Best Resistor for auto Application

Hi,

I was wondering what your guys experience using the luxeon Stars for 12 volt Auto power. I think the Car actually makes 13.6 volts. I'm currently using a 45 ohm 3 watt resistor on the amber 1 watts and it seems to get kinda warm to the touch. I'm I pushing it to hard?? I did my ohms law and this is what it said to use. Any help would be great. Thanks

ski...

A 3 watt resistor is supposed to get warm, though not hot enough that you burn yourself touching it.

I'd worry a little more about voltage spikes from the car damaging the LED. The 12 volt power in a car is very noisy because of electromagnetic hash produced by the engine and spark plugs.

With the alternator running I usually get 14.4-14.6V. What are you going to do with this Luxeon, how frequently will it be on, what duty cycle? You may want to use a linear voltage regulator or something to keep it within a more specific voltage range. Or use an LVR with a sense resistor to run it in current regulation mode, that would be even better. This assumes that you are really concerned with keeping it at a constant power level. Realize that the difference between battery-only (12.0V) to alternator on (14.4V) is so wide - 2.4V, the entire Vf of an amber Luxeon, that a single resistor is not sufficient to keep it running satisfactorily under all conditions. If you use a resistor large enough to run it safely at 14.4V, then it will be seriously underdrivem at 12V. If you target 12V, it will be seriously overdriven with the alternator. If you target 13.something, you're still pushing it, just not as badly...

I'm not sure how you 'do the math' but when I did I got slightly different numbers.

First off, as another poster pointed out, expect close to 15 Volts running hard with the headlights off. This less say 2.5 Volts (Vf for the red LED) gives us 12.5 'across the resistor'. If we want full current (.35 Amps) we need 12.5/.35 or 36 Ohms. Since our Voltage is 12.5 and our current .35 our power (in Watts since voltage and current were in Volts and Amps) is 12.5 times .35 or 4.375 Watts, a higher wattage resistor is called for, I'd go with ten or so. For a 'handy Radio Shack solution', perhaps four 150 Ohm, two Watters in parallel (effective 37.5 Ohms, 8 Watt)?

You're running 12.5/45, about 277 mA. This current times the 12.5 Volts, 3.46 Watts, I bet it did get warm.......

This setup, contrary to another poster, is fairly stable within reasonable voltages. At say 12.0 Volts we have 12.0 less 2.5 or 9.5 Volts 'across the resistor'. Divided by 36 Ohms is still 264 mA a fairly modest 25% drop that will be hard indeed to spot in most cases. And while noise in auto electric's is a concern, as are spikes (by far the worst is when the starter and it's solenoid 'drop out') I wouldn't worry too much. Noise averages zero so doesn't contribute to heat and any spikes will be short in duration, and attenuated by our series resistor and their effect should pass without damage.

Or so I see it.

Doug Owen

I just had a 'shower thought' (usually my best), this may be just the time for the 'constant current 317'!

Gotta get to work, but this can be food for thought for the drive in.

Cheers and good day to all our readers...

Doug Owen

Hello there,

Since you are using an amber LS and running off of a
12v automobile system you will see resistor heating
due to power dissipation as high as 2.88 watts or even
more. This will certainly heat up a 3 watt resistor!

What i would recommend is to use two 100 ohm resistors
in parallel, each rated at 10 watts. This will not only
give you a higher power rating, it will distribute the
heat to two resistors instead of one. This means each
resistor only has to dissipate one-half of the total heat.
Since they are both 10 watters, each one will only have
to dissipate a maximum of one-eighth to one-quarter
of their rating, which is a very safe operating area indeed.

I dont think it's too bad of an idea to use resistors,
after all they are the simplest and most reliable drive
system. You wont find the regulation as bad as if you
were to drive from say a 6v system with the same percent
change in voltage.

Here is what you can expect from using two 100 ohm, 10 watt
resistors in parallel (50 ohms) for a battery voltage of
11.5 to 14.5 volts using an amber type LS LED (2.5v):

1. At 11.5v, you will see about 180ma, and the resistors
will have to dissipate 1.62w, or 0.81 watts each,
which is very low.
2. At 14.5v, you will see about 240ma, and the resistors
will have to dissipate 2.88w, or 1.44 watts each.

Both of these numbers are very low and shouldnt cause any
problems at all. Worst case is 3.125 watts, which works
out to 1.6 watts per resistor, which is very safe.

The percent regulation in light output is about 14%, which
isnt too bad really.

If you go to a series regulator circuit, you will have to
provide a heat sink that can dissipate at least 3 watts.

Even though the efficiency is not good with either of these
methods, it's not really an issue unless you are planning
to run many strings of these LEDs, in which case you would
want to look into a step down switching regulator.

Good luck with your LED circuits,
Al

Pull up a crate, it's story time.


My sister has a Volvo 850. Some of them, but fortunately not hers, have a problem. The relay on the power door locks sticks causing all 6 (4 Doors, Truck, Gasgap) lock motors to fry. They solved it using a 2-tier solution if you will...they fixed the relay so it doesn't stick and then, just in case, the new motors have a special type of resistor on them (Don't recall the name). The resistor increases it's resistancy as it heats up so that instead of the motor butning up, the resistor more or less cuts off the electrity to the motor.


Anybody have any clue of the type of resistor I speak of?

Saaby, strange you should mention this... came accross a very similar problem with a friend fixing a commercial microwave oven where it has door release solenoids just the other day.

What your after is a PTC, positive temperature coefficient thermistor, guessing Digikey, Jameco et al, the usual suspects will be the people to search around.

possibly more than you wanted to know:
http://www.chipcenter.com/eexpert/akruger/akruger030.html

HTH
Adam

Once again, we really need to know more about just where and how this LED will be used. The change from 350mA to 270mA is a 30% drop in light output, certainly noticable under many conditions. The question is whether it's actually *important* and that depends on the purpose of the light.

For my front turn signals, it was important to me that the marker light output would not only meet SAE specs at all times, but would be *constant*, not just "in the range", which is why I used a transistor driver instead of just some resistors. The transistor is also rated for some 40-60V across its inputs, so I didn't bother with any other spike-suppression measures.

If it's just for a dome light or some other interior accent lighting, I suppose it wouldn't matter. But who knows, if you're using the dome light because you need to find something, I'd shoot for a full 350mA worth of illumination under all conditions, and thus use a driver.

Hello again,

I know the white LS is rated for 350ma average current but
isnt the Amber LS rated for 170ma average current or
something lower like that (1 watt models)?

Also, usually a 30% total change is viewed as +/- 15%,
but in this application it may be better to look at it as a
full 30 percent change.

Good luck with your LED circuits,
Al

Nope, all the 1W LS's are rated for 350mA. So really only the red/orange/amber ones are 1W parts, all the rest are 1.25W.