LEDs for less power consumption

My remark about LED death from dimmers applies only to AC bulbs with an internal capacitor.
LEDs on 12V can be easily dimmed by reducing the current with a variable resistor, or if the inevitable loss of energy bothers you, by switching the LED on and off very quickly with variable ON/OFF duration.

 

Ah, so pulse width modulation works on LEDs? (on and off very quickly)

I'm a big fan of pulse width modulation and have been using it with halogen lights for many years now, as well as to charge my batteries.

 

Yes!
LEDs with pulse modulation can even reach much higher brightness levels than on DC.
Any LED with 25 mA maximum current can be supplied 250 mA at 5-7 % duty cycle giving twice the brightness.

 

CDK I (hardly ever) argue with you but can you point to some data on this?

I would have thought that "area under the curve" would be closely proportional to perceived brightness.

Or IS this "Human perceived brightness" (Which is what we probably care about) VS radiometric received energy ??

Hmmm....

 

Hi Terry,

Easy to explain, if you drive a LED at double it's rated power for half the time so fast your eyes do not see it go on and off then you achieve higher brightness. The LED's watts remains the same so it lives.

BUT

you guys must be carefull of switched LED drivers. They switch these things at high frequency for more efficiency etc etc and there were cases where the emmissions generated as a result inteference with your ship's instruments.

It's going to be somewhat shitty if you can use either the lights OR the GPS, or lights OR the radio...

 

Most opto manufacturers use the phrase "luminous intensity" without further clarification; I've never had the need to dig deeper.

White LEDs are in fact UV LEDs with a coating that converts UV to visible radiation.
The radiation output vs forward current graph is almost a straight line between 5% an 100% current and usually stops there because the point of maximum dissipation is reached. But beyond that the output stays linear with the current until at least 10 times the maximum value for DC application.

In pulse applications the LED can be 10 times as bright, but only with a small duty cycle so the maximum chip temp is not exceeded. Although the chip itself is switched off 95% of the time, the converting phosphor is much slower and can be "pumped" to a brightness far above that in DC mode.

I am a bit pressed for time right now, so I cannot point to any data to substantiate this, but if you have a NE555 or similar chip a test is quickly done.

 

Here is a better and easier circuit than a 555. Use a 4093 nand gate.

The diode charges the cap up faster through the lower value so on time is less tha 1/10th of the on time. You will have to play with the values to get the ducty cycle you require, the switch timing is the same as that given in the 555 data sheet... it's only a resistor charging a cap.

 

Oh, now that makes sense. Hmmmm. Thanks.

I can easily do some different duty cycles with an Arduino, and I'll try it out.

 

Maybe LEDs would work in opaque projector!

 

PWM is great if you want to "pump" the most brightness out of an LED but if you are considering "dimming" the LED's, why not use an adjustable voltage regulator like the LM317 to adjust the current in the LED's.

Of course with the 250 to 350 ma. LED current we are talking about here, it will be hard to find 1.5 ohms to 5 ohms trimpot so why not just parallel another resistor through a switch. That way you get high and low setting. LM317 advantage is it is rated at 1.5 amps so we can control a bank of lights (with multiple LED's in series). LM350 is the same type but rated at 3 amps.

 

One thing to watch out for is polrity. The better LED lamps will have a full wave bridge rectifier in them so they don't care about which way they're wired up. I got some that did not and i had to add it. Sure, it shouldn't be a big deal, but I feel that having them not be polarity sensitive reduces the number of things that can go wrong.

 

The LED is a diode and will pass voltage only one way except it is not practical to use it as a rectifier because of large voltage drop and (1.2 to 3.8 volts) and low current current capacity (5 ma up to 800 ma for the super LED).

You can try experimenting with it using a low voltage AC. 6 to 12 volts will do. Series a 1,000 ohms 1/4 watt resistor to limit current flow and it lights up either way.

If you use a DC supply, then it is polarity sensitive. It will light up only one way.

Most line AC powered LED's have a built in voltage step down and current limiter built in the housing as shown in the picture. All LED likes to be current controlled so in some cases, you see something like this but with DC input voltage from 6 to 36 volts. Alongside the picture is the LM317 I am experimenting on. I am changing my 220 VAC supply to DC 24 volts, my solar power output.

 

I just soldered on a full wave bridge rectifier and fed it DC, but the LED no longer cares which polarity it's hooked to.

 

You are technically correct because you are reverse biasing the LED to break down the PN junction. Not a very efficient way though as typical forward voltage drop is 1.7 volt whereas in reverse bias it takes about 5 volts.

It is like putting a car in reverse and driving backward to go forward. You wont burn out the LED but it will be inefficient as it takes 5 volts to light it up versus the 1.7 volt when forward biased.

 

We're talking about different things. A full wave bridge rectifier works because no matter which way the current coming in is polarized, the output will always be the same polarity. They're mostly used for converting AC to DC, but can also be used to convert DC from one polarity to another. In this case I'm using it so that no matter which polarity is coming in, the output will be the correct one for the LED.

http://en.wikipedia.org/wiki/Bridge_rectifier