Heat generated by LED's?

davehal9000 · Post by **davehal9000** » Tue Feb 03, 2009 3:58 pm

I'm looking at doing my first model lighting job and am concerned about heat buildup melting the model. I'm looking at one rig with three LED's, another with one, both with their own 9V batteries. Both I believe will be completely enclosed on the plastic of the model. Any need to worry, or should I find a way to route everything through fiberoptics instead?

TREKKRIFFIC · Post by **TREKKRIFFIC** » Tue Feb 03, 2009 4:19 pm

LED's are quite cool especially when compared to incandescent bulbs. You should be fine.

photoguy · Post by **photoguy** » Tue Feb 03, 2009 5:38 pm

And 9V is probably WAY more voltage than you'll need. I just finished a lighting project with twelve LEDs in parallel, plus a flickering LED circuit, and a blinking 555 timer running two more hi-output LEDs, all off of a 3V wall wart. I would think 9V isn't going to give you much, except a shorter LED lifespan.

Stilgar · Post by **Stilgar** » Tue Feb 03, 2009 6:25 pm

FYI...I did the PL 350 Enterprise with hundreds of LEDS and it uses a 5V power supply....
I have never noticed any heat issues. (But I dont leave it turned on for hours....)

http://home.comcast.net/~stilgar99/

Stilgar

Kylwell · Post by **Kylwell** » Tue Feb 03, 2009 6:47 pm

9v batteries also have a stunningly low amperage resulting in low run times. Get a 4 AA or AAA holder and run with that.

Sparky · Post by **Sparky** » Tue Feb 03, 2009 6:55 pm

Check out the youtube video about 9 volt batteries. They are made of AAAA cells I believe. So in a pinch you can field strip a 9 volt to get very low run time 'A' size cells (fits AA,AAA,or AAAA) for a critical device needing them.

Just an interesting fact about 9 volts.

Pat Amaral · Post by **Pat Amaral** » Tue Feb 03, 2009 7:01 pm

You didn't say if you have a current limiting resister in line with that LED that's running off 9v by itself. With that much voltage, you're going to need one, otherwise the LED WILL get hot (but only for a second before it quickly burns out

). You also didn't say what color the LED is.

White ones usually only need about 3.6v (red ones need about 2v). Assuming you're using white, you'll need to put a 270 ohm resister in line between one of the battery connections and the LED to keep the current within safe limits. If the LED is red, the resister should be 350 ohms (360 or 390 ohms might be easier to find and will still work). Build the circuit like I've described and it should run cool enough.

Here's the math:

9v - 3.6v(white LED voltage) = 5.4v

5.4v / 20mA (current needed to drive the LED) = 270 ohms (resister value).

or

9v - 2v (red LED voltage) = 7v

7v / 20mA = 350 ohms (like I said before 360 or 390 ohms are more common and either should provide sufficient current)

If they're white and they're connected in series (connected in line with each other) the three LEDs in the other circuit should be fine without a resister although they might run a little dim which shouldn't be a problem. If the LEDs are red though, you should use a 150 ohm resister to make sure those LEDs run cool. Without a resister, they might work ok for a while but they won't last very long and they will most definately run a little hotter. If these are being run in parallel, you'll need to use the rules for one LED like above.

I think I got all that right. Again, you didn't provide certain details so I wanted to make sure this was covered. If you've already accounted for the resister, great! Maybe this will help somebody else.

Cheers,

Hope it helps,

Sparky · Post by **Sparky** » Tue Feb 03, 2009 7:06 pm

LEDs do run color, but high output ones still get very hot. Not that modeler's would need or wnat to spend the money on those types of LEDs.

If you use a dropping resistor, please remember the power dropped in the resistor is converted to thermal energy. It's best to design the power system (batteries or ac converter) and the LED configuration for efficient use the voltage and current.

But the short answer is as mentioned, LEDs in a sealed model will run very cool.

Pat Amaral · Post by **Pat Amaral** » Tue Feb 03, 2009 7:13 pm

For the record, 1/4 watt resisters should be fine for the applications that I described above. The circuits as decribed shouldn't generate much heat at all.

davehal9000 · Post by **davehal9000** » Tue Feb 03, 2009 10:02 pm

One was a ultra bright blue LED with a 470 ohm resister. The other will have three blinking LED's, two blue and one white, I believe. I'm trying to get sort of a flickering effect with the latter, but I may be barking up the wrong tree on that one.

Sparky · Post by **Sparky** » Wed Feb 04, 2009 12:08 pm

you should be fine with that number of LEDs and plan dropping resistors. Its the 200 + LED setups that need to be laid out so that waste power is not an issue.

Here's a circuit that I was able to get to work with an LED, tweaking the bias resistors. We wanted an LED that could )through fiber optics) appear to be an arc or plasma cutter in operation.

http://www.awrr.com/flicker.html

The one thing I observed is that you should not look directly into the LED will fine tuning or event trying to observe the effect. The subtle flickering will be lost in the larger pulses, (eye saturation maybe). If you took an LED wired to the circuit and shined it on a wall you could see much more 'random' flickering form bright spots to more subtle surges in light.

davehal9000 · Post by **davehal9000** » Wed Feb 04, 2009 2:16 pm

Wish I understood this stuff. It's all Greek to me.

Sparky · Post by **Sparky** » Wed Feb 04, 2009 3:01 pm

What effect are trying to achieve with the 3 flickering LEDs? Sparks/explosion or engine glow/throb?

davehal9000 · Post by **davehal9000** » Wed Feb 04, 2009 3:49 pm

Engine glow/throb, with a blue LED or combination of LEDs. I found a site that had simple enough plans to do the other lights I need, instruments and all, but the plans they had for the engines seem too repetative, looking more like a strobe than flickering.

Sparky · Post by **Sparky** » Wed Feb 04, 2009 4:08 pm

Here is a throbbing circuit in operation. Well 2 circuits the Yellow LEDs are in series and have their own circuit set to one throbbing rate, and the red LEDs are in series on another throbbing circuit with its one settings:

http://www.kc6sye.com/images/images_05_ ... _small.avi

davehal9000 · Post by **davehal9000** » Wed Feb 04, 2009 4:17 pm

http://www.hobbylinc.com/htm/mtr/mtr10010101.htm

Found this one. Should be able to change out the red for blue and white

Sparky · Post by **Sparky** » Wed Feb 04, 2009 4:33 pm

Those appear to be light bulbs and not LEDs. It might even, basically, be the same circuit from the other railroad sight.

Edit;
I checked Walther's website and they have a much closer-up view, the bulbs are light bulbs and not LEDs. The flicker effect may not translate to LEDs (I'm 90% sure that it won't).

Filament based lights generate light over a different range of voltage/current, so a circuit setup to generate flickering levels of voltage/current in a light bulb, will not work for LEDs. The circuit can do it for LEDs but it has to be re-tuned so that the level it is flickering with is in the right spot for an LED.
It won't work for an LED out of the box the same as it does for light bulbs.

Madman Lighting · Post by **Madman Lighting** » Wed Feb 04, 2009 7:00 pm

Sparky,

I looked at the minor creation in the video, yours perhaps?

Yes, I'm very impressed.

-John C.

Sparky · Post by **Sparky** » Thu Feb 05, 2009 10:32 am

The design was Tony Augustin. I just put in the neon rope to hide the obvious pipe threads, and the LEDs replaced some other gribblies that kept popping off every time we opened the neck joint or even tightened it.