c_span

In the further adventures of "doing one job, he finds another" with my 928.. I think I need a bit of help, please.

I recently replaced the hood switch in my 91 S4 as this was giving me false alarms (and unstable continuity readings with my multimeter). Fitted a new one and, touch wood, no more false alarms.

Now "WYIT" I decided to upgrade the under-hood light to an LED cluster. Fitted it and now the engine bay is bathed in a pool of bright light.. cool.. but.. when I press the hood switch, the light doesn't turn off completely, it just goes very dim.

When I put a multimeter across the light terminals, it seems to go from 12v -> 6v with the hood switch "closed". I'm not especially bothered about the current draw, especially as it's LED, but it doesn't seem right to me?

Grateful for any suggestions what might be causing this..

WallyP

The hood switch also acts as a sensor for the alarm system - possible cross-talk?

Nicole

I just found out I have exactly the same situation - SOMETIMES.

To be exact: Last week, when my car was in the shop, the mechanic pointed out the behavior that you describe, and demonstrated it to me.

I was able to duplicate again at Sharktoberfest. So I bought a new switch.

When I got home tonight, I thought I'd replace the switch real quick, but decided to test the old switch one more time: This time it turned the light off completely.

I'm just as puzzled about this as you are. The only difference is that tonight, the engine is still warm (and the engine compartment has been well heated all day).

I don't have or had any false alarms, but I had a broken switch before, which turned the light on again once you depressed it all the way down.

Weird...

c_span

Thanks for getting back to me.

So could it be "They All Do That, Sir".. or yours is broke the same as mine!

Pretty sure the switch is OK, it's brand new. And ISTR I put my multimeter on it just to see it went zero resistance/infinite resistance. So where does the wiring go from there?

Alan

On the S4 cars this switch does 3 things:

1) Turns on the hood light (regardless of light switch)
2) Acts as a sensor for the hood alarm
3) Controls the disable of the fan "after-running"

From what Nicole notes it may be related to item #3

I'm not sure quite why this would be - but normally the bulb acts as the pull-up on this line for the feeds to the alarm and to the fan controller. With an LED this will not pull all the way up to battery voltage when the 'bulb' is off - it will be probably 1.4v below the battery voltage at least - if its a polarity independant LED bulb then it likely is an even bigger difference due to extra series diodes.

A way to work around this may be a resisitor in parallel to the LED bulb - its likely to NOT need to replicate the original bulb on resistance - it can probably be a few hundred ohms - so will not add much additional current draw when on - however you'd need to test this out...

Alan

c_span

Interesting.. so it could be because the LED cluster is higher resistance, so therefore I would need to add a resistance to bring the total "across the terminals" nearer to what the original bulb would have been?

That would make sense, because a quick measurement on the festoon bulb shows a resistance ~7 Ohms

From the spec sheet of the LEDs, it says 180mA, which is presumably @ 12V so that's ~67 Ohms?

But if I put a couple hundred ohms on there in parallel, won't that mean the overall resistance of the system is higher than the LEDs (and even further from the incandescent bulb)?

(sorry it's been a while since I did V=IR)

Alan

No anything in parallel will make the resistance lower.

In this case its likely the fact that there is a significant voltage drop across the LED (vs none across the bulb when off). If is is related to the absolute resistance it would mean you have a leakage path there...

I'm not suggesting matching the original bulb on resistance - in this case this would be a pretty worthless exercise right? (when cold (off) the bulb resistance is far far less anyway... prob <1 ohm)

I'd experiment with a parallel resistor of maybe 200ohms to start - go down to 100 ohms if need be...

I'm still surprised by your measurement of 12v vs 6v though... its possible you have another problem here...

Personally my strategy is to stay incandescent and to ensure I have a way to switch it off when desired...

Alan

c_span

Just to update, a 220 Ohm resistor across the terminals seems to work fine. Light now fully off when switch is closed and nice and bright when open

Thanks very much for the help

Nicole

I'm not sure I understand how this resistor is wired and where. Do you have any pictures?

Wouldn't a resistor draw additional current - which would be worse than having the light glow a bit?

I'm a bit confused here...

c_span

the resistor just goes across the two terminals (the flat blades of the light connector have 2 holes in them, that you can poke each leg of the resistor through and then push the connector on to "hold" it all in place).



As i understand it, the resistor convinces the circuitry to give you effectively zero volts at the socket (as opposed to the slight voltage that was there before, causing the light to glow dimly) - and once you're at zero volts it won't be drawing any current.

Hope the above makes sense!