LED's and constant current

[user4572]

I'm playing with an array of 6x5 LEDs. I wanted to see what the total current draw would be will all of the LEDs on.
When I first fired it up my meter read .9a and then as I left them on it slowly crept up to 1.5a

Do LED's draw more current as they are left on?

I thought it was supposed to be a constant draw...

I'm going to use them as a flasher unit so the wont be left on all the time but I want to make sure I get it right the first time..

Thanks for any input..

 

[BobK]

Yes, they are supposed to have a constant current flow, but they do not do that themselves, you need a circuit that does that.

It sounds like you supplied the LEDs with a constant voltage. As they heat up they will draw more and more current at the same voltage. This is called thermal runaway and it is why you cannot simply apply a voltage source to an LED.

Read our resource on driving LEDs.


Bob

 

[user4572]

Yes, they are supposed to have a constant current flow, but they do not do that themselves, you need a circuit that does that.

It sounds like you supplied the LEDs with a constant voltage. As they heat up they will draw more and more current at the same voltage. This is called thermal runaway and it is why you cannot simply apply a voltage source to an LED.

Read our resource on driving LEDs.


Bob

Sorry I should have been more specific. I have my board setup like #4, multiple strings in parallel.. My other creations have been 2 rows of 3 LEDs, each row with a resistor and I didn't have any issues. This one used more LED's and resistors so I wanted to make sure I didn't mess it up.
I even used the LED calculator that's in the link to make the circuit.

I fired it up again and noticed that a couple LEDs are not as bright as the others and 1 is flickering. I'm going to replace those tomorrow.

 

[(*steve*)]

I'm not sure what you mean by #4.

can you post the schematic?

 

[Harald Kapp]

As they heat up they will draw more and more current at the same voltage. This is called thermal runaway and it is why you cannot simply apply a voltage source to an LED.

True. And applies partly to current limiting by a resistor, too. As the voltage across the diode drops, the voltage across the resistor increases, driving more current through the circuit. Nevertheless, due to the non-linear characteristic of the LED thermal runaway is avoided because a small change in LED voltage will lead to a huge change in current and thus a huge a huge change in voltage across the resistor which in turn will counteract the change in voltage across the diode.

multiple strings in parallel

Each string should have its own current limiting resistor.

 

[BobK]

If the LEDs are properly connected with resistors, I cannot see the current slowly rising from 0.9A to 1.5A. How about a schematic?

Also, are they heat sinked? High power LEDs need heat sinking, and the star board they might be on is not a heat sink, it is used to mount them to a heat sink.

Bob

 

[user4572]

I've uploaded the schematic I used from the led wizard and a picture of the board. This is the same setup I've been using for a couple years now.
My mistake..its 6x4 not 6x5...
I've never needed a heat sink before with these LEDs. they are each .5 watt LEDs
Normally these are not left on. They flash so they do build up a little heat over time but never like what I've seen and always a constant current draw.

 

[(*steve*)]

Go ask the wizard for the resistor value for 5 LEDs in series. The 6.8 ohm resistor is too small to be effective in limiting the current. This is covered in the resource on driving LEDs.

 

[BobK]

I hate those LED calculators. I have seen at least one that will recommend 0Ω for the resistor if the forward voltages add up to the power supply voltage.

And BTW 2.16V x 100mA is not 0.5W.

Bob

 

[user4572]

I've been using that calculator for a couple years now without issue. This is also the biggest LED flasher I've built.
My last class in electronics was back in the early 90's and then I switched to computers so this is now a hobby for me. I guess I'll have to find my old books and verify everything before I do anymore big projects.

Thanks

 

[BobK]

No, you have an an issue right now. It led you to believe that a 6.8Ω resistor dropping 0.6V was enough to control the current to your LED chains. It clearly was not.

The effectiveness of a resistor as a current limiter depends on what percentage of the source voltage it is dropping. Somewhere around 90% would make it very effective, 25% would make it suitable, 3%, which is what you design does, not not effective. The calculator should have told you this.

Bob

 

[Herschel Peeler]

Sorry I should have been more specific. I have my board setup like #4, multiple strings in parallel.. My other creations have been 2 rows of 3 LEDs, each row with a resistor and I didn't have any issues. This one used more LED's and resistors so I wanted to make sure I didn't mess it up.
I even used the LED calculator that's in the link to make the circuit.

I fired it up again and noticed that a couple LEDs are not as bright as the others and 1 is flickering. I'm going to replace those tomorrow.

Re: "I fired it up again and noticed that a couple LEDs are not as bright as the others"
That sounds like you over stressed the LEDs.

 

[(*steve*)]

The calculator should have told you this.

I agree, "should". But find me one that does, or even better, one that refuses to offer this as a solution

 

[BobK]

My point exactly.

Bob