borland

After replacing the incandescent brake lights with LED bulbs, you’re cruise control may no longer function. This is because the cruise control amplifier on our 928’s sources an electrical ground through any of the brake light bulb filaments. Because amplifier sinks such a small amount of current, the LED bulbs don’t look electrically grounded. For safety reasons, the amplifier also deactivates the cruise function when the brakes are applied because it then sees 12 volts instead of ground.

You can remedy this problem by either installing at least an incandescent bulb or a load resistor at the third brake light. Alternatively, on my 90’S4, I installed a relay to modify the ground detect. With the relay installed, everything functions in the same way as before.

Any changes to the cruise control electronics may affect the safety of the vehicle, so I can’t recommend this modification to anyone. This post just documents what I did.

My original fix plan was to use a simple PNP transistor switch, but the cruise amplifier is unable to switch-on due to a voltage drop across the transistor exceeding 0.6 volts. What I ended up doing was to locate a relay inside the amplifier module that sourced the ground through the amplifier’s pin 12. From pin 8 I wired the relay coil to the brake light side so it sees 12 volts when the brake lights go on. From pin 8, I cut a trace on the amplifier’s circuit board and wired the relay contacts to ground. This is for the later models which have amplifier with the 911 part number.

I sourced a miniature DPDT relay, but a SPDT was only needed. This relay is so tiny, I can’t hear it activate when I apply the brakes. I covered the relay with ¾” shrink wrap tubing and hot glued the relay to the top of the large capacitors shown in the photos.

Here’s some images showing some details.






I could have bought similar parts at RadioShack, but I already had a this relay on hand. http://www.allelectronics.com/make-a...RELAY/-/1.html

Happy cruising!

Ed Scherer

Thanks for posting this. I'm going to be revisiting this issue within the next couple of weeks.

I hadn't even noticed that since replacing all my incandescent bulbs with LEDs that I lost my cruise control, since I haven't done any significant road trips since these changes.

I definitely want to address this prior to going to the 2009 OCIC in DFW. Speaking of which, are you going to be there?

borland

Ed,

For the trip to DFW, you could just install the one 21 watt incandescent in the third brake light socket. That should allow your cruise control to function again.

What might be simpler, if possible, is to just install one 20 watt load resistor in the third brake light housing.

I originally wanted to install a solid state solution. I designed two transistor switches (first bipolar, then JFET), but gave up and went with the mini relay.

I won't be going but look forward to seeing photos of OCIC next month.

My latest project is to install H3 LEDs next to the fog lights in the 'additional high beam" slots. Run them as DRLs. That would defeat the S4's wimpy daytime 'flash-to-pass' feature, so I'm looking at adding an extra relay and two diodes to make the headlights pop up for a daytime 'flash-to-pass' alternative. This may never get done, but if it does, I'll post the results.

Ed Scherer

Yeah, I might just go for the quick-'n'-dirty solution (load resistor). It really bugs me to do that (typical engineer; hate wasting energy), but sometimes expediency wins out.

Thanks for bringing the cruise control issue to my attention, though. I wouldn't have even noticed it until too late, because I never use it around town (well, until I just tried to test if after seeing this thread).

Tampa 928s

Do you have a schematic of the cruise control module? I need to repair mine but want to know how it's designed.

Alan

There isn't a lot of room there...plus it gets very hot (in a plasitic housing) - so I'd be inclined to put a screw down metal one somewhere where you can attach it to sheetmetal. I have one for my center brake light - its on the inner skin inside the sunroof area - easy wiring (just be careful drilling).

Alan

Ed Scherer

Barry, 77tony was asking me about brake light LEDs and I was reminded of the cruise control / brake light interaction and the nice work you presented in this thread.

In reviewing this material, it occurs to me that there might be a really easy solid state solution: an N channel power MOSFET (as an alternative to what you're doing with a relay in the bulb control module), as it doesn't suffer from the voltage drop you see with BJTs and JFETs. Did you ever try that?

I might give it a try this weekend and report back.

borland

Ed,

Without the aid of a circuit diagram for the controller, I tried to trace the circuit board components to see if I could engineer a solid state solution, but at the time, I found them too hard to trace. So, I went with my relay approach and that worked just fine.

I also recall experimenting with a transistor switch, but the voltage drop the controller was looking for across the brake light bulbs was too low. So my transistor switch wasn't sensitive enough.

I'm sure someone could come up with something like what your suggesting with the right approach. Good luck with your attempt.

Alan

Sometimes a resistor to ground may be just simpler & more reliable... I'd bet it doesn't actually have to be that low a value either - why not experiment with say a 1K ohm resisitor - if that works then it will be less than 1/4w dissipation so tiny power requirement/size & could easily live in the brake light housing (unlike a power ballast resistor).

Borland did you experiment much with this in terms of needed Rx value to ground?

Alan

77tony

Have already made the switch to the OB hatch glass w/out the later 3rd brake light. Already built in a 4" LED strip into the back of the rear headliner, so an incadecent bulb in that location is out of the question. Seems to me that installing a resistor above the rear headliner would get to hot in that restricted space ? Also, the cruise control is inoperable. Looking forward to seeing Ed's possible solution this weekend. TIA. T

Ed Scherer

Alan, a few months ago, I was experimenting with load resistors at the CHMSL location. A 25 Ω load resistor was not sufficient to allow the cruise control to work. A 6 Ω load resistor was sufficient. That was a rather disappointing finding.

I really dislike using load resistors (or "hidden" relays, for that matter), so I'm interested in a solid state solution if it's not too klugy.

Alan

Under the sunroof cover is not a restriced space at all - really tons of room - there is also the inner skin of the roof (with cavity) for direct metal to metal connections - easily screwed down.

Ed - I too don't really like the idea of a load resistor or a relay either. But in this case even for a solid state solution you will still need to sense that the brakes are not active and it seems the cruise brain provides a significant pull up - which will make that hard to do without a much lower resistance to ground anyway...

Now why on earth does it need as low as 6 ohms to ground though - I can see that the bulb control will need that much ballast to not alert - but can the cruise brain be sourcing so much current that it can't sense a low level with much less (like more than an order of magnitude less)?

So different way to think about it - to need this - in normal incandescent operation it must be dumping say 50+mA into the brake filaments the whole time the brakes are off... (assuming it needs a Vil of ~<0.3v) Really? Why does it even need any kind of pull-up functionally?

I do find this very hard to believe... may be worth trying again

Alan

Ed Scherer

Good questions. Not having dug into the circuitry (or even taken any measurements to see what it's doing when the cruise control is activated), I just don't know.

It's not the only thing that counts on getting a "near ground" through incandescent bulbs; the turn signal indicator light is another example. Kind of strange how it (turn signal lights and associated indicator light) works, but when you look at typical turn signal flasher circuits, it becomes clear.

I could hardly believe what I was experiencing myself when I was testing the load resistors. I repeated it a few times myself! I did this with some leads dropping down from the CHMSL and various load resistors to clip on. While driving. Swapping those resistors around while stopped briefly was fun (they get hot!)

I'd be really happy to see someone else reproduce those results.

BTW, the reason I suggested a MOSFET is that you can get cheap power MOSFETs: they provide a very low resistance source/drain channel with a high-impedance (probably not important in this case), voltage-controlled gate. Almost a drop-in replacement for a relay used in this configuration, assuming that I'm correctly understanding the circuitry in the bulb controller that Barry described.

Alan

Ed - this was specifically testing cruise functionality and not just for bulb warnings right?

Alan

Ed Scherer

Yup. I disabled my bulb warning circuit years ago.