Picking the right resistor?

[bfromcolo]

Hi, this is my first post so be kind.

I am fooling around with some CREE XR-E emitters and a 9.6v (8xAA 2600mah NiMH) toy car power source. If I wanted to wire two of the LEDs in series I need a resistor also.

Based on the LED specs voltage at given currents is:
3.3v - .350A
3.5v - .700A
3.7v - 1.00A

At 9.6v and .700A I would need the resistor to drop 2.6V, so R = 3.7 ohm

At 11.2v (assuming 1.4v per cell fully charged) and 1A, the resistor needs to drop 3.8v, so R = 3.8 ohm, to prevent frying the LEDs.

So it appears a 4 ohm resistor should work right? Give me good brightness over a few hours of run time? I plotted the 3 A/V points and trying to guesstimate where 4 ohms would put me on the graph, and at 9.6v it should be around 625mA and 3.45v, or about 2.25 W. I can't find a chart telling me what light output an XR-E would provide at that current level, but at 80 lm per Watt can I expect

2 LED x 2.25W x 80 ln/W = 360 lumens

and reasonable brightness run time of

80% (2600 maH / 625 mA) = 3.3 hours

I was looking at the specs for the LM317, but if the reported drop of 3V in the chip is true, I would never get enough current/voltage to the LEDs in series to work right?

It may be hard to believe I have a BSEE after reading all that, but these kinds of LEDs were not around back in the day. Any help you can provide confirming or correcting my logic would be appreciated.

B from CO

 

[Marduke]

I would recommend using the proper driver board. Several cheap ones can be found at http://www.dealextreme.com/products.dx/category.917

 

[VidPro]

sure 4 ohms, works great do it. get a nice fat 1 watt resister, they sell really nice 4 ohm resisters that will fit in about anything.

when your batteries are low, it will use way less juice, so it will just keep running , which is great.
if the battereis are just charged, you can bother to WOW sombody with it.

when it goes from 1amp to about 300ma it looks about 1/2 bright, after that it looks , , , like time to charge

dont forget that the "Vf" thing in the leds goes down over time and when the temps rise.

check the actual LOAD voltage of the ni-mhys, they might be 1.4+ hot off the charger , but via wire an connections and all it will be less. check how the led voltage is when its actually in the application (heat). check the total current in the application.

so get a few extra resisters, like say some 1/2W 10ohms you can parellel, then you can parellel 2x10ohm for 5 ohms, add one more for only 3.3ohms, and one more for 2.5ohms. the resister calculators are a good start, but adjusting down after its in place is often nessisary because of resistance that is already in all the connections and wire and all.

If your MAX voltage your getting through the curcuit, when the battery is FULLY charged, is below the 1Amp, and you have a good heat sinc it will work good even after things change over time. do NOT run over the 1000ma trying to overdrive, because then when things change you will drive the led too hot.

i recommend it, and after wiring and switches and all, you might find that its a bit less (or more) than the 4ohms to get spec at full charge., but that is a good place to start till its wired how it will be finally.

then put in a 2Way switch so you can have 2 levels, say with a higher resister, it will run forever, both because its hard to break, damage, and there are no parts that have MTBFs, so there is little that can go wrong. PLUS you are far less likly to reverse charge the SERIES batteries when one of the battereis goes flat on you.

the 317 itself has only about 1.5v drop, but a total curcuit using the 317 has reverse polarity protection, and therfore is the added voltage drop of the cheap diodes too. you CAN potentially set up a 317 with no reverse polarity and have a max loss of 1.5v and that loss comes out as major heat on the 317 itself.