# Pulsating LED to increase current but keep heat down.

1. Jan 7, 2010

### soupnatizzle

i was thinking of using a 555 timer to create a low duty cycle TTL square wave to power an array of Inferred LEDS at a very high current for short segments of time.

basically i want to get really high outputs from an led and keep it from frying itself by turning it on and off really really fast to keep the heat down. possible?

2. Jan 7, 2010

### dlgoff

Sure it's possible. Less time on, less intensity also. Can you explain what your application is?

Welcome to PF

3. Jan 7, 2010

### Bob S

I would think that the light output is proportional to current I, while the voltage drop V across an LED increases at higher current. So the heating, = I·V, increases faster than the light output. If you have a series resistor R, the heating of the resistor equals I2R.
Bob S

4. Jan 7, 2010

### soupnatizzle

im looking to get much higher intensity light out of 10 inferred leds. im looking to setup a camcorder with inferred night vision and i want to get as much from the leds as possible.

so i want to pulse the input at say 60hz or higher to give the effect of constant on while increasing the current to the LEDS for a small duty cycle of say 10%, so that i can up the Amps to idk several times the rated value without destroying the LEDS.

something along the lines of, if you touch your hand to a hotplate for a few seconds o it has time to heat up and burn you, but if you quickly touch the surface of the hotplate for a fraction of a second you dont get burnt.

say i used a transistor to feed the leds current but keep the voltage = to 1.5 or so volts

Last edited: Jan 7, 2010
5. Jan 7, 2010

### Bob S

Is there any way to synchronize the LED pulses with the camcorder frame image readout. so you don't have the beat-frequency alternating image lightening and darkening? The commercial portable solid-state television camcorders do it.
Bob S

6. Jan 7, 2010

### soupnatizzle

i wouldnt know how to do that, but the camcorder records at 30fps

7. Jan 7, 2010

### mgb_phys

Niavely I would have thought this would be worse for heating.
If you pulse the LEd there must be a part of the cycle when you are ramping up the voltage into the LED but it isn't emitting (or not emitting fully) yet is still absorbing energy.
Wouldn't this lead to a greater power consumption in the diode than running it at lower power continually?

Overdriving an LED with pulses to get it to emit much more than it's rated power is common - but it's inefficient and you need to heatsink it to get rid of the waste heat.
Remember that when LEDs heat up you get a lot less light/watt out of them.

As Bob says running it at regular power for a shorter duty cycle in sync with frame could help, you will have a lower average power (and heat) but the same light as far as the camera is concerned.

ps, it's failry easy to detect an end of frame pulse from the video output even if the camera doesn't have a sync out

8. Jan 7, 2010

### whome9

The eye responds to peak brightness more than to avg. brightness.

9. Jan 7, 2010

### dlgoff

This may be of interest.
http://www.photonics.com/ShowImage.aspx?img=aHR0cCUzYSUyZiUyZnd3dy5waG90b25pY3MuY29tJTJmaW1hZ2VzJTJmc3BlY3RyYSUyZmZlYXR1cmVzJTJmMjAwNCUyZm1heSUyZm12bGlnaHRpbmdfbW9yaXRleC5qcGc=

Last edited by a moderator: May 4, 2017
10. Jan 8, 2010

### Mike_In_Plano

This might actually work, if you sync up with the frames.

Many of the old camcorders had a way to lock to external lighting to prevent beating against the room's fluorescent lighting.

If your just buying a separate camera board and using the camcorder's view finder, than you may want to make a sync detector circuit and turn off the LEDs during the sync portions of the video.

There's any number of sync detector or sync separator circuits floating around, and it would be a pretty cool way to keep your LEDs cooler/brighter, and/or increase your battery time.

Best Luck,

Mike