LED limit with zener.

[bandi12]

Hy. I bought some RGB LEDs ( 2 pin , it changes the color automatically with a small internal chip) , max voltage is 6V.
So i putted two LEDs in serial and powered with 9V , all good for couple of day's (4-5) after that one of them burned out , so i replaced it, after a couple of days again it burned out.

So I am thinking that the chip inside at a moment say's "i don't need current" ( for a short period of time) and it will "push" all the 9V in the second LED. So i was thinking to insert for every LED in paralel a zener so when one of the LEDs want to push all the 9V into the second LED then a 4V7 zener will open and limit the current to 4.7V so the secound one dosent burn out. Ofc all this "theory" from google/wiki , i am not shure if i made the circuit in the correct way ( and i realy don't want to burn out more LEDs) so somebody whit more skill can check it out ? Thx for the help.

Something like this.

 

[Gryd3]

Hy. I bought some RGB LEDs ( 2 pin , it changes the color automatically with a small internal chip) , max voltage is 6V.
So i putted two LEDs in serial and powered with 9V , all good for couple of day's (4-5) after that one of them burned out , so i replaced it, after a couple of days again it burned out.

So I am thinking that the chip inside at a moment say's "i don't need current" ( for a short period of time) and it will "push" all the 9V in the second LED. So i was thinking to insert for every LED in paralel a zener so when one of the LEDs want to push all the 9V into the second LED then a 4V7 zener will open and limit the current to 4.7V so the secound one dosent burn out. Ofc all this "theory" from google/wiki , i am not shure if i made the circuit in the correct way ( and i realy don't want to burn out more LEDs) so somebody whit more skill can check it out ? Thx for the help.

Something like this.

View attachment 17099

Do you have a data sheet, or link to the LEDs that we can see?
You could most likely solve this with a single resistor... but I want to try to figure out how that internal IC operates.

 

[bandi12]

It's "OptoSupply OSTB5131A"

This is all i can find : http://datasheet.octopart.com/OSTB5131A-OptoSupply-datasheet-26746456.pdf , but it dosen't give any info about the IC.
And it work's like this :

 

[Gryd3]

I think your best bet is to operate them in parallel within the rated voltage of 4.5V - 6V (Use 3-4 AA, or AAA Batteries)
I read through that little document and I cannot tell if that LED simply alternates between the 3 LED dies within, or if it does actually bypass the LED dies that would submit the other LED to a higher voltage.

..Also, the LED in the video is operating on 3V... which is below the minimum rating on the spec sheet provided. The video is simply to show what the LED behaves like, and is not the actual LED you have?

 

[bandi12]

It's the same LEDs i have. Well they are inside a kind of chandelier with glass cube globes, it has three arms and each arm has 2 globes. So on each arm I have 2 of them in serial and after that the pair of two are in parallel with the other arms. Each globe has 4 LEDs so in total are 24 LEDs. The problem is that i have another LEDs to ( white , power LEDs 1W and 3W) so each arm has a lot of wires and it will be pretty hard to fit an extra one in to link all of them in parallel. ( I have a 3A regulator , in parallel all of them will "eat" 2.5-2.6A si the regulator will be pushed to the limit , so bk of this i chose to group them by two).

So i think i need to mount it down fit and extra wire , switch the regulator to a 5A one.

 

[Gryd3]

I am still unsure on the exact cause of the failure.
Using the zener like you proposed will clamp the voltage across the LED to a specific voltage to prevent it from going any higher, but the zener will essentially short out the difference in voltage and a lot of current will flow through the zener. This will either kill the zener, or shorten it's life considerably which would then lead the the LED dying again. You can use a resistor to limit the zener current, but this resistor will need to be in series with the LED as well which could very well stop it from working.

Without knowing the exact cause, I can only recommend using them in parallel unless someone else has a better idea.

If this is out of the question, you could grasp for straws and install a 100Ω resistor for every pair of LEDs you have.
The LEDs will draw anywhere from 10-40mA which would cause the resistor to drop the 9V down between 5V - 8V .. it's a wide operating range .. and will drop the voltage more when the LEDs are pulling more current, and less when they aren't.
The data sheet also does not list a Forward Voltage Drop... but an operational voltage, so it's hard to determine how much voltage actually gets dropped across each LED...

 

[bandi12]

On an other forum somebody also recomanded that zener methode. Well i think i will kill two more LEDs , wire up with the zenner , hook it up on a 12V battery with a 9V regulator , and leave it ON for 2-3 day's , let's see if something will smoke up , if yes then i will rewire all of them in parallel.

 

[Gryd3]

On an other forum somebody also recomanded that zener methode. Well i think i will kill two more LEDs , wire up with the zenner , hook it up on a 12V battery with a 9V regulator , and leave it ON for 2-3 day's , let's see if something will smoke up , if yes then i will rewire all of them in parallel.

If you have a multi-meter, measure the voltage across one of the LEDs and see what it's at.
This will give you more details into how the voltage is dropping across the LEDs when you connect those things in series.

 

[BobK]

It should not take any more wires going into the globe to put them in parallel. Just run two wire from outside to the first one, then two wires from the first to the second.

Bob

 

[bandi12]

Well , actually i would need to put only one more wire from the outside globe (there the 9V goes) to the second globe on the same arm. Then in the second globe cut the wire from the first LEDs negative pin and link it to the ground link the LEDs + in together and link it to the new wire. The wire outside of the arm's it will look ugly ( the arms are holding the globes and they are +/- 1 cm square and maybe 20-30 cm long and each globe is 15 cm from the other one on the same arm)

I will try to measure the voltage , but i don't think the multi-meter will have time to measure the "very fast" droppings of the IC, maybe the IC send the 9V to the second one only for 10-20 ms.

 

[Gryd3]

Well , actually i would need to put only one more wire from the outside globe (there the 9V goes) to the second globe on the same arm. Then in the second globe cut the wire from the first LEDs negative pin and link it to the ground link the LEDs + in together and link it to the new wire. The wire outside of the arm's it will look ugly ( the arms are holding the globes and they are +/- 1 cm square and maybe 20-30 cm long and each globe is 15 cm from the other one on the same arm)

I will try to measure the voltage , but i don't think the multi-meter will have time to measure the "very fast" droppings of the IC, maybe the IC send the 9V to the second one only for 10-20 ms.

You may see a more slight gradual change as it alters color though.

 

[bandi12]

Yes , at the begin it draw's 25-30mA ( only red ) and at the cycle end ( all three colours ) it draw's around 85-90mA. I will measure the Voltage to tomorow.

And that guy who told me about the zener's , say'd that in the datasheet there is "Quiescent Current = 1uA" meaning that the IC has a "blocked" moment when most possible the IC dosen't command any diodes so all the current will be forced to the secound LED.

I will try how he told me with the zener's and let's see what happend's