LED Brake light clusters - DIY

[alexanderfitu]

Hi all


I recently made some LED light clusters for my fazer.


I wrote up the process here:


http://fotifixes.com/2013/01/16/led-brak...rs-fzs600/


I know you can buy LED light clusters for the fazer, but they aren't as bright as mine


Lemme know if anybody has any questions

[Dead Eye]

Looks pretty awesome, nice homebrew mod there

[alexanderfitu]

cheers! they flicker a bit when the indicators are on because  of the volt drop, so I'm working on some indicator boards too

[Dead Eye]

I'm no electronics expert and this could be one of the most stupid things I've ever said, but wouldn't a capacitor help?

[Panthor]

+1 for a really good mod that looks really awesome and will definately help you bring seen

[Panthor]

+1 for a really good mod that looks really awesome and will definately help you bring seen

[alexanderfitu]

I'm no electronics expert and this could be one of the most stupid things I've ever said, but wouldn't a capacitor help?

neither am I  :'( I'mlearnin as I go along. a cap should help though.it willbe in the next revision

[alexanderfitu]

I'm no electronics expert and this could be one of the most stupid things I've ever said, but wouldn't a capacitor help?

neither am I  :'( I'mlearnin as I go along. a cap should help though.it willbe in the next revision

[Fazerider]

Neat job. Admittedly I'm under the influence, but doesn't 1Ω and six diodes in series mean the current for each chain is only really being limited by the forward voltage drop of the diodes?

[alexanderfitu]

Neat job. Admittedly I'm under the influence, but doesn't 1Ω and six diodes in series mean the current for each chain is only really being limited by the forward voltage drop of the diodes?

Technically yes, but the LEDS have a high forward current of 100ma so they require a fair bit of current (As they are so bright). I initally designed the circuit for 13.5v with 15 ohm resistors, but after adding the diodes etc and measuring the volt drop I ended up dropping them to 1 ohm for 12v.


I used an LED array calculator to help with the maths.

[Fazerider]

Once my blood/alcohol ration is down tomorrow I'll probably have to admit I should have left the arithmetic until morning but I think you'd do better to leave at least one diode out of each chain. That way the array will cope better with voltage fluctuations due to the indicators etc. It'll also protect the LEDs because the forward voltage drop varies with temperature.
Also, I don't want to be behind you when you apply the brake at night... so far as I can estimate that's going to be about 6x as bright as two 21w bulbs. Bearable in daylight perhaps, but overkill in the dark surely?

[alexanderfitu]

Ah that would make it better I suppose, it does seem very sensitive to voltage drops etc.


If I removed an led from each chain, what would the resistors need to change to?


Its the first time I have built an LED array, if I get some time, I will re-do them with less leds in each chain.


regarding the brightness, they arent stupidly bright, the RED housing dissipates a fair bit of the light, I would say 50% brighter than the original bulbs.

[alexanderfitu]

If I remove 1 led from each chain, and change the resistors to 10ohm, do you think this would help with the flickering with indicators on etc?


I wonder if wiring caps across the powersupply would be easier ?

[Lumpy]

get a white / clear tail light cover. Sorted.

[Fazerider]

What is the forward voltage drop for your LEDs at their normal operating current? I've been assuming it would be 1.9V, but they may be different depending on the manufacturer.
If we go with a figure of 1.9V for the moment, five in series would give a total drop of 9.5V.
If you reckon on a 13V supply rail then the resistor has to drop the difference of 3.5V, if you want 100mA then its value needs to be 35Ω. If the supply drops to 12V then the current will reduce to 71mA, just a 29% reduction in brightness (whereas with 6 LEDs and the appropriate resistor the change would be 62%).
It is quite dependent on the diode characteristics though, if you've a link to a datasheet for them it'd be better than me guessing.
A capacitor would have to be enormous to have any effect, unless you've one of those giant 1F caps used for in-car sound systems going spare, I don't think that's the way to go.
I'll take your word for the brightness being OK, its hard to translate cd into lumens to compare with an incandescent lamp... but I've been caught behind the occasional vehicle with excessively bright brake lights and it's not good for night vision!

[alexanderfitu]

The forward voltage is 2.0 to 2.2 volts, forward current is 100ma.

Also I get a 1v drop in the wiring so if the battery voltage is 13v I get 12 at the loom, additionally the diode drops 0.5 -0.7 volts.

Its a schottky diode, but im not sure of the exact one, as its one I had lying around.

going by those above specs, a 10 ohm resistor with 5 leds in strings would give around 100ma?

[Fazerider]

OK, so if we use 2.1V as the value, that gives 10.5V drop across a chain of 5 LEDs. If we stick with 13V as the supply for the moment (in order to make sure the LEDs don't get over-driven it's best to take the highest voltage you're likely to get) that gives 2.5V across the resistor and 25Ω would be the right value to give 100mA. Then, if the voltage drops to 12V the current will reduce to 60mA.
If you suit the resistor for a 12V supply you'd use 15Ω and should the voltage rise to 13V (with the headlights off and at high revs for example) then the current would rise to 167mA.
In fact, the current swings wouldn't be quite that severe because the forward voltage drop of the diode would also increase... it's what is limiting the current with your present circuit.


As for the tail light, I'm not sure what current you'll be getting at the moment. With such a narrow voltage margin for the resistor to use it's well into the knee region of the LED's curve. Perhaps 30mA per LED would be a good starting point, with 5 in series and assuming a 1.9V forward voltage at that lower current that gives 9.5V.
A 12V rail then would be reduced to 11.5 by the Schottky diode (think you've drawn that in the wrong place for the LED strip, it's there to stop the clocks and sidelight lighting up with the brake when the headlights are off). So the total resistance a chain needs would be 67Ω... however, there is already 25Ω present so the remaining 42Ω could be provided by a single resistor of 3.5Ω (twelve 42Ω resistors in parallel). The nearest preferred values are 27Ω and 3.3Ω but you might need to experiment with the current for the tail light.

[alexanderfitu]

Thanks for the help with the calculations.

my diode is dropping close to 0.7 volts and the wiring a bit more than a volt, so I went with 5 leds per string with 15 ohm resistors, this has helped make the tail lights brighter when on at night (previously pretty badly dim), i stuck with the same 20ohm 0.5w resistor I had in place and its working much better.

Thanks

[Fazerider]

OK... that would give a tail light current of about 50mA so about half brightness and Vf wouldn't be much reduced. I'm a bit surprised you find that enough of a difference in brightness though, the incandescent bulbs you've replaced go from 5W to 26W.
Cars with OEM LED light clusters appear to control the current on a pulsed basis: a long exposure photo of moving cars shows a string of dots rather than a continuous streak.

[alexanderfitu]

In reality its a bit dimmer than half, as the original brightness was barely illuminated. I suspect the brightness curve  of the LEDS is very " curvy" if that makes sense.
IE 50mA is 25% brightness rather than 50%. I have noticed that as well, but PWM stuff on this is a bit ott