Luxeon turn signals

Candle Power Forums

Help Support Candle Power:

highlandsun

Enlightened
Joined
Aug 11, 2002
Messages
607
City & State/Province
Los Angeles, CA
Generally I see folks here all over these LED mods because the LED version of a light is more efficient, provides better light longer, etc... I.e., there are many tangible benefits to doing a conversion. When I contemplate replacing my car's front turn signals with amber Luxeon stars, I don't see a ton of compelling benefits, and I do see a huge expense. Would you consider spending this kind of money yourself?

I actually started down this path because 3 of the 4 bulbs in my original front corner lamps had burned out within a couple weeks of each other. (two 194s, one 1157 burned out.) These were the original bulbs, on my '97 Probe GTS. On this car, unless you have very long/thin hands and arms and can reach through the wheelwell, you have to remove the front bumper cover to get at these bulbs. That is quite a lot of hassle to replace a couple lousy bulbs, and certainly not something I'd want to do more than once.

At this point, the idea of a long-life LED starts to sound really good. But still, incandescent signal bulbs are like 20 cents each, and getting the equivalent light output from LEDs will cost well over $100. It's hard to justify this mod on a dollar-value criteria.

Anyway, not being one to shy away from spending money on my car, I ordered Ford's Euro-Spec corner lamps, which have two external mounting screws which let you get access to change the bulbs without having to remove the entire bumper cover. How clever; I wonder why Ford didn't do the same thing for their US lamps. Anyway, that was $120 each already spent. These Euro-Spec lamps also have clear lenses, which is kinda cool, if not overly trendy...

At this point I could say my original problem is now solved - the bulbs are easily replacable, end of story.

But no... The Euro-Spec lamp doesn't use the same bulbs as the US-spec, nor are they mounted in exactly the same configuration. The US lamp uses an 1157 in front and a 194 on the side. The Euro lamp uses a 194 farther in front and an 1156, also in front, with a blank reflector on the side. This leaves me with very little side-illumination, something that you generally would like to have, particularly for a *turn signal*.

Aside from that, things just look ugly with a clear 194 mounted there next to an amber 1157 bulb. (Fortunately the lamps use the same harness connectors, so I can use my US harness and keep the 1157 marker/flasher functionality. The 1156 bulb has a different mounting ring but if you twist & push a while you can make the 1157 fit.) The 194 has that weak, sickly yellowish coloring, which is all the more obvious next to my (very white) HID headlamps, amber 1157, and dichroic yellow fog lights.

I quickly decided to ditch the incandescents and try those drop-in LED bulb replacements. So, now I have a uniform amber color in the 194 and 1157 locations, but their light output is a joke. I'm amazed that selling these things is legal. A 194 puts out 25 lumens, this 4-LED 194 replacement probably puts out 4 lumens, tops. The high beam on an amber 1157 puts out 300 lumens, the little 1157 LED bulb probably puts out 15 lumens at best. (They're just clusters of 5mm LEDs and I seldom see any individual LED with anywhere near 1 lumen output.)

So, I've started designing with some 1W amber Luxeons... At the moment I'm planning on 4 in series; they will be extremely underdriven in the normal marker/parking light role. The corner light only needs to put out 50 lumens; by using 4 1W Luxeons I can space them evenly in the corner lamp and light up the dead spot that is currently just a flat reflector. For flashing they'll be slightly overdriven. The total output from 4 ambers at normal spec (350mA) should only be 144 lumens, still far short of the 300 from the incandescent. But at the moment I'm not willing to buy the extra Luxeons that would only be used for a fraction of their existence...

Considering that I've already spent over $1000 on HID headlamps (2 different kits) and a couple hundred on the corner lamps thus far, I guess it's not such a huge leap to spend another couple hundred to make this happen. But sometimes I wonder if it's really worth the effort and the cost, especially looking at the cheap price of incandescent bulbs. And while LED life is certainly a consideration, electrical efficiency really isn't; in fact it's somewhat of a drawback since the ultra-low current consumption makes my turn-signal flasher think there's a bulb missing...

So - in a similar situation, would you do this mod, or would you just buy some decent incandescent bulbs and call it a day?
 
Hey Howard, I've been through the long stage you are going through. As soon as I got into LEDs, I wanted to be the first to use them in my car and impress everyone. But it became pretty much a dream and that alone. The 1157 and marker bulbs they sell are a sorry excuse for "bright" as you probably know. I'd rather have the brightness than the dimmer instant on/off.

For them to be bright, you would most likely need a large array of LEDs, and they would definately be brighter outside the diffracted signal lens, like the LED brake lights are. I've been almost tempted to try LEDtronics' 1157 LED bulb, but at almost $50 each I just couldn't put myself to it - I would need 4 alone for my taillights. I've even tried making my own array of LEDs, but it didn't work out as well as I wanted.

Concerning luxeons, you're right on that you will need more than one. I haven't tried it because its just a lot of effort for something that will probably not please me as much as my halogen (front) park/turn signals
smile.gif
I've also gone through the trouble of relaying all of my lights already.

Take a look at http://www.radiantz.com - they custom build and sell arrays for auto use. After giving up on the LED bulb thing, I looked around the web for auto-LED conversions. I saw this one site where a jeep was fitted with truck LED signals, which I thought was awesome - picture the brightness and effect those would create, and they look great. So I started thinking about the possibilities of replacing my whole taillight lens with truck LED lights that I have been researching, but I couldn't hack away at mine. I bought a cheap 52-LED amber array made for a truck, to use for one of my front park/turn signals - it is bright! It's about the same size as the lens on my car, I just never put in the effort to retrofit it because it didn't seem easy to secure it and it is in a very difficult position to play around with, so it just sits in my room (for now). My car's lights would probably be hacked up by now if I had a job to buy all these things, but then again it could turn out to be a huge mess and disappointment.

So I kind of gave up with the LED deal, and being depressed as to seeing everyone with those cool windshield washer LEDs that don't even know the significance of them - just think they're a "cool addition" to their souped up cars.

The only LEDs in my car as of now are my luxeon dome light, bright red blinking alarm light, and my spoiler LED brake light.
 
The LEDs arrived today. Cool. Testing them with my LiIon cellphone batteries and some NiMH AAs. I was expecting a single LS to draw 130mA at 2.4V but instead I see 85mA at 2.6V (from two NiMH AAs, open circuit voltage 2.7V). Oh well. I'll have to measure all 8 individually to get an idea of their range.

With 4 in series and 14.0V coming from the flasher they'll only see a max of 3.5V which is within spec. A single one hooked up to my LiIon (4.1V open circuit voltage) lit up fine, but obviously got warm.

Designing the circuit is easy enough, but making it waterproof, that may be a puzzle.
 
Well, keep me updated here. I'd like to see pictures if possible - what car is being used for this experiment and what signals are being fitted?
 
As an Amazon Associate we earn from qualifying purchases. Product prices and availability are accurate as of the date/time indicated and are subject to change.
Wow, 2 in series from a 4.0V LiIon actually turns on with 8mA current flow. Dim, but visible. Adding an NiMH AA in series brings it up to 5.2V @ 112mA. The problem with testing with individual NiMH cells seems to be that they really can't handle the current demand on their own, too much voltage drop...
 
I went and bought the rest of my parts today. Using a PNP transistor to provide a current source. Using a 12V wall wart to drive the prototype circuit.

Here's the drive circuit:

luxeon2.GIF


The 1.7V LED is Radio Shack 276-307, I used it because it was pretty much the only one they had in the store. If anyone can point me at a ~2V amber LED please let me know; I'd rather not have a useless red LED in this circuit.

The PNP transistor drops the 1.7V from the LED to 1.0V. 1.0V divided by the 10ohm resistor yields 100mA, so the transistor will only conduct 100mA. The 4 amber Luxeon Stars connect in series between the PNP collector and ground.

I use a PNP configuration (instead of NPN) so that I can also feed +12V directly from the flasher output. At 12V they will only be about 3 times brighter than normal. At 14V (alternator running) they should be about 5 times brighter, which will meet the DOT requirements.

Now all that's left is to mount them into the turn signal housings.
 
Looks like I may just use a LM285 1.2V diode and a 5ohm resistor instead of the 1.7V LED and 10ohm resistor. Certainly the thing being an LED isn't of any benefit since the Luxeon Stars themselves provide sufficient light.

The meter I have only measures up to 200mA. (I ordered one of those cheap Cen-Tek's yesterday but it will be a while before it arrives.) Since I had nothing else handy, I stuck a 10ohm resistor in series with two Stars and ran 12V into it. The Stars pulled 6.70V and the resistor 5.30V, so that gives 530mA. This is good news; I will definitely be at or above 550mA with 4 stars in series driven by the 14V alternator output. Couldn't run it this way for long, there was a burning smell. (The resistor was nowhere near big enough for the power level..)

I haven't decided yet whether to add a load resistor to the circuit or not. Since my car has an electronic/hybrid flasher, it will flash at double speed because the current draw of the flashing LEDs is too low. If I use a load resistor I'll have to add a diode too, to prevent it from sucking up any current in the running light mode. That'll drop my 14.4V down to 14.0 (schottky diode), but I guess that's still OK.
 
The above circuit caught my eye for its simplicity, and so I started playing a bit. I came up with the following, which is only slightly more complex, has a slightly lower 'headroom' requirement, and allows one to select two different regulated current levels (for run and flash, for example).

linear_auto.gif


In the above circuit, the high value resistor supplies base current to the main power transistor. This causes current to flow though the current sense resistors. As more current flows though the sense resistors, the voltage across them causes base current in the signal transistors, which steals base current from the main power transistor. The equilibrium is reached when the voltage across the sense resistor is about 1 diode drop.

I selected the TIP127 transistor because it is _cheap_ from digikey, and is a darlington pair that requires much less base current. It is in a TO220 package which uses the PCB for heat sinking. It should have no problem dissipating a couple of watts (the package is rated for 10s of watts, but there isn't much heat sink provided.)

The selected device is not 'high precision'; the sense resistors are 5% because of their low value. Additionally, the base voltage needed to trigger the small signal transistors will vary anywhere from 0.4V to 0.8V; however most of this variation is caused by temperature, and the device will reduce the drive current fed to the LED as it gets warmer. But it should do a reasonable job of keeping the LED current stable over a wide range of input voltages, and the components should be able to tolerate an automotive environment.

I have a PCB layout that is a bit less than 1 inch square

Design comments welcome!

-Jon
 
Argh, I just ordered some other parts from DigiKey a few days ago, should have waited then I could order these too. Now I can't try this circuit for even longer...

I got my sample 1.8V LEDs from Sloan Corp. Promptly fried one of them by inserting it into the circuit while power was applied. Duh... Oh well, live and learn... The 1.8V LED that remains ran the circuit at 115mA. Still good, but the color wasn't a perfect match. My DigiKey order was for the 1.2V zener reference diodes and 5ohm resistors, so that will no longer be an issue.
 
Been a bit busy to tear my car apart but I've been playing with this some more. Measured 14.8V on the marker lights with the alternator running. I couldn't get a good reading on the flasher because my meter's response was too slow, but it looked like about 14.3V peak. It turns out I need to add a diode on the flasher anyway, because the front and rear flashers are a single circuit on each side. A decent schottky diode (1N5817 20V, 1A) will drop the voltage another .4V or so. Once again, I tried using two Luxeons in series to simulate this with my power supply set at 7.5V. I measured 660mA at around 7.3V going thru these two (vs 100mA at 4.70V). Definitely in the right ballpark. Assuming around 13.9V after the diode drop, should put it into the 500mA region. Off to Frye's tomorrow to find some diodes....
 
I'd rather not think about that...

The LEDs themselves would probably survive, at least. It would be a pain to get another pair of Euro-spec lamps from Ford to modify...
 
By the way, I've added some beam shots comparing the Luxeons with the original incandescent signal lamps. At the bottom of the page.
 
Back
Top