Go Down

Topic: Hot LEDs (Read 2292 times) previous topic - next topic

Fej42

Hey guys!

What I'm doing is: I'm making a sensor bar to use a Wii Remote as a cursor. It's rather simple, here's the schematic:



It uses:

- A USB jack for power
- A toggle switch to turn it on and off
- A blue power LED, supposed to run at 5v with 30mA
- A 1 ohm, 1/4 watt resistor for the blue LED
- Two IR LEDs, packaging says to use 1.2v, website says to use 1.28v, both say 100mA
- A 33 ohm, 1/2 watt resistor for the IR LEDs

I used an online calculator to get the resistor values.

My problem is, the IR LEDs are getting really hot. They're on top of a piece of plastic (monitor), and if they run for more than 2 minutes, I'm afraid that they will melt the plastic. How can I cool these down? Are heatsinks necessary? Or should I just get a bigger resistor?

Thanks everyone!

nexekho

I'd imagine they're supposed to get hot, the ones in the official sensor bar get very hot even through the case.  IR LEDs are usually only used for short bursts in the case of remote controls.

Techone

@Fej42

For the IR Led : ( 5 V - ( 1.28 + 1.28 ) ) / 100mA = 24.4 ohms    33 is OK   ( 5 - ( 1.28+1.28 ) / 33 = 73.939393 mA or 74 mA

Power dissipation :  1.28 * 74 mA = 94.72 mW per IR LED    The resistor : 33 * 74 mA ^2 = 180.708 mW

For the Power LED :  5 V * 30 mA = 150 mW    The resistor : 1 * 30 mA ^2 = 0.9 mW


As for getting hot, well that one is ???  According to calculation is not suppose to get THAT HOT !!!  The IR led is lower than 1/4 Watt power dissipation.  If the current is correct.

Check for short, maybe LED is faulty ( shorted ? )  Check for current --> put a DMV in  series with the circuit. Plce the DMV in mA Mode. And check voltage too.  Use an another supply voltage, NOT your computer !!!       

@nexekho

According to calculations... NO nut suppose to get hot.

Jack Christensen

Is the blue power LED getting hot? It would be an odd one if it really is supposed to run on 5V directly. Not sure I'd trust Radio Schlock's specs.
MCP79411/12 RTC ... "One Million Ohms" ATtiny kit ... available at http://www.tindie.com/stores/JChristensen/

dc42

Check that the series resistor you are using for the ir leds really is 33 ohms.
Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

nexekho


@nexekho

According to calculations... NO nut suppose to get hot.


Well the official sensor bar does so I'm not sure what your argument is here.

TFel

100mA though a 5mm LED will always get hot.

If you are worried about the heat affecting your monitor, put them in an enclosure or lower the current.

Fej42

I had slightly bad data - the PC is putting out ~4.5v, not 5v.


Is the blue power LED getting hot? It would be an odd one if it really is supposed to run on 5V directly. Not sure I'd trust Radio Schlock's specs.


Nope. It's practically cold.


100mA though a 5mm LED will always get hot.

If you are worried about the heat affecting your monitor, put them in an enclosure or lower the current.


How low should I make the current? 50mA? I'm worried that the LEDs won't be bright enough, but hey, I can't see them, so all I can do is trial and error, I guess.

Also, would this enclosure provide any kind of heatsink capability?

TFel

I was thinking more along the lines of a plastic project box you can put them to insulate your monitor from them. If you mount the LEDs on a PCB and mount the PCB inside a project box with a cutout in the side for the LEDs to poke though, you should be fine.

pwillard

If you have a camera... take a picture of the bar when ON.  Some cameras can see that wavelength.  Worth a try.

Techone

@Fej42

Can you measure the current of the IR LED and voltage across the IR Led ?  I am curious to know why it is hot. It is hot like, "I burning my finger here ..." or just warm... because a 1/4 Watt at the IR LED ( according to calculation ) can not be that HOT .. I am having a hard time to believe this. A 100 W light buld is " I am burning my finger here ..." is HOT to touch, but NOT 1/4 W !!

I know you can not see the IR LED, but you can still measure the current & voltage.... TIP : place a DMV ( in current measument mode ) in parallel with the switch to measure the current. Just don't turn ON the switch, the meter will connect your circuit. But, you will get a total current value ( according to Kirkov Current Law -- I total = I 1 + I 2 )   

dc42

What's not clear to me is whether the 100mA is a continuous rating or a pulsed one.
Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

MarkT

Firstly IR LEDs are rated for pulsed operation I suspect - that's how they are almost always used.

Secondly LEDs voltage ratings are very approximate (wide variation with batch and temperature).  You should not be using a 1 ohm resistor for the blue LED, that will risk overloading the USB port.  To drive LEDs you want to set the current, not the voltage, hence the common practice of a series resistor and a power supply voltage greater than the LED requires (typical blue LEDs are about 3.2 to 4.0V range at room temperature BTW, suggesting a 47ohm series resistor - that Radio shack one seems unusual - perhaps its violet instead of blue?  Perhaps the specs are nonsense?).
[ I won't respond to messages, use the forum please ]

Fej42


@Fej42

Can you measure the current of the IR LED and voltage across the IR Led ?  I am curious to know why it is hot. It is hot like, "I burning my finger here ..." or just warm... because a 1/4 Watt at the IR LED ( according to calculation ) can not be that HOT .. I am having a hard time to believe this. A 100 W light buld is " I am burning my finger here ..." is HOT to touch, but NOT 1/4 W !!

I know you can not see the IR LED, but you can still measure the current & voltage.... TIP : place a DMV ( in current measument mode ) in parallel with the switch to measure the current. Just don't turn ON the switch, the meter will connect your circuit. But, you will get a total current value ( according to Kirkov Current Law -- I total = I 1 + I 2 )   


I can't measure the current right now, but it was definitely "I'm burning my finger here". Heck, it's possible it could have gotten to "it's melting the tape on it right now".


Firstly IR LEDs are rated for pulsed operation I suspect - that's how they are almost always used.

Secondly LEDs voltage ratings are very approximate (wide variation with batch and temperature).  You should not be using a 1 ohm resistor for the blue LED, that will risk overloading the USB port.  To drive LEDs you want to set the current, not the voltage, hence the common practice of a series resistor and a power supply voltage greater than the LED requires (typical blue LEDs are about 3.2 to 4.0V range at room temperature BTW, suggesting a 47ohm series resistor - that Radio shack one seems unusual - perhaps its violet instead of blue?  Perhaps the specs are nonsense?).



It's blue, and the specs don't seem nonsense. It's a weird LED I guess.

Anyway, I was able to fix the problem. I replaced the 33 ohm resistor with a 47 ohm one. It basically halved the current. It also cooled the LED significantly. I looked at it with a camera (thanks pwillard!) and they seemed plenty bright enough. They work fine.

Thanks for all the help everyone!

Don't forget that IR is just another word for HEAT!  You're driving these IR LEDs pretty hard (20 mA is probably more than enough) and they are emitting a lot of IR radiation!  Your finger perceives that radiation as heat!  As mentioned earlier, they are usually pulsed and not left on continuously - that will only make them appear hotter.

Go Up