Led series help needed

Hi

I hope this is the right forum for my question.

I have 4 strawhat 3.6V leds in series on a 13.5V car dashboard.

within a short time they start flickering and eventually one of them dies.

If i turn down on the rheostat the flickering stops, but this is not usefull.

do these leds need to be paired or what is going wrong?

Thanks for your help!

Welcome to CPF, dirodima!

You have too many of those 3.6V LEDs in series to power with 13.5V (and no series resistor? :duh2: ). Split it up in two strings of [LED + LED + resistor], and wire those strings in parallel.

Take the specs of those LEDs, search online for "led calculator" or similar, put in the values, done.

RetroTechie said:

Welcome to CPF, dirodima!

You have too many of those 3.6V LEDs in series to power with 13.5V (and no series resistor? :duh2: ). Split it up in two strings of [LED + LED + resistor], and wire those strings in parallel.

Take the specs of those LEDs, search online for "led calculator" or similar, put in the values, done.

Click to expand...

Thanks a lot RetroTechie!

I will definately try that. My leds are 100Ma, so, i will use some 68 Ohm, 2 or 3 Watt resistors 2 by 2.

But out of curiosity, the power doesn't exceed 3.6V x 4. So, what happens, are multible leds in series unstable?

Again, thanks a lot!

3.6 x 4 = 14.4
too much voltage for 13.5 volt supply.
flickering is probably caused by car's charging system. they usually just go dim or out if not enough supply voltage.

The fact that the LEDs fail prematurely is not suggestive of having too many LEDs or dropouts in the charging system. About the only thing that can make an LED fail suddenly and catastrophically is a substantial overcurrent or overtemperature (usually caused by overcurrent) condition. A negative transient could be bad too, but I don't know what the failure looks like for that. Anyone care to chime in on that?

An automotive electrical system is not a friendly environment for LEDs. There are frequent and substantial voltage transients and surges that be deadly to LEDs. These are caused by a combination of factors. The alternator can be one of the worst culprits, when combined with a large load that can be suddenly removed, like the starter, highbeams, audio system, or (heaven forbid) a winch. Even the door locks can create a transient. A driver of some sort, even the series resistor suggested by RetroTechie is critical to long, reliable LED service. For low power lights, a transient suppressor and series resistor can provide excellent protection.

I often hear people mention that LEDs flicker before failing. That doesn't make sense to me, but I hear it often enough to believe that it must be real. Even the transients in an automotive system should rarely cause visible flicker. If anyone can explain the physics of it, please do.

Thank you all for you input.

In my (unlearned) opinion, 4 leds in series, typical 3.6 volt, minimum 3.2 volts, should work fine between 12.8 - 14.4 volt supply.

If i'm correct, this type of LED is "simply" an arc drawn between two fragile surfaces. Fragile to too much arc, meaning to much heat, from to high ampere, drawn due to too high voltage
(LEDs didn't exist when i went to school). i'm aware of the concept of ripple and transients, but shouldn't those be consistent over time to overburn the surfaces.

In a car battery, you have a huge capacitor, but maybe it is reacting to slow to protect against transients.

Would a small capacitor over the LEDs have this protecting effect?

The car battery is in fact comparable to a large capacitor. It does have impedance, especially as it ages, but it's the impedance of the wiring that's the worst culprit at high frequencies.

A capacitor will often not help at all, because it will primarily only shift the frequency at which the transient will ring. It may well reduce the maximum voltage (or not), but it will not reduce the energy available by much. It's very difficult to characterize the impedance of the system you are placing it in, making it just as difficult to predict the results.

A transient suppressor would be a much better choice, as it's designed to both limit the maximum voltage and absorb energy. If you could put some resistance between the transient suppressor and the LEDs, even better.

Maybe the worst thing that is somewhat likely to ever happen to your LEDs is to get a jump from your friendly local tow-truck. He has a 24 volt system and heavy cables with strong clips, and will raise the voltage of your electrical system way over 14.4V. Any component or system installed by an OEM or sold by a reputable third party is required to survive this. I forget exactly what the spec I'm working to right now says, but I think it's 26 volts for 5 minutes. I was more concerned with the 76V load dump the spec says it has to survive, even though this is an event the vast majority of vehicles will never experience.

Also, you need to understand that just because the typical voltage of your LED is 3.6V does not mean you can run it at 3.6V without any trouble. That's right. If you connect a 3.6V (typical) LED to a 3.6V supply, you may substantially overdrive the LED. Or you may substantially underdrive it. Unlike a regular light bulb which operates at a fixed voltage and draws as much current as it wants (a big rush at the beginning followed by a rapid settling down to a relatively constant value), an LED should be operated at a fixed current, and will settle at whatever voltage it wants.

The specs mean that if you put the rated current through it, it will typically settle around 3.6V, but it could be higher or lower, as low as 3.2V in some cases. If you have a case where rated current would result in a voltage of 3.2V, or even 3.4V, but you apply 3.6V, the LED will draw well above rated current. Unlike a light bulb, where a 10% increase in voltage results in less than 10% increase in current, in an LED a 10% increase in voltage results in substantially more than a 10% increase in current.