Help, Headlights won't turn off! SOLVED!

[Mars Attacks]

nah its both headlights i'm afraid and they are on dipped headlights as opposed to side lights as well so its very strange, have had to unplug the clusters again today so that i don't have to keep disconnecting the battery.

not sure what else to try, can only think it must be an earthing problem somewhere. more of a mk1 / mk2 man myself so not sure where all the earth straps are on a mk3.

[javalin]

have you tried taking the bulbs out of the headlight's one, by one?  Can't remember exactly how it works, but another bulb (or broken wire) can end up providing an "earth" path for another.

You can check ground levels and voltage flows with a multimeter.  £5 Maplins or so.  Ground is checked using the volts scale and putting a probe on the neg battery post, and the neg of whatever your testing - should read < 0.5v

J

[Mars Attacks]

cheers i'll try that later thanks.

[AudiA8Quattro]

If the lights stay on when you disconnect the switch then you must have a constant earth.
I think the lights have a constant live feed.

[Mars Attacks]

yeah that makes sense, just need to find it, and if its right that a defective headlight bulb might cause this then we might be getting closer to resolving it! hopefully. cheers

[jezza16v]

The headlights are switched via a relay aren't they? The relay contacts may have welded themselves together? Pull the relay and see if they go off. (I don't know which relay it is though)

[Mars Attacks]

i didn't think they had a relay, but if anyone knows if they do, the number would be helpful please.

[Mars Attacks]

ok so over lunch i've been out and had a quick look at the car, i removed the bulbs one by one but everything looked fine and the lights remain on when i reconnect.

One thing did occur to me, perhaps they have been uprated in the past (we've had the car for 2 years and i've never looked at the headlamps as they have never been a problem.) and perhaps a relay was added to the system which might have stuck open.

In that case, where would it be? with the other relays on the fuse box, and if so which slot would it occupy? any ideas if anyone has uprated the headlamps themselves?

Thanks again, sorry i'm useless i know!

[javalin]

hiya

seems like there isn't a relay which surprises me.

http://faculty.ccp.edu/faculty/dreed/Campingart/jettatech/techset/index.htm

I think the volt meter is your friend on this one, you need to trace the power back from the lights and find which bit is causing them to stay on.  For example:  if you find that the headlight switch is always live both sides, or that removing the main-beam blub turns them off, relays, etc, then you've found the problem.

>One thing did occur to me, perhaps they have been uprated in the past (we've had the car for 2 years and i've never looked at the headlamps as they have never been a problem.) and perhaps a relay was added to the system which might have stuck open.
a good thought.  I think they're usually close to the headlights - so follow the wires back and look for a small box on the wires. (Good opportunity to look for breaks, cuts, damage etc whilst you're there).

Thinking about this more.  If it were me - i'd start at one end and work forwards. i.e.  Pull the headlight switch and confirm (using voltmeter, or test lamp) that power is deffo off when switch is off.  Then i'd try to follow the power to the headlight - something is powering it just to find out what.  I'd check the power feeds to the light clusters, i.e. that the wires are live when the switches are live in the cabin - again start with all off, and power one by one.

A Haynes manual would help with the wiring diagrams too.

Keep us updated...

James

[Mars Attacks]

thanks again, yeah will have to pick up a multimeter and figure out the source of the problem or i could be in there all day trying to guess what could be the problem. cheers

update - i've been out to the car again and had a good look round the wiring and no sign of any inline relays so i'll eliminate that and look for the source of the power via a multimeter.