Headlight Voltage Weirdness

[sped372]

I was driving the Ghia the other day when I noticed the bundle of wires under the dash (from the column up into the trunk) was hot! I turned off the lights and the bundle cooled back down, slowly. We rarely drive the car at night (or with the lights on) so I presume this has been going on since we've owned it. I started digging into the electrical system this evening and noticed some peculiarities. The system looks pretty much unmolested from inside the trunk. I will end up cutting open the bundle that was hot (it has a very nice electrical wrap job on it over the entire length of the bundle) but I have not gotten that far yet. I wanted to do a preliminary investigation first.

Lights all work, so do the blinkers. They have always seemed somewhat dim, but not much moreso than on the Bus or my old Beetle. Hi/Low relay works. Oil & Gen lights in the speedo are bright. Blinker indicator typically signals once, and then doesn't (the actual blinker lights continue to blink).

As I started tracing things my major observation is a voltage disparity between the front and back of the car, depending on whether the headlights are on or not. It doesn't seem correct to me.

Engine off, ignition key off, headlight switch off
Voltage directly at battery: 12.5
Voltage at + side of voltage regulator: 12.5
Voltage at 'always hot' terminal of fuse box: 12.5
Voltage at left headlight fuse: 0

Engine off, ignition key on, headlight switch off
Voltage directly at battery: 12.3
Voltage at + side of voltage regulator: 12.3
Voltage at 'always hot' terminal of fuse box: 12.3
Voltage at left headlight fuse: 0

Engine off, ignition key on, headlight switch on
Voltage directly at battery: 12.3
Voltage at + side of voltage regulator: 12.2
Voltage at 'always hot' terminal of fuse box: 9.5
Voltage at left headlight fuse: 8.7

I checked a few other points up front (at the headlight switch, at the hi/low relay, at the flasher relay) and they all get 'pulled down' whenever the headlights are on. I didn't start the engine but it didn't seem like it should make a difference since I have essentially full voltage at the battery regardless of whether the lights are on or not.

I think my next check will be to either unplug the headlights to see if eliminating the draw allows the voltage to 'stay up' or else dig into the bundle under the column. Probably the headlights first since it doesn't require surgery.

If anyone has experienced this or has any thoughts I'd love to hear them. I'll try to keep this posted updated with my findings for posterity.

[grandfatherjim]

Corroded connections can cause this kind of stuff: turning on headlights draws lots of current. Current must through through for example, the little pointed tips of fuses where they spring-load in their respective little holes.
A good place to start is to go through all those types of spots - anywhere there is a connection - and use very fine i.e. 600 or higher sandpaper to clean them up. Then apply a light layer of dielectric grease to prevent further issues.
Grounds tend to be particular culprits.
Only once these are dealt with, does it become possible to start troubleshooting.
Jim

[dingo]

Do a Volt-drop test with a meter on both positive and negative side of the lighting system..and you can pinpoint exactly where the current is meeting resistance (i.e. corrosion or bad contact)

[sped372]

Do a Volt-drop test with a meter on both positive and negative side of the lighting system..and you can pinpoint exactly where the current is meeting resistance (i.e. corrosion or bad contact)

Good thinking.

...use very fine i.e. 600 or higher sandpaper to clean them up. Then apply a light layer of dielectric grease to prevent further issues.

That's what I'm in the middle of right now. I wanted to start there and work through as many as I could before doing anything more "destructive" - some of those connections are a major PIA where they're located way up under the recess of the trunk!

[Amskeptic]

Do a Volt-drop test with a meter on both positive and negative side of the lighting system..and you can pinpoint exactly where the current is meeting resistance (i.e. corrosion or bad contact)

Good thinking.

...use very fine i.e. 600 or higher sandpaper to clean them up. Then apply a light layer of dielectric grease to prevent further issues.

That's what I'm in the middle of right now. I wanted to start there and work through as many as I could before doing anything more "destructive" - some of those connections are a major PIA where they're located way up under the recess of the trunk!

A ghia has a hand-built body that may have a little more difficulty collaring the electrons and getting them back to the battery. Make absolutely certain that all grounds and ground paths are shiny perfect/protected from oxidation. Every brown wire deserves as much care as any red or black wire. Body-to-floor pan must have good electrical path too. This entire system is only as good as its weakest link. Check especially the big wires and battery cables. Even an intact looking (+) cable from battery to starter can look lovely on the outside, carry 12 volts easily at low amps, and still be a corroded mess under the insulation near terminals.
Colin

[sped372]

I've only found one thing that has been incorrect so far, the 30 terminal on the headlight dimmer relay was connected to fuse #2 instead of fuse #10. I was excited when I found that because it seemed to fit the symptoms, since fuse #2 is powered by a thinner-gauge wire through the column (activated by the ignition switch) and at first I thought it was channeling all the power for both headlights. More research shows that terminal 30 on the relay only actually comes into play when the headlights are off and you momentarily pull the stalk back for a 'flash to pass' function. So, in those cases it was definitely overloading the column wire, but not while the headlights were on normally (that current comes from the black/white wire between the headlight switch and the dimmer relay, on terminal 56, not the wire on terminal 30).

I've continued to wire brush and grease all the connections and everything is slowly improving. Side marker lamps now work and the turn signals are no longer flaky (used to be very susceptible to bumping the bulbs in the socket and losing connection). The high beam indicator and even the defrost indicator in the speedo now work. The blinker indicator works *more* consistently, but still not 100% of the time.

I now have 11-11.5 volts to the headlights when they're on (used to be 9ish) so it's a big improvement. I haven't yet pulled the emergency flasher switch to do those connections yet. I'm still seeing a solid 0.5v drop across the headlight switch itself (red/30 to black/white) so there is obviously some resistance internal to it, not sure if it's possible to open it or not. I'm reluctant to replace it with a shoddy new part so I'll press on and see what other improvements can be made.

Also, I unwrapped the bundle under the column and every wire has a splice in it but they are all mapped correctly. Obviously the column was removed/replaced at some point. They're just butt-splice crimp connectors, though and they feel a little loose so I'll probably redo them and add some solder.

[grandfatherjim]

If the headlight switch is the same as on a 72 bus, it is possible, albeit finicky, to open and service it.

[Amskeptic]

More research shows that terminal 30 on the relay only actually comes into play when the headlights are off and you momentarily pull the stalk back for a 'flash to pass' function.

I don't believe VW had that feature until much later. Is that at US Ghia or an imported European?
Colin

[sped372]

I'm not sure, do Ghias have M-codes like buses? The 30 terminal threw me off for awhile because it isn't on any of the wiring diagrams for the US-spec models. I ended up figuring it out by cross referencing the non-US wiring diagrams and also through some forum hunting. The side-marker configuration in the front blinker assemblies matches the US-spec wiring diagram.

It looks like the currently available replacement dimmers have the 30 terminal, though, so it may have been swapped in later. It's not much newer than anything else, though, if at all. The wire to 30 looks period correct as well with terminals that match the other connections, but it is black unlike those shown by the non-US wiring diagrams. Dunno, have to do some more research.

[sped372]

do Ghias have M-codes like buses?

Supposedly they do but they aren't stamped on a plate like on buses, you have to send for your birth cert...
http://www.thesamba.com/vw/archives/info/mcodeskg.php
http://www.thesamba.com/vw/forum/viewto ... ight=codes

[Hippie]

The part of the harness that's getting hot first is where the bad (high electrical resistance connection) will be. The resistance causes the heat. There may be more than one in more than just one area though.

[sped372]

I had to leave the garage in frustration because I could not get the headlight switch escutcheon started. I can stand outside the driver's door and reach my left arm through the trunk and snake it up into place through the wiring to hold the switch in place, and also from the same position reach my right arm through the open window to hold the escutcheon. I can not see anything because my eyes are halfway between!

I think I need a helper unless someone has a helpful tip. Escutcheon tool on order but that probably won't help me get it started!

[hambone]

I have noticed that the older escutcheons have longer threads. I had to toss my new one because the treads were too "short" to engage, installing a flasher switch on a bus.

[sped372]

Well it's the same one I took off there a couple days ago so I know it "works" - now it's just a matter of getting the switch and the escutcheon all coaxial-happy-like in a confined space.

[Hippie]

If you think about it, access--or rather the lack of it--is the key roadblock of nearly any mechanical repair endeavor.
You must endevour to persevere.