to twist them vs the LED working. And the polarity, too, of course.

can you test all of the lights for functionality before putting the cluster back in place? I'm thinking not because of the electrical plugin...

 

[fury fan]

You can, you just need to look at the traces and figure out which pins feed which lights, then use jumper wires in the correct locations.
That is how I discovered how finicky the LEDs/sockets were.

What @rags and @Michael E Mason said is on-target too.

I did not use the emery cloth to clean the traces, it surely would've helped.
My exercise was on a spare cluster, to figure out what LEDs/colors to use in each location, so no need to clean the traces at that stage.
When I go to put LEDs in the car, I surely would remove the cluster from the car, check the pin connections, and do some trace-cleaning at the sockets.

 

[Gerald Morris]

When I was evaluating LEDs in a 65-66 cluster (here), I found that the bulb sockets were VERY finicky about how much to twist them vs the LED working. And the polarity, too, of course.

Ditto. Some LED 194s DO work on dash lights. I use a couple right now on Gertrude, but only a couple. Polarity matters w LEDs, when its DC. Best to shop for auto specific LEDs if you want to illuminate the dash. Bayonet socket sorts forgive a lot more. In such locations, LEDs really do well.

 

[Gerald Morris]

can you test all of the lights for functionality before putting the cluster back in place? I'm thinking not because of the electrical plugin...

Yes. Just hook the board up to a good power source, observing polarity carefully. Even that perverse 1965-66 board can be harnessed up. Use some clips, CAREFULLY to avoid damaging the old copper plated circuit paths. You might use copper foil in spots where the old stuff has oxidized or otherwise worn away. DO solder carefully if you foil a circuit.

 

[rags]

on my '65, the issue was the ground points for the circuit board. it grounds at its attaching points to the cluster nacelle. once a year i would have to reach up with an open end wrench and tighten the nuts. symptom was that the directional indicators wouldn't work correctly. both would flash dimly when one was activated. upside is that for testing, negative goes to the the nacelle or any of the circuit board attaching points. could positive go to any of the terminals? i'm thinking yes. they would be one each for left and right turn signal, oil light, temp light, cluster lighting, power for the voltage limiter. six terminals total? could it be that easy? i forgot. keep in mind that an AC power source in the correct voltage range can be used to test lighting. an LED can be powered with AC but it will only be lit half the time, 30 times per second on 60 cycle AC.

 

[Gerald Morris]

on my '65, the issue was the ground points for the circuit board. it grounds at its attaching points to the cluster nacelle. once a year i would have to reach up with an open end wrench and tighten the nuts. symptom was that the directional indicators wouldn't work correctly. both would flash dimly when one was activated. upside is that for testing, negative goes to the the nacelle or any of the circuit board attaching points. could positive go to any of the terminals? i'm thinking yes. they would be one each for left and right turn signal, oil light, temp light, cluster lighting, power for the voltage limiter. six terminals total? could it be that easy? i forgot. keep in mind that an AC power source in the correct voltage range can be used to test lighting. an LED can be powered with AC but it will only be lit half the time, 30 times per second on 60 cycle AC.

Yep! I'd say pick one of those ground points, stick a ring terminal on top of it, solder it onto the board, and have enough pigtail on that ring terminal to make it to the firewall, where one could bond it through to a jumper also connected to the battery. I tun a bonding jumper across my engine block to pick up the coil, distributor, alternator, then on to the firewall. Painted and rusted metal surfaces conduct poorly, and a few good bonding jumpers run around your 50-60 yr old C body helps things quite a bit.

 

[darth_linux]

You can, you just need to look at the traces and figure out which pins feed which lights, then use jumper wires in the correct locations.
That is how I discovered how finicky the LEDs/sockets were.

What @rags and @Michael E Mason said is on-target too.

I did not use the emery cloth to clean the traces, it surely would've helped.
My exercise was on a spare cluster, to figure out what LEDs/colors to use in each location, so no need to clean the traces at that stage.
When I go to put LEDs in the car, I surely would remove the cluster from the car, check the pin connections, and do some trace-cleaning at the sockets.

I was able to get all of my dash illumination bulbs to light by applying 12v to the correct pin of the "pin ring" but I couldn't get any of the 'IDIOT" lights to light up with this method, even though I didn't install LEDs and the incandescent 194 bulbs are known to work. Not sure why this would be the case since I can get the bulbs to light by removing the socket and testing it that way, but not once installed.

 

[D Cluley]

Maybe you know this, but the idiot lights ground through the sending unit, not the housing of the cluster, so to test you need to apply ground and power to the correct 2 pins in the connector.

I've also had issues with some corrosion on the circuit board so the little ears on the plastic bulb socket didn't make a good connection with the circuit board. Sometimes twisting the socket a little while powered up will make it light.

 

[darth_linux]

Maybe you know this, but the idiot lights ground through the sending unit, not the housing of the cluster, so to test you need to apply ground and power to the correct 2 pins in the connector.

I've also had issues with some corrosion on the circuit board so the little ears on the plastic bulb socket didn't make a good connection with the circuit board. Sometimes twisting the socket a little while powered up will make it light.

Actually, I did not know this, but it makes perfect sense after my tests and observations with my instrument cluster. Now that I'm looking more carefully, I do see that the oil light and the e-brake pedal light share a common pin for ground with the hot/cold ground trace. The high-beam light has it's own 12V pin and shares ground with the illumination trace. Each of those lights also have their own pin for 12v power.

Thank you!

 

[darth_linux]

So, last weekend while I had the panel that lives under the steering wheel removed so I could install my tachometer, I thought I'd try installing one of my LED bulbs in the instrument cluster. I reached up to unscrew the nearest light socket, pulled it out, put the LED in, screwed the socket back in and turned on the dash lights to see what it looked like. Nothing. I thought the orientation might be off so I pulled it out, reversed it, plugged it back in, turned the dash lights on again. Still nothing. Perplexed, I put the original bulb back in and went back to installing the tach.

After getting the tach installed, I was ready to start the engine. Turned the key to "on" and noticed that the OIL light failed to light up. Turns out this is the bulb I had been swapping out. Started the engine, checked the tach, everything seemed in order except no OIL light with engine stopped and key ON. I went under the hood and checked for voltage at the sending unit to see if it was just a bad bulb. Then I was really perplexed because there was no voltage getting to the sending unit anymore. I posted a question on this forum, futzed around with the test light, couldn't find the correct wire at the bulkhead because a PO had painted absolutely everything with a thick coat of black paint. Decided it wasn't worth continuing on with at the moment, and mentally postponed the examination until summer.

THEN I noticed that my fuel gauge wasn't doing anything anymore either. It wasn't reading correctly before, but at least the needle moved. Now, the needle is pegged on E and doesn't budge no matter what.

Today I had the car up on a lift and I checked for voltage getting to the fuel tank sending unit - nothing. I cleaned off the ground strap real good, also wire brushed the metal fuel line and noticed that the sending unit looks to have been replaced in the last 5 years or so. Replaced the ground strap and checked again - still nothing coming to the tank.

SUMMARY - I changed out and replaced back to original the OIL light in the dash and now the OIL light doesn't work anymore and neither does my fuel gauge. No voltage is getting to either sending unit. What could I have done?

TIA

UPDATE:
After removing the instrument cluster, replacing all of the bulbs with LEDs, thoroughly testing the fuel gauge, replacing the voltage limiter (original was still good) and lubing the speedometer, here's what I think I did.

I think I bumped the 3 pin connection (orange, black, dark blue) that gives 12v to the lower gauge lights, voltage limiter input and voltage limiter output that goes to the fuel tank and broke pin number three, which I discovered was loose/broken upon disassembly and I have subsequently resoldered and repaired. That would explain the fuel gauge suddenly being frozen.

The oil light not working was probably just a dirty connection where the bulb socket screws into the circuit board.

Both things are working again now, but the fuel gauge still reads way low. Probably time for a new sending unit.

 

[cbarge]

but the fuel gauge still reads way low. Probably time for a new sending unit.

Dont jump the gun just yet.
The new electronic voltage limiters tsjes a few minutes to operate properly..
Who's to say it was reading too high before?
Besides, the calibraton of new sending units aee known to be way off.
The best and easiest way to have a bang on ges guage is use a meter match.
Other members here use it andI have one as well.
Fuel Sender calibration using Meter Match

 

[darth_linux]

Dont jump the gun just yet.
The new electronic voltage limiters tsjes a few minutes to operate properly..
Who's to say it was reading too high before?
Besides, the calibraton of new sending units aee known to be way off.
The best and easiest way to have a bang on ges guage is use a meter match.
Other members here use it andI have one as well.
Fuel Sender calibration using Meter Match

I already confirmed my voltage limiter is working correctly, as well as is the fuel gauge itself, and all of the circuits on the circuit board. Also, I checked the wiring back to the tank as well as the ground strap. All in good shape. The only thing left is the sending unit.

 

[cbarge]

I already confirmed my voltage limiter is working correctly, as well as is the fuel gauge itself, and all of the circuits on the circuit board. Also, I checked the wiring back to the tank as well as the ground strap. All in good shape. The only thing left is the sending unit.

The meter match calibrates/compensates for an inaccurate sending unit.
So you said the guage is working..therefore ditto the sender..just not 100% accurate.
I would do the easy way out and install the meter match before changing the sending unit.
TechnoVersions - MeterMatch for Analog Gauge Correction
Hope this helps

 

[darth_linux]

The meter match calibrates/compensates for an inaccurate sending unit.
So you said the guage is working..therefore ditto the sender..just not 100% accurate.
I would do the easy way out and install the meter match before changing the sending unit.
TechnoVersions - MeterMatch for Analog Gauge Correction
Hope this helps

This is a good idea, if in fact the sending unit is not faulty to begin with. My suspicion is that I have a cheap aftermarket sender that has a float that has "sank", hence always showing near empty even when full.

Once I pull the sending unit and test it, I'll either replace it with new, or if my current sending unit is good but just "off," I'll take another look at the MeterMatch unit.

Thanks for the tip!

 

[Newport 664door]

Is voltage supplied to the idiot lights in the same way as it is sent to the gauges?

I'll add my 2 cents, when my ammeter wires burned up( that's another story) I went under the dash to pull the 2 burned black and red wires, there is a 3 prong connector also attached by the ammeter. These wires supply power to fuel gauge and interior lighting. You may have disturbed that connector