Which model do you have? My 67 has a dash switch to allow/defeat the instrument lights coming on with the headlights. Later cars had a ‘dimmer’ arrangement, I think. You also mention just the tach & speedo - do the lights on the other gauges work? They are probably on the same fuse/supply, so if some work and some don’t, it’s probably bulb failure.

 

Yes, a really weird Italian thing, 60s Fiats have a switch to turn the dash lights off too.

I get the idea of a dimmer (but not on an Alfa as they're dim enough already!) but I can't imagine any circumstances where I'd want them switched off.

 

It’s a 72 GTV 2000. The indicator lights work, such as parking brake indicator, turn signal, high/low beam indicator, alternator light, etc. Swapping the bulbs (and socket thing) they work in other locations, so I know it’s not the bulb or socket.

EDIT: So is that probably the intention of the "dimmer rheostat"? It controls the brightness of the illumination, and then the regardless of brightness setting the lights are turned on/off with the headlights?

 

That's the way it works on a Spider, headlights must be on and then the instrument brightness is controlled by the dimmer. You could bypass the dimmer to see if that makes a difference.

 

That's the way it works on a Spider, headlights must be on

Actually, having either the headlights or parking lights on will illuminate the dash lights.

You could bypass the dimmer to see if that makes a difference

And the dimmer may well be where the problem lies; those things aren't particularly reliable.

 

Yeah, I've tried that (bypassing it straight to ground, I think that makes sense, but clearly I'm no electrician), immediately blows the #5 fuse when the headlights are turned on. This seems generally unlikely, but what are the chance the board in the instrument cluster has issues? It seems like something pretty failure proof, but as I fiddled with the car today it seems to be pointing that way. I can pull the #7 fuse and the high beam indicator still works... that should cut off all power to the wire that powers that light (A3 on the plug in my diagram). I started this whole thing to resolve some vague gremlins, including an oddly behaving fuel gauge (sender is confirmed good) and a temperature gauge that reacts to turning on the headlights. Not trying to introduce a ton of variables or go after multiple issues at once, but I'm increasingly suspicious they could all be caused by the same root issue.

 

Yeah, I've tried that (bypassing it straight to ground, I think that makes sense, but clearly I'm no electrician), immediately blows the #5 fuse when the headlights are turned on.

You're right!

By "bypassing it" 6alfas meant interconnecting the two wires that attach to the rheostat; not connecting one of them to ground.

 

The rheostat just introduces resistance and drops the current to the light bulbs as more resistance is applied. To bypass it you would connect the three wires together (the two marked G are probably on the same spade already) so the switch is no longer in line with the circuit. If you have a multimeter, you could remove the switch and check the resistance across the two connectors. If it is a dead open at all times (infinite resistance), you know the switch is bad. On a good switch the resistance should vary from a dead short (0 ohms) to something more than that that if working properly.

 

There are three bulbs to illuminate the instruments and now would be a good time to replace them with led bulbs and actually see the instruments at night. Some are polarity sensitive and need to be inserted a certain way. It's a 50-50 chance they will work the first time you put them in, if not just rotate 180 degrees. Also with a led in the low fuel indicator it will be easily seen even during the day. One place I'd not use a led is in the green parking/headlight slot as they are distractingly bright at night. There are two spaded wires that go to the back of the cluster, the pink one is for the hand throttle indicator and the other is a black ground wire for the cluster deviecs so make sure they are conntected to the correct spades on the back of the cluster. You can make a short wire with male spade connectors on each end and jump across the dimmer. I've never really understood why there is a dimmer in the first place since the instrument ligths are so dim to begin with. Hope that helps.

 

Guessing here… but maybe it’s coz the headlights were so poor (they all were back then, remember the healthy market for driving lights?…) that in some circumstances you wanted absolutely no light pollution in order to maximise forward vision, especially beyond the lit city streets.

 

Ha - I read that as "lit cigarettes", which in 70s Italy was probably also correct.

 

Looping back here just in case anyone is ever searching and finds it. It ended up being a ground issue - which I was suspicious of but had a misconception that was stopping me from solving. Previous owner removed the ground spade from the board and only left the spade for the throttle light, which I incorrectly assumed was the ground. The small bolts with nuts on the board are grounds, so I just put a spade on the one that was the original ground location, and everything started working. Explains why I was seeing such inconsistent behavior, ended up being a simple solution. LEDs installed, I'll never even know how bad the original lights were. And now a few other things are working too!

 

Worth mentioning : on the US 2000 GTV, the dimmer does not control the instrument lights which are directly connected to the parking lamp switch. It controls the mandatory small bulbs that try to illuminate various switches such as the wiper or heater fan.