2 LM317 in series each driving a 10W Led

[tesla952]

Hi all...

Premiss (i will exclude forward voltage as i use them current regulated, but both of them have different voltages):
-1x 10W Blue LED (Current 1300 MAh)
-1x 10W Red LED (Current 1300 Mah)

The two leds are mounted on an old CPU heatink for dispersing the heat.

The situation is: I have a power supply of 15V 5A, and have two LM317, each driving a led. First thing i did is that i connected the LM317in parallel, and it wasnt working. Then, just for curiosity i connected the LM317 in series and it was working, so i left it like that.
Reason i'm using 2 LM317 is that the leds use each different voltages, and that LM317 handle only up to 1.5A .


Problem is that the circuit is not working as expected. isuspect that the RED led is drawing all the current and sabotaging it for the blue led which, after a while, dims and the red led gets all bright.


What do you think of my circuit? How could i connect two 10W leds in a better way? Do anybody know where the problem with my circuit? (Yes, i know i probably sound like a noob, but by making mistakes we learn a lot more )

Thanks in advance for any help

 

[KeepItSimpleStupid]

Look at the datasheet for the LM317. There is a suggested way to turn it into a current regulator with a single resistor. That's what I would do.

 

[tesla952]

I did, i made it a current regulator by using a single resistor. But i still have the problem when i connected both in series :/

 

[ronv]

Maybe you could post your circuit?

 

[ronsimpson]

I can not see what you are doing!
"I did, i made it a current regulator by using a single resistor. But i still have the problem when i connected both in series :/ "
Normally you don't connect current regulators in series.
Never parallel LEDs! Never nerver parallel LEDs of different colors.

If you drop 15 volts down to about 3 volts (at 1.3A) then the LM317 will run HOT.

 

[KeepItSimpleStupid]

Did you use the "Precision Current Limiter" circuit on page 19 https://www.electro-tech-online.com/custompdfs/2011/01/LM117.pdf

MAh is typically Milliamp Hours. Is that the correct unit? Do you mean 1300 mA?

 

[tesla952]

LOL! Yes i meant mA, it was night over here and i can now see that maybe i was too sleepy to do serious things on the computer

Here comes a scratch of the circuit i made in Paint on the netbook (i dont have any serious software on this pc)..

Reason why i thought of using 15V 5A is that ( i might be wrong) i thought that with current regulators the voltage wasn't important since it regulated the current itself and that the dropout voltage was still there as when it's used as a voltage regulator.

Using the 1.3Ohm resistor the LM317 should give and ouput of ~0.96 A with around 1.2W dissipated so it should be fine for driving the led.

Details about the leds:

- 10W Red 600 lumen (Forward voltage 10V, Fordard Current 1000 mA)
-10W Blue 600 lumen (Forward voltage 10-12V, Forward current 1000 mA)

What do you think i missed out in the circuit?

Thanks again for any help

 

[KeepItSimpleStupid]

1. Bet the LEDs are getting to hot. They need to be heatsinked.
2. Regulator is oscillating.
3. Regulator shutting down. No heatsink.

Why a cap across the LEDs?

 

[MrAl]

Hello there,

Since BOTH of your LEDs use the very same current (presumably 1300ma) then you can connect them in series and use ONE single LM317 to regulate the current rather than two. The current will be the same through both LEDs.
Since the voltage is around 3v for both LEDs, that will be around 6v for both which means the regulator will drop about 9v at 1.3 amps which is more than 10 watts of power to dissipate. That's quite a waste of power. You'll need a nice heatsink for the LM317 too if you really want to run it that way.
To get some of the power out of the regulator you can use a dropping resistor or a few 3 amp diodes in series, but a nicer way might be to just add yet another LED in series which will bring the string up to around 9v and waste less power. You may even get away with four LEDs in series which would mean more brightness and less power wasted.

Of course the LEDs themselves absolutely need the proper heatsinking too.

 

[ccurtis]

A couple of things:

1. The blue LED can require up to 12V to operate which is 3 volts less than your supply voltage. The 317 can typically operate with a 2 volt difference, so you are running close there. Make sure you have a true 15 volts with little ripple when both circuits are connected to your power supply.

2. From the data sheet: "Although the LM117 is stable with no output capacitors, like any feedback circuit, certain values of external capacitance can cause excessive ringing. This occurs with values between 500 pF and 5000 pF. A 1 μF solid tantalum (or 25 μF aluminum electrolytic) on the output swamps this effect and insures stability." You have installed 0.1 uF.

 

[KeepItSimpleStupid]

MrAl:

If the LED's have a Vf of 10V then 3 V+2*10 V > 15 V so no go.
The circuit the OP has is not a parallel regulator circuit. They are independent. As designed it may have issues, but it's a good start.

The OP should pay attention to the drop across each regulator as it begins to fail.

 

[nulldevice]

Do you have the mounting tabs of the LM317 regulators on the same heat sink or otherwise electrically connected? If so you have the outputs connected together. Not good if so. The mounting tab and output pin are one and the same.

 

[MrAl]

MrAl:

If the LED's have a Vf of 10V then 3 V+2*10 V > 15 V so no go.
The circuit the OP has is not a parallel regulator circuit. They are independent. As designed it may have issues, but it's a good start.

The OP should pay attention to the drop across each regulator as it begins to fail.

Hello again,

Yes i tied to edit the post but it would not let me change the response for some reason.

With 10v LEDs and a 12v supply nothing will work using an LM317. With a 12v LED even worse.
The only way to regulate a 12v LED with a 12v supply is to use a regulated boost circuit.