High Efficiency LED for Joule Thief, etc.

[TinselKoala]

Here's something I've been playing around with. It's a high-efficiency LED,  Philips Lumileds MXA7-PW57-H001, coupled with a AP2502 current-sink regulator chip. The AP2502 has 4 current-sink outputs each regulated to sink 20 mA. So over the supply voltage range of 2.9 to 6 volts, it will keep 20 mA current flowing in the LED, per output. I'm only using one output and one LED here. The chip also has an Enable pin which allows PWM modulation of brightness by varying the duty cycle of the voltage applied to the pin. The chip eliminates the need for a power-wasting series resistor to regulate current. The LED is just about the highest efficiency LED available on the market.

The parts were kindly sent to me by a friend.

I'll be doing some more tests with these devices as time goes by. It will be very interesting to see how long a single CR2032 button cell will keep this LED glowing intensely brightly at 20 mA. The LED is actually nowhere near full brightness at 20 mA; at 100 mA it is blindingly bright, obscenely bright. 

http://www.youtube.com/watch?v=BtcF2aMC_ew
http://www.youtube.com/watch?v=X9wxuRZV-Ro

The LEDs and the current-sink chip are available from DigiKey. The LEDs cost about 70 cents US for a single one, or about 10 dollars for 25 of them. The AP2502 is about the same price. These are both surface mount devices. SparkFun sells a breakout board for SOT-23-6 devices for about a dollar each, that will turn the tiny chip into a standard 6-pin DIP footprint.

[MarkE]

Each of the pins 3, 4, 5, and 6 sink 20mA when pin 1 is high.  The AP2502 is intended to be used as an area back light driver.

[TinselKoala]

Each of the pins 3, 4, 5, and 6 sink 20mA when pin 1 is high.  The AP2502 is intended to be used as an area back light driver.

So the chip must be operating its output mosfets in the linear region, yes? With Enable tied 100 percent High, the LEDs aren't pulsed, I don't think, so current regulation must be accomplished by the op-amps regulating the voltage to the Gates of the mosfets as the LED supply voltage is varied. Is that right?

[MarkE]

So the chip must be operating its output mosfets in the linear region, yes? With Enable tied 100 percent High, the LEDs aren't pulsed, I don't think, so current regulation must be accomplished by the op-amps regulating the voltage to the Gates of the mosfets as the LED supply voltage is varied. Is that right?

Yes, that is how the device works.  The idea is that if you have a lithium battery that is nominally 3.6V in a hand-held product, then you get good efficiency by just linear regulating to a white LED that can run on 2.9V.  Most of the time the battery will be between 3.4V and 3.7V giving you 78% - 85% efficiency in a small space at really low cost. 

JT circuits are current sources of sorts so they don't need the additional regulator to protect the LED.  The regulator flattens the flyback voltage down slope, converging rms and average LED current, increasing efficiency.  Depending on the LED parameters and the particular JT design, the regulator can realize increased net efficiency. 

Those cheapy blue oval flash lights from Harbor Freight just use a simple current limiting resistor.  So, they are quite inefficient and the brightness starts dropping sharply in less than an hour.  A current regulator of any kind makes that better.  A high efficiency switching regulator is more ideal.  A pair of those clip-on 1.5V regulators that have been overhyped would work great with the board that you have.  The efficiency after those regulators would be about 180 lumens per Watt.

[ramset]

Tinsel
A sizzling Topic ATM ...
Think PINK
I am very Happy to see "you" working on this .

https://www.youtube.com/watch?v=ILzWmw53Wwo

http://www.terraspheresystems.com/index.php/news-media/19-news/41-urban-farming-20-no-soil-no-sun
From Above link  I square ft of Vertical farming WILL YIELD 64 head of lettuce per arum, where as 1 square foot of Land farming
?? MAY ??   yield 3 heads per anum.[Ideal conditions]

Establishing Crop yields per square foot against energy required to run the LED's would be wonderful to know ,do they have to run at full power ?
I do understand that PWM would NOT be acceptable for growing crops ,you can trick the eye But not mother nature [thx Brad].
Or would PWM be appropriate for part of the plants 24/7 365 artificial day ??

This is the future.....

Chet K