Converting 12v to 10v to power LED tail lights.

[Petes540i6]

Greetings experts:

I am trying to figure out how to convert a 12v system used to power incandescent tail lights in a car so I can use LED lamps that are in an upgraded version of a tail light assembly. The specific use here is to put in LED "CELIS" lamps from an 2001-03 BMW five series into a 1998 5 series.

(The extent of my electronics experiences is building two devices to cancel "bulb out" warnings, which required little more than soldering a few items from radio shack together and putting them in small project boxes.)

I understand that the appropriate voltage-out I need is 10v. I do not know what amps it needs (and frankly, I don't know if amps matter in this instance, but it appears to be an issue based upon my limited Internet research.)

Also, I suspect the "volts in" can vary quite a bit because I am using a car battery that is used for starting the car, recharged by an alternator, used to run electronics, etc.

My questions are these: First and foremost, how should I build something that will fit this application? How might I figure out the appropriate amps (if necessary)? How can I address the fact that voltage-in might fluctuate quite a bit?

Thanks in advance.

 

[Diver300]

LEDs need to be supplied with a fixed current (measured in amps). From what I can see on the web, the BMW tail lights run at about 6 V, and 10 V is just what the supply circuit happens to supply when the LEDs are disconnected.

Therefore you need to know the current that the lamps take. It is also really important that you supply each lamp separately, with a separately regulated. BMW E39 Euro-spec CELIS "LED" Tail Light Retro-Fit Kit shows a regulator for each tail light.

That is because the voltage needed to supply an LED does not change much when you change the current. If you supply an LED with 1 mA, you might get 1.9 V across it. Increase the current to 20 mA and it might only go up to 2.0V. But that is for one LED, at one temperature. The voltage will vary between LEDs and it will change with temperature. If you put two LEDs in parallel, you could get most of the current flowing in one, and hardly any in the other. The one with most of the current will burn out, so then all of the current flows in the other, and that burns out.

I think that the 6V on the BMW LED lights is because they are made with 3 LEDs in series. That way the voltages add up, and the current is exactly the same in each one.

 

[Petes540i6]

Diver300

Thanks for the reply. My intention is to build a "resistor/regulator" for each side, akin to the control boxes that come with HELLA's retrofit kit. (The retrofit kit is difficult to purchase without also buying the updated lights, and I already have the lights.)

The tail lights appear to be made up of three approximately six-inch long LED's running adjacent to one another. While only one wire runs into the lamp for the LED's, I do not know if they are set up in such a way that the LED's are in essence "in-line" or whether the one wire feeds electricity to them concurrently.

Is there a way for me to measure (or safely estimate) the current that will be used for the LED's?

When you use the term "temperature" do you mean the temperature of the air in which they are used or the "temperature" (color) of the light? (At present, I do not know whether the LEDs are emitting red light, or they are white behind a red lens.)

Also, what type of device should I be considering constructing? In the thread below are images of the device one person constructed, but he has not yet shared his plans with me for its construction.

What a fun-filled weekend->Retrofit Galore! - Bimmerforums - The Ultimate BMW Forum

Thank you again.

 

[Petes540i6]

Thank you Diver300.

I drafted and submitted a post a few days ago (or at least thought I did) that might be in the moderation queue, but nonetheless, I have learned more about the light systems.

I have been advised that the "North America Spec" LED lights can be run on the vehicle without the need for the Hella CELIS control units. Apparently, the North America lights have a few extra LEDs, and can run on the "12v" system just fine. The "Euro Spec" lights need to be run on 10v with the regulator, apparently. I will be using the North America Spec lights.

One person on a BMW chat room has advised me that he is running NA Spec lights in his car without a voltage regulator for two years without an issue.

Also, it appears that my concerns about fluctuating voltages may be unfounded. The lights of the vehicle are powered or governed somehow by a "Light Control Module" that appears to serve as a regulator. If I take a reading of the battery of my car while not running, I get an expected constant reading of 12.44v. However, when I take a reading of the volts coming to the parking light bulb, I get a fluctuating figure between 11.88v and 11.95v. It is constantly variating at figures between these two voltages. Thus it appears to me that the LCM is taking a very active role in the regulation of the voltage to these lights.

It makes me wonder if the "current" is also regulated by the LCM. But because of my limited understanding of electronics in general, I do not know if the current is something that can be regulated from the source of the electricity, or is more a matter of that which is receiving it (in this instance, the LEDs.)

I have not yet tested the LCM's regulation under the stress of staring the car. It would be interesting if it manages to keep the voltage between these figures under that stress.

 

[Nigel Goodwin]

The factory LED lights I've seen just use simple resistors to feed the LED's, with three or four LED's in series. The voltage in a car is relatively 'regulated' anyway.

 

[Ubergeek63]

Greetings experts:

I am trying to figure out how to convert a 12v system used to power incandescent tail lights in a car so I can use LED lamps that are in an upgraded version of a tail light assembly. The specific use here is to put in LED "CELIS" lamps from an 2001-03 BMW five series into a 1998 5 series.

(The extent of my electronics experiences is building two devices to cancel "bulb out" warnings, which required little more than soldering a few items from radio shack together and putting them in small project boxes.)

I understand that the appropriate voltage-out I need is 10v. I do not know what amps it needs (and frankly, I don't know if amps matter in this instance, but it appears to be an issue based upon my limited Internet research.)

Also, I suspect the "volts in" can vary quite a bit because I am using a car battery that is used for starting the car, recharged by an alternator, used to run electronics, etc.

My questions are these: First and foremost, how should I build something that will fit this application? How might I figure out the appropriate amps (if necessary)? How can I address the fact that voltage-in might fluctuate quite a bit?

Thanks in advance.

Well technically you need to survive a 80V for over a 250mS per ISO-7637. And no matter how many hobbyists tell you otherwise, that is what the INTERNATIONAL STANDARDS ORGANIZATION requires of the industry.

And no matter how many hobbyists or even measurements tell you otherwise, you risk blowing out your expensive LEDs if you do not protect against it.

The other thing that most people seem to miss is that load dumps are inductive and when you clamp it you EXTEND the pulse length since the same energy at the same current is dissipated by a lower voltage.

dan

 

[ronv]

What keeps me from jumping in on this is that I don't know what is inside the little regulator modules. It looks a little more complicated than just a resistor. Could be a constant current regulator, but if it is, what current? You could probably get away with a power resistor. Maybe that is a way to creep up on it. Start with 50 ohms and see how bright and work your way down to the brightness you feel safe with. Then you can either build a regulator or if the resistor doesn't get to big for the current just use it.

 

[Ubergeek63]

What keeps me from jumping in on this is that I don't know what is inside the little regulator modules. It looks a little more complicated than just a resistor. Could be a constant current regulator, but if it is, what current? You could probably get away with a power resistor. Maybe that is a way to creep up on it. Start with 50 ohms and see how bright and work your way down to the brightness you feel safe with. Then you can either build a regulator or if the resistor doesn't get to big for the current just use it.

There is no way for a simple resistor to perform adequately. In a vehicle electrical system you need a constant current source with a 100V compliance at the rated (NOT MAX) LED current

 

[Petes540i6]

OK Here's the latest with the project. I have both tail-lights in the car. Note that the LEDs are only the "parking light" portion of the light. The other bulbs (reverse, turn, brake) are all conventional incandescent/filament bulbs.

The Light Control Module was "unhappy" with the LED's and causing a number of "symptoms." These are not unusual from others' experience: Things like: (i) causing a turn signal to flash at 2X speed, even though the bulbs for the turn signals are good and the turn signals might otherwise be thought not to be related to the LEDs (ii) dimly lighting either a reverse or brake bulb on one side of the vehicle even though that bulb should be "off", and (iii) a subtle flickering of the LED's.

This changed when I put a small 5W incandescent bulb in parallel with the LED's. All the symptoms identified above disappear.

The question I have now is how is related to the fact that I am currently running two 5W bulbs in the trunk of my car. I'd rather have a "resistor," or whatever I need to have the LCM "think" that there is a bulb parallel to the LEDs so I do not have to run this bulb there.

Thus, my question is, "How do I build a load resistor to "trick" the car into thinking a standard 12v 6W bulb is in the taillight?"

Thanks again.

 

[ben7]

You could use a 22ohm, 10 watt resistor as a dummy load. It would dissipate around 6.5 watts at 12volts, but the car's voltage might vary, so at 14 volts it would dissipate about 9 watts!

 

[Petes540i6]

You could use a 22ohm, 10 watt resistor as a dummy load. It would dissipate around 6.5 watts at 12volts, but the car's voltage might vary, so at 14 volts it would dissipate about 9 watts!

Thank you. I have since read that a 25 Ohm, 25 Watt load resistor will work, which are readily available. Does this seem plausible?

Also, would such a load resistor be fine to run in place of any other 12v 6W conventional parking lot bulb? I ask because the car originally had two parking light bulbs on each side, and now has one "set" of LED's on each side powered by one parking light bulb wire. Thus, I have an extra parking light wire and getting a "bulb out" warning from the LCM that I would like to defeat with a load resistor.

Thank you again.

 

[Petes540i6]

Edit: Here's an example of the load resistor kit I have in mind:

Load Resistor Kits

 

[Diver300]

Those load resistors are just the right thing.

It is highly unlikely that you need to take the full load of an incandescent bulb. If the car expects 5 or 6 W, even taking 3 W (1/4 Amp) would almost certainly be OK. The resistance (along with the supply voltage) is what controls the power that it taken. A 47 ohm resistor will take 1/4 Amp (approx) at 12 V.

You then need a resistor power rating well above what the resistor will take. If you go for 47 ohms, that's about 3 W at 12 V, or about 4 W at 14 V. You should use a 10 W or more resistor if you don't want it to be too hot to touch.

The real problem with LEDs is that they take no current at low voltages. Also, if you put a small current through them, the voltage is quite large.

For example, if you have a 5 W bulb, the bulb failure circuit might put 1 mA though it when turned off. That might give 1mV across the bulb and it won't get anywhere near hot enough to glow. If you have an equally bright LED, if you put 1 mA though it you might get 6 V across it, and it will glow. The bulb failure circuit sees the 6 V and shows a failure.

Even a 1 k resistor, which does not need to be more than 1/4 W, might fool the bulb failure circuit. However, the exact current used to detect blown bulbs varies.

 

[ben7]

I'd agree with Diver300, as long as there is a good current flow(500mA or so) there should be enough to trick the LCM. Yeah, the led's probably only use 20-50mA, not much at all.