Headlights won't turn on…where are my electrical people?

[IBallEngineer]

I tested ground on Terminal 6 with DRL relay connected, no ground unless it was on FLASH. Terminal 7 with DRL relay disconnected also had no ground unless it was on FLASH.

Jumpering at Connector C13 at the stalk between Terminal 8 and 13 did not allow LOW to come on or constant HIGH, but FLASH still worked.

Well that certainly seems to condemn the switch. Could you confirm that shorting terminal 13 at the stalk to ground turns on the headlights? And as long as you are there, ground terminal 7, that should activate Hi beams. If you still don't get lo beams with 13 grounded, ground terminal 8, this will bypass the cluster.

I'm pretty sure if you ground 13 or 7 your lights will work, otherwise you have a bad relay AND a bad switch. But we can test the switch independently of the relay if this doesn't work.

:couch2: MORE PEANUTS PLEASE

 

[TheDurk]

Well that certainly seems to condemn the switch. Could you confirm that shorting terminal 13 at the stalk to ground turns on the headlights? And as long as you are there, ground terminal 7, that should activate Hi beams. If you still don't get lo beams with 13 grounded, ground terminal 8, this will bypass the cluster.

I'm pretty sure if you ground 13 or 7 your lights will work, otherwise you have a bad relay AND a bad switch. But we can test the switch independently of the relay if this doesn't work.

:couch2: MORE PEANUTS PLEASE

I didn't see yours until I had finished editing mine. We are on the exact same wave length. I'll get us some popcorn.

 

[Fishwerks]

Well that certainly seems to condemn the switch. Could you confirm that shorting terminal 13 at the stalk to ground turns on the headlights? And as long as you are there, ground terminal 7, that should activate Hi beams. If you still don't get lo beams with 13 grounded, ground terminal 8, this will bypass the cluster.

I'm pretty sure if you ground 13 or 7 your lights will work, otherwise you have a bad relay AND a bad switch. But we can test the switch independently of the relay if this doesn't work.

:couch2: MORE PEANUTS PLEASE

Any advice on how to short to ground? Is hooking my test probe alligator clip to a ground, then getting the probe tip into the terminal an appropriate way? The way I am understanding it, these terminals on the stalk should have power?

layball:

 

[TheDurk]

Any advice on how to short to ground? Is hooking my test probe alligator clip to a ground, then getting the probe tip into the terminal an appropriate way? The way I am understanding it, these terminals on the stalk should have power?

layball:

Not really. Terminal 16 is just a ground. That why I suggested you use it. A good chassis ground of unpainted metal will have the same result. Terminal 13 has a relay coil, a diode and a switch between it and 12V+, but it's a ground when the stalk is on HEAD and the stalk is good so no power should be measured there.

 

[rideexileex]

If I remember correctly, I think that Flash is on a different circuit than the normal highbeams, lowbeams, etc... Would need to check the wiring diagrams. Correct me if I'm wrong - I'm working on 3 years of memory

 

[TheDurk]

If I remember correctly, I think that Flash is on a different circuit than the normal highbeams, lowbeams, etc... Would need to check the wiring diagrams. Correct me if I'm wrong - I'm working on 3 years of memory

Not really. It takes a separate path through the stalk obviously, by-passes the cluster auto-off cut-out and then it's the same through several components.

That's why were pretty sure if it's not the cluster cut-out, then it's the stalk switch. The main DRL relay remains a suspect just because it's involved in everything.

 

[rideexileex]

Not really. It takes a separate path through the stalk obviously, by-passes the cluster auto-off cut-out and then it's the same through several components.

That's why were pretty sure if it's not the cluster cut-out, then it's the stalk switch. The main DRL relay remains a suspect just because it's involved in everything.

That's what I meant - different circuit as in not going through the auto-off function. +1 memory.

I would skip troubleshooting the switch until the very last item... The switch is a mechanical object, and is the least likely to fail out of anything. If wiring changes were made, its most likely to have affected the 'computers' / sensitive electronics. A fuse would have blown before the switch would melt.

 

[IBallEngineer]

I would skip troubleshooting the switch until the very last item... The switch is a mechanical object, and is the least likely to fail out of anything. .

Not sure if you've read this entire thread, but we are there! No ground to the DRL, no ground out to the cluster. 99.99998% sure it's the switch.

For the record in a DNA test that is a 100% match!

 

[rideexileex]

I'll dig for my diagrams tonight... IRRC, the ground comes from the cluster to the switch... When the cluster fails, it acts like an open

 

[IBallEngineer]

I'll dig for my diagrams tonight... IRRC, the ground comes from the cluster to the switch... When the cluster fails, it acts like an open

No you got it right, the DRL receives a ground trigger from the switch, but through the cluster. The OP jumpered terminals 8 and 13 at the stalk, effectively bypassing the cluster and still had no effect. But we still need to bypass the switch, with and with out the cluster in the circuit.

:couch2:

 

[TheDurk]

Not sure if you've read this entire thread, but we are there! No ground to the DRL, no ground out to the cluster. 99.99998% sure it's the switch.

For the record in a DNA test that is a 100% match!

Agree the switch is bad, but still some relevant chance of another fail elsewhere until we get those lights to come on.

 

[TheDurk]

That's what I meant - different circuit as in not going through the auto-off function. +1 memory.

I would skip troubleshooting the switch until the very last item... The switch is a mechanical object, and is the least likely to fail out of anything. If wiring changes were made, its most likely to have affected the 'computers' / sensitive electronics. A fuse would have blown before the switch would melt.

Not necessarily. The stalk is designed to handle max current of a relay coil. The system is fused at 10A and 7.5A and that is much higher than the stalk is meant to take. Plus these stalks fail with reasonable frequency at this age without any help.

Even BMW 7-series stalks like to die at 15 years. Ask me how I know.

 

[Grenvilleter]

Still watch'in and bringing more beer.

Remember 1st post -

" I hit the switch, smelled like hot plastic (not a good sign) almost immediately and the fogs didn't turn on so I switched it off."

 

[Fishwerks]

Not really. Terminal 16 is just a ground. That why I suggested you use it. A good chassis ground of unpainted metal will have the same result. Terminal 13 has a relay coil, a diode and a switch between it and 12V+, but it's a ground when the stalk is on HEAD and the stalk is good so no power should be measured there.

With stalk on HEAD I did not measure any power between Terminal 16 and Terminal 13. With stalk on OFF I did not measure any power either. Should there be power there with stalk on OFF?

 

[TheDurk]

With stalk on HEAD I did not measure any power between Terminal 16 and Terminal 13. With stalk on OFF I did not measure any power either. Should there be power there with stalk on OFF?

No, not then either because on stalk=OFF ground is cut off from the HEAD relay which opens and removes the 12V+ from the other end of the circuit. So no power to measure.

How are we coming on the stalk test?

 

[Fishwerks]

With terminal 8 grounded at the stalk the LOWS and HIGH both work.

 

[TheDurk]

With terminal 8 grounded at the stalk the LOWS and HIGH both work.

What about when you ground 13? We need that to know if it is the stalk or the cluster cut-out. But excellent progress!!

 

[IBallEngineer]

What about when you ground 13? We need that to know if it is the stalk or the cluster cut-out. But excellent progress!!

13 is the decider, although if grounding 13 doesn't turn on lamps we know the cluster is bad, we still can't rule out the switch. We would then have to TEST for ground at 13.

Luckily I got more peanuts.

:couch2:

 

[Fishwerks]

13 is red with yellow stripe correct? Not yellow with red stripe?

 

[IBallEngineer]

13 is red with yellow stripe correct? Not yellow with red stripe?

Correct, it should be the only red and yellow one.

DOH, there I go thinking before I read. Yellow with red is there, but not what you want.