I have looked around and know that you can pulse a LED at a higher than normal forward current, and due to how our vision works, it will appear brigher than at a constant current. The example I saw was 200% current in a 50% duty cycle. But the thing I was wondering was, how high can you push the LEDs, and can it be something like that example, or does the LED need more "cooling" time than that would allow.? One datasheet for a LED I saw said the peak forward current was 60mA compared to the standard forward current of 25 mA. The note at the bottom said that the peak current was a 1mS pulse in a 1/20 duty cycle. So 2.4X the current, (and hence brighness) but 1/20th the on time? Any one have any practical experience with this, rather than just web pages explaining the theroy, without providing any solid numbers. Ive seen some app notes, too, but for some reason I just dont thing that a LED could handle 4x current in a 25% duty cycle. I know it works math wise.. but there is no such thing as a free lunch so... If things work out right. I might eventually make a strobe light with LEDs, and it would be nice to know what kind of pulses they can handle. Would a .001 mS pulse at 10-15X current, and a obscenely slow duty cycle, adjustable from 1-120 pulses per second be feasable? Obviousley more than one LED would be used, but if I can get much more power from a cluster of the same size, that would be nice. Other projects too... just that the stobe might be fun to have and mess around with.
Not sure, but one thing I would be worried about is if the driver froze. I thought about doing this for my car lighting, but I worried what would happen if it glitched and they got 100% duty. Some of them are in hard to get at positions and I didn't want to replace them.
Being that LED's themselves are pretty cheap...why not just try it and see what happens Worst you could do is blow a LED or two.
i know that they do this on high end flashlights to save power, but have the same brightness... so if less power=same brightness, same power=more brightness and IMO the answer to your "nothings free" question is "in this case it is" LEDs are just more effecent in a pulsating mode.
Pulsing light is a common technique to improve the signal to noise ratio in measurements. You illuminate the item very bright for a short moment and "gate" the sensor so it is only active when the light is on. Here is an example http://www.sciner.com/MCP/GIC.htm device. The human eye is somewhat similar, but on a slower time scale (sorry, no good link). That's why you can see better with high intensity pulses than with constant low intensity illumination. In addition pulsing a LED allows to run it at its most efficient current during the pulse. Many types of LEDs overheat if that current is applied constantly. Here is an accumulation of LED info in simple terms: http://ipeqwww.epfl.ch/qd/pdf/ASLS10_Stanley.pdf . Pulsing the LED is not that risky. A simple NE555 makes a decent oscillator for experimenting. If you are really concerned about the oscillator "freezing" - e.g. because you use a software generated oscillator signal - couple it through a capacitor. That way a DC signal won't reach the LED. In that configuration it is advisable to use two LEDs anti-parallel, so one uses the charge current and the other the discharge current of the capacitor.
That sort of pulsing is handy for sending a coded signal (like a remote), as the range will be longer than if the max continuous current was used. Kingbright's specs seem more optimistic than your example, they base their pulse ratings on 0.1mS 10% duty, and, for example, some blues and whites are 30mA continuous, 150mA pulsed; 5x current, 1/10 time. Their hyper-red, they say 200mA pulsed. I've made a 555/single led strobe, but didn't push the current much, must dig it out & test to destruction.
Now that has to be one of the most awesome sayings I've heard all year You might want to copyright that term
Ok I just was wondering what I should expect. I have a feeling I'll be seeing spots here soon... Hmmm I guess I'll get a bunch an play around... as for varying pulse widths... I'll just use stuff that I have laying around... a few transistors.. some caps... | I wonder what will provide the pulse signal oh well for overkill V
I thought i should fill in a little here, i had a look at this a while ago, and it seems that while pulsing is really efficient for IR LEDs, it's not supposed to be as efficient for shorter wavelengths. The IR LEDs basically put out double the light at double the current if done in short pulses. The visible light LEDs are supposed to output a lot less than double the light if the current is doubled. From that i concluded that PWM controlling is good for lowering the light output, but it's not that efficient att raising the output. This information was from a LED manufacturer, and ofcourse i do not still have the link... As for duty cycles and maximum forward currents, look at the data sheet.
I think all the LED data expressed above is based on data sheets. Every LED (from manufacturer to manufacturer) is going to be different as far as peak current and duty cycles. I don't think they would list it as a spec if it hadn't previously been tested at some point. I believe IR LEDS are pulsed in that fashion to allow for a more "usable distance", but I'm not entirely sure about that.
I think that with one LED, it would be like a 1 cylander engine, not too good. but get 10 LEDs, that meens at least 1 is alwase on, and its like a V10, very good
theshadow27, 1 LED always on out of 10 is defeating the object of this trickery. The eye perceives changes in light very quickly, so by lighting an LED at a low duty cycle the LED will appear brighter than an LED driven constantly at the same current. The numbers in question are always printed on the datasheet and will vary from manufacturer to manufacturer. There is some truth in what Ghlargh said. This increased brightness is much more prevalent for IR LEDs than with standard LEDs, infact the effect is only really noticable at low (<20mA) currents in standard LEDs. Beware of slow rise/fall times from your oscillator if you're pushing the LED at high currents - it's very easy to burn out an LED at high currents because the LED is on for longer than you calculated...
