buck 18V to 5V, then use a resistor down to 3.05V?

I have a Ryobi 18V incandescent flashlight/work light. I'm thinking of using one of those car cigar port USB adapters (one with a buck driver in it) to do two things.

1. to allow me to use the 18V lithium battery pack as a power bank
2. to give me a 5.05 Volt source to power an LED instead of the incandescent bulb.

I ordered a couple XM-L2 leds, and a Nanjg AK47 driver in 1050ma, and 1400ma. I'm told it can handle 6V, so it shouldn't have a problem with 5.05V from the buck driver.

Here's some information:
The XM-L2 forward voltage is 3.05V @ 1500ma
So I plugged Into a calculator..
http://led.linear1.org/1led.wiz
source V... 5.05V
Forward V.... 3.05V
Diode forward current (ma) ... 1500


The wizard recommends a 3.4 Watt capable or greater 1.5 ohm resistor. The color code for 1.5 ohms is brown green gold.

Link to this solution: http://led.linear1.org/1led.wiz?VS=5.01;VF=3.05;ID=1500

Here's my question:
If I didn't want to get the multi-mode Nanjg driver, but would rather just have one mode.... on and off, would putting a 1.5 ohm resistor inline, after the buck driver, drive the emitter at 1500ma?

I've tried this, the one thing I can tell you is that lil usb driver circuit isn't going to like driving that LED. It'll get seriously and possibly dangerously hot, and I mean really hot. I tried both a 1.2A and a 2.3A and both were just piping hot within a couple of mins being drawn at max output.

Use a dedicated (buck!) LED driver instead. That will reduce waste heat a LOT over [5V out + series resistor] @ 1.5A. And probably hold up longer, too.

USB 2.0 is limited to 1.0A maximum. That adapter may not even be happy at that level. You may be able to read the output ratings printed on it (some are difficult or impossible to read). You should not expect it to live long or prosper if you overload it.

If you stay within the rating of the adapter, what you want to do should work reasonably well. You would expect the current to increase a little as the LED heats up, but not a lot, and certainly not enough to worry about thermal runaway (I say this mainly to preempt those who might otherwise suggest the possibility).

However, to add to what RetroTechie points out, the 5V to 3V conversion using a resistor or linear regulator is only 60% efficient. That means you will be throwing away 40% of the battery's energy (as heat) before it even gets to the LED. Multiply that by the efficiency of the buck regulator, which might be around 90%, to get an overall 54% efficiency for your regulator (90% = 0.90; 60% = 0.60; 0.90 * 0.60 = 0.54 = 54%). If you were to use a single-stage buck-type LED driver, you might maintain around 90% overall.

Thank you gentlemen.

Are there buck drivers that deliver multiple outputs? If so can you suggest one?

Initially I mainly wanted to adapt the light so that I can use the batteries to power a USB port so that other items can be run from it, IE. use it as a power bank. Then, I noted that the USB, when connected, the indicator LED draws about 10ma, which is a pretty high parasitic drain. I cut the LED out, and the output voltage went to 5.5 Volts. Is that too high to use it to charge cell phones, etc?

Also DIWdiver, I have two different adapters one, this one is 1.0A and another states that it is 2.1A (which I think will make it Apple compatible.)

-----------------------------------------

I was going to wire the USB buck, hot all the time, and put the output 5.5 volts to the hot input of the nanjg driver, and the nanjg's hot output to the LED. Then connect the LED neg to the NEG output of the nanjg driver, and put the switch inline between the battery neg and the neg to the driver.

Another option is to use the resistor in place of the nanjg driver, but that wouldn't give me any modes.

Now that I see how energy wasteful both of the above options are, I guess the better way is to use two different buck drivers?

So I ordered a buck driver.
https://www.fasttech.com/p/1253701

I believe that if I set it at about 3.0 volts, an XM-L2 emitter NW will pull almost 1400ma, and will put out almost 500 lumens.

The buck driver came a while ago, and today I finally got to put things together. Everything worked fine when I used jumper wires. I set the buck at 2.82 volts, and then at 3.0 volts. I'll keep it at 3.0 volts. I heated the reflector to fill it with some solder, and messed up some of the shiny reflector... actually made it more orange peel which smoothed out the beam a bit. Ah very nice. So I soldered the emitter, and the power to the buck driver, BUT I carelessly switched the polarity when I soldered the power to the board.
I burnt out the buck driver, so now I have to wait for another one to arrive.