7

I have recently acquired some Cree LED headlight bulbs for my ageing 5-series. They are amazing enough to actually make the vehicle safe to operate at night. The stock incandescent bulbs were woefully inadequate.

Unfortunately, and legal grey areas of the usage of such lighting technology aside; the vehicle's light control module sends short pulses to the bulbs when the lights are off to determine if they are blown. These pulses aren't adequate to illuminate an incandescent bulb; but are completely adequate to make LED bulbs periodically flash like crazy, making people believe that I'm driving...an inferior but also ageing vehicle.

Anyhow, I'm trying to design a circuit that will achieve the following:

  • Be very, very simple to make.
  • Supress the short bulb-test flashes from illuminating the LEDs.
  • Not affect the brightness of the LEDs when the lights are on.
  • Not make the light control unit think the bulbs are blown (I'm happy if the circuit bypasses this function altogether and always fools the LCU into thinking they're not).

The only thing I've been able to come up with, via my limited and often dangerous electrical engineering knowledge, are expensive time-delay relays. Attaching the largest capacitors I have (2200uF) across the bulb connectors has no discernible effect (though I'm willing to buy larger ones if necessary).

Can anyone suggest a design, or where I should be looking for information?

I'm also not averse to a circuit that will apply a simple transient over time to full brightness. It'll make the lights behave more like incandescent lights and therefore look less of a legal grey area!

| improve this question | |
  • 6
    You can disable the check. A google search for "bmw 5 series bulb test pulse" will get you information. – Andrew Morton Jan 5 '17 at 20:29
  • 1
    I was about to suggest a cheap DPDT relay wired with the lights-on switch to disconnect the LEDs and connect some dummy loads to keep the test happy, but Andrew's disabling the test in software sounds far superior. – Neil_UK Jan 5 '17 at 20:31
  • 1
    What does the pulse train going to the LEDs on the power bus look like? Would just putting an appropriately-sized resistor in parallel with the LEDs be enough to fool the module into thinking that an incandescent is actually attached? – Bitrex Jan 5 '17 at 23:27
  • 1
    @AndrewMorton I went with your solution. Using PASoft BMW Scanner I was easily able to reprogram the LCM not to perform the check or output the warning message. If you want to peg that in as an answer I'll accept it. – Toby Wilson Jan 7 '17 at 18:43
  • 1
    Are these LEDs designed to be used in headlights? A quick look at the Cree website doesn't show any automotive products. So did you just install some general-purpose LEDs? If the latter, your lighting geometry is probably wrong and you're blinding oncoming traffic. – Hobbes Jan 11 '17 at 14:16
4

schematic

simulate this circuit – Schematic created using CircuitLab

Something like this may work. The R1, R2, C values in the diagram has a time constant of around 1 sec, so it should reject pulse of a few 100ms. I really have no idea what is the duration of the pulse you need to reject, so you need to adjust the values.

Choose R3 with just low enough resistance to pass the check. Choose M1 with low RDS-on for low loss when fully on.

An issue to look out for is that M1 will have extended transition time which fits in with your slow turn on criteria but could dissipate significant power during the transition. But the transition should only happen at very low frequency, i.e. when being switched, so may not need special attention.

| improve this answer | |
  • Forgive my ignorance; what are M1 & M2? – Toby Wilson Jan 6 '17 at 8:41
  • 1
    M1 is the label for the P-channel MOSFET and M2 is the label for the N-channel MOSFET. – rioraxe Jan 6 '17 at 18:41
  • +1 you have the basic idea .This could be refined to become a product or maybe built into your LED fittings. – Autistic Jan 7 '17 at 21:29
2

Credit to @Andrew Morton for this answer; which he is happy for me to submit.

I simply used PA Soft to disable the checks in the Light Control Module (LCM) by going into "module coding" (or similar?) and unchecking the checks, then writing the coding data to the module. Make sure you read the coding data from the module first!

Consequentially, this means the module will NEVER check or report headlights again unless this is undone; so bear this in mind if you may revert back to incandescent bulbs in future.

| improve this answer | |
1

I would suggest a resettable fuse or polyfuse installed across the bulbs. This may delay your actual turning ON of your headlights.

| improve this answer | |
  • This is about the simplest idea that could work, the only potential issue is the very hot temperatures you can find in a car- Think being in the sun in Phoenix in July. That may prevent the PTC from working properly, but even if that's the case the lights would still function. – John D Jan 5 '17 at 20:25
  • I'm not sure how this would work; by "delay turning ON the headlights" do you mean opening the bonnet to physically reset the fuse? – Toby Wilson Jan 7 '17 at 18:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.