My first experiments with LEDs: disappointing... :(

mailint

Newly Enlightened
Joined
Sep 22, 2007
Messages
165
Today I received my first LEDs to play with: a Cree XR-E, a Luxeon Rebel, a SSC Z-power P4, all pre-mounted on a star or PCB.

I immediately tested them directly connected to 2xAAA lithium Energizer batteries and I was disappointed by the results.

All were of similar brightness, the batteries went down from about 3.260V without load to 2.850V with the LED connected/turned on.

The currents were about 50-70mA the first time. Then, on a second measurement after a minute they were 40-60mA. Then, on a third measurement after a minute they were 30-50mA.

All the measurements last 1-3 seconds.

Why so low currents with direct connection to batteries? and why does them become even lower when tested after some time? please help me to understand...
 

Nereus

Enlightened
Joined
Mar 11, 2005
Messages
509
Location
Espoo, Finland, Northern Europe
The voltage of 2 batteries in series is too low. Try connecting 3 batteries in series. On the other hand, 2 batteries in series provides a very long runtime.

If you want to use 2 batteries only and to get greater brightness you need a step-up converter.

-N
 

matrixshaman

Flashlight Enthusiast
Joined
Jan 17, 2005
Messages
3,410
Location
Outside the Matrix
AAA batteries are pretty wimpy any way for those high power LED's AA's will have around 3 times the current available that AAA's do IIRC. Do you have any flashlights with these type LED's? With the right driver they are blinding bright and can throw a long ways not to mention light up a room quite well with the right setup. It just takes the right combination of power source and optics or reflectors to get dazzling brightness :)
 

Gunner12

Flashaholic
Joined
Dec 18, 2006
Messages
10,063
Location
Bay Area, CA
They usually run brightest with 3.7v or 3.5 for the Rebel 100. The AAA batteries will sag in voltage because of their low capacity and high internal resistance. AA batteries will work better. Don't forget a heat sink if you plan on running them for a long time.

Get a cheap driver from Dealextreme/Kaidomain and place that between the LED and enough batteries.

The LEDs are usually meant to be driven between 100 mA to 1000 mA.
 

mailint

Newly Enlightened
Joined
Sep 22, 2007
Messages
165
The voltage of 2 batteries in series is too low. Try connecting 3 batteries in series. On the other hand, 2 batteries in series provides a very long runtime.

If you want to use 2 batteries only and to get greater brightness you need a step-up converter.

-N

On Wikipedia (http://en.wikipedia.org/wiki/Boost_converter) I read that with the boost/step-up converter "Since power (V*I) must be conserved, the output current is lowered from the source current."
Ok, I obtain higher voltage but if I obtain a lower current than 70mA I'm disappointed anyway...
Fenix is selling a 2xAA (ok, not 2xAAA but not so much different I believe) flashlight called L2D RB100 with declared 175 lumen. I don't think that it can be so bright at only 50mA...... (anyway I just ordered it so I'll make some measurements and I'll try to power it with 2xAAA to see what happens...)
 

mailint

Newly Enlightened
Joined
Sep 22, 2007
Messages
165
AAA batteries are pretty wimpy any way for those high power LED's AA's will have around 3 times the current available that AAA's do IIRC.

I also have a 125mW green laser pointer powered by 2xAAA and when it runs the current is around 650mA. How can it be possible?
 

LED-FX

Enlightened
Joined
Jul 23, 2001
Messages
630
Location
Edinburgh UK
Because at 2.85V the light emitting Diode is hardly conduct, once you get over 3V typically for a white,and blue and green, LED conduction avalanches and current consumption rises steeply. The current is available from the batteries but not at such a low voltage.
 

Gunner12

Flashaholic
Joined
Dec 18, 2006
Messages
10,063
Location
Bay Area, CA
They meant power as in watts. Voltage is only half of the info you need, multiply by amperage and you get watts.

For example, assuming the impossible 100% efficient driver. The driver draws 2 amp form a 1.5v battery.
1.5v x 2 amp = 3 watt

Now if the LED runs at 3.7v(varies)
3 watt/3.7v = 0.8108 amp = 810.0 mA

If it is connected to a 100v device
3 watt/100v = 0.03 amp = 30 mA

I've read somewhere that the Fenix L2D is running the LED at up to 900 mA.

Assuming 3.6v LED and a 70% efficient driver.

0.9 amp x 3.6v = 3.24 watt to LED
3.24 watt/70% efficiency = 4.63 watt draw from the battery
4.63 watt/3v batteries = 1.54 amp draw

Example of a good boost driver, the RV7 1 AA board has been reported to supply 1100 mA to a LED from a 1.2v battery while drawing 4.5 amp from the battery(post).

Doing the math:

1.2v x 4.5 amp = 5.4watt draw from the battery
about 3.7v x 1.1 amp = 4.07 watt to the LED

About 75.4% efficiency.


Another thing is that a LED draws different amperage depending on the voltage. If the Cree was given unlimited current but a set voltage, for example 3v, it will not draw a high amount of current and burn it self out. Vice versa, if the LED was given a set current, it will only have a certain voltage. Look at this thread, his Rebel 100(data) draws 300 mA at 3.04v but only 20 mA at 2.74v. His P4(data) draws 100 mA at 3.00v and 20 mA at 2.77v.

I think what I'm trying to get to is that a boost driver draws as much current from the battery as needed to provide a certain current to the LED.

I'm hoping that made some kind of sense.

Also, rechargeable batteries and lithium batteries sag much less under high current draw then Alkaline batteries due to a lower internal resistance.
 
Last edited:

tebore

Flashlight Enthusiast
Joined
May 10, 2006
Messages
2,141
Location
Toronto, Ontario. CAN.
Those ain't fresh AAA Lithiums are they? They shouldn't start at 3.2 and sag to 2.8 so fast. I've had AA lithiums that usually test 3.4 and hold up till about half way then show some sag.
 

mailint

Newly Enlightened
Joined
Sep 22, 2007
Messages
165
Thanks Gunner for your very good explaination!

Do you also know what's the best boost driver for 2AAA - Rebel 100?

Tebore, you're correct. They're not new...
 

Gunner12

Flashaholic
Joined
Dec 18, 2006
Messages
10,063
Location
Bay Area, CA
The best relatively cheap driver for 3v will probably be the DX 1.5-4.2v Circuit Board 3 pack. Around 71% efficient with a 3v supply. The smallest would be the Micropuck from Luxeon(about 80% efficiency according to data sheet). There are three versions, 350 mA, 400 mA or 500 mA. The most powerful/efficient would be the Badboy step up drivers(85% according to specs). There are four versions, 300 mA, 400 mA, 500 mA, and 750 mA.

Those will all work with the Cree, Seoul and Rebel LEDs.
 

LEDcandle

Flashlight Enthusiast
Joined
Aug 15, 2005
Messages
1,943
Location
Mushroom S'pore
Fenix is selling a 2xAA (ok, not 2xAAA but not so much different I believe) flashlight called L2D RB100 with declared 175 lumen. I don't think that it can be so bright at only 50mA...... (anyway I just ordered it so I'll make some measurements and I'll try to power it with 2xAAA to see what happens...)

That's because Fenix lights have electronic converters inside and are not direct drive. to get 175 lumen, the high or turbo modes usually push 700ma -1A, similar for many other lights.

Most of the new LEDs need a voltage of about 3.7v or more to pack some punch. Two lithium 1.7v AAAs (and not even new!!) are apparently not enough.
 

mailint

Newly Enlightened
Joined
Sep 22, 2007
Messages
165
The best relatively cheap driver for 3v will probably be the DX 1.5-4.2v Circuit Board 3 pack. Around 71% efficient with a 3v supply. The smallest would be the Micropuck from Luxeon(about 80% efficiency according to data sheet). There are three versions, 350 mA, 400 mA or 500 mA. The most powerful/efficient would be the Badboy step up drivers(85% according to specs). There are four versions, 300 mA, 400 mA, 500 mA, and 750 mA.

Those will all work with the Cree, Seoul and Rebel LEDs.

Thanks for the great informations!
I'm mostly interested in the smallest >500mA - 3V. The Micropuck seems thick, sure that they're the smallest?
 

mailint

Newly Enlightened
Joined
Sep 22, 2007
Messages
165
That's because Fenix lights have electronic converters inside and are not direct drive. to get 175 lumen, the high or turbo modes usually push 700ma -1A, similar for many other lights.

Most of the new LEDs need a voltage of about 3.7v or more to pack some punch. Two lithium 1.7v AAAs (and not even new!!) are apparently not enough.

Thanks for the info!
I'm definitively convinced that I need a step up converter :)
 

IMSabbel

Enlightened
Joined
Dec 4, 2004
Messages
921
Thanks for the info!
I'm definitively convinced that I need a step up converter :)

Also, dont underestimate the difference between AA and AAA.
AAA cells are less than half as big as AA, and the internal resitance (which causes the voltage drop as higher currents) scales with internal surface area.
 

rscanady

Enlightened
Joined
Feb 26, 2005
Messages
785
Location
Beaverton, OR
You should also familiarize yourself with the current/voltage curves for LED's. This may also help you understand what is going on.
 

mailint

Newly Enlightened
Joined
Sep 22, 2007
Messages
165
You should also familiarize yourself with the current/voltage curves for LED's. This may also help you understand what is going on.

I knew the current/voltage curves for LED's. The problem was that I didn't imaginated so much voltage drop of the AAA batteries...
Thanks IMSabbel for the explanation!
 

hank

Flashlight Enthusiast
Joined
Apr 12, 2001
Messages
1,561
Location
Berkeley CA
I ordered a couple of Micropucks, they seem fine. One question -- do they somehow adjust to the different requirements of say a red or amber, compared to a blue or white, Luxeon? I gather so -- their preassembled puck-and-Luxeon can be any color. But how does it know what it's driving?
 

tonycollinet

Newly Enlightened
Joined
Oct 15, 2007
Messages
35
It doesn't know what it is driving, but it does know what current you want. Either you have bought a particular current rating, or you have set the current with a potentiometer.

It measures the current - if it is too low, it increases the voltage. If too high, the voltage is reduced.

In other words, it is a constant current source.

The measurement is done 10's or 100's of thousand times a second - so you don't see any significant variation in the voltage/current as it is controlled.
 
Top