Project "TORNADO" - S x S 18650

HarryN

Flashlight Enthusiast
Joined
Jan 22, 2004
Messages
3,978
Location
Pleasanton (Bay Area), CA, USA
Hi, A long time ago when I started the little "Breeze" side x side light, I also started a S x S 18650. The body is still in the cabinet, and lately, I am getting the itch again.

As usual, it will not be a quickly completed project.

LEDs will probably be 2 x Lux Vs - especially since I have a pair of XWOS parts in the drawer.

Optics - my special combo mix.

Timing - who knows

General concept - Headless Horseman

Take care

Harry
 
Last edited:
Why do you choose to use an old Lux V? Wouldn't a current gen emitter be better?

Well he said he's got em so might as well use them. I would still use X bins today especially XWOS if I could get my hands on them.
 
Hi Spencer - your question about the Lux V is a good one. I guess the answer is a bit complex, but here goes.

a) I am a bit over zealous on the LED binning codes
- I have tried a variety of LEDs of various brands.
- If the light is just for the kids to use and loose - who cares.
- If the light is for me and it is based on white LEDs, then it is going to be a WO tint (or realllllly close), and the package has to put out a lot of light

b) Modern Lux Vs are made with modern LED die
- I am perhaps more familiar than many with the LED wafer / die manufactuing processes used by various companies.
- I have sold, and continue to sell a fair amount of equipment to this market
- Lumileds isn't sticking junk into the Lux V, the die remain technically competitive

c) Comparing LED Die
- Single die devices are not enough light for me at the drive currents I want to use
- The K2-TFFC is getting closer. Of course it is much easier to focus than a 4 die package, but even the upper (available) bins of this are not at the light output of a Lux V X bin (bin 200 - 250 Lumens at 700ma)
- LED rated output can be quite highly rated, but behave much more mediocre in real world condtions, esp. due to temperature effects.
- Lumileds die (especially those in the last year) have some enhancements to dramatically decrease the temperature sensitivity.
 
Last edited:
Continuing the reply to Spencer:

d) Light Management
- It is challenging enough to get the beam a person likes with a single die package, and 4 die packages are kind of a pain.
- The SS has a "forward" pattern which means a reflector has limited use, but a lens works ok if you know what you are doing
- The Lux V has a lambertian pattern and some special optics to make its tint and output very even over a wide output angle. This gives great light extraction, but is also a pain to capture and focus
- I have a LOT of experience with the Lux V, esp. with the WXOS already in the "Breeze" light. When I say "a lot", I mean several years of pain, anguish, breakthrough, and thousands of dollars spent getting there. I am not in a position or mood to throw all of that away and start over with the SS package unless it really can beat out the XWOS in every binning category - hands down. So far, I would say the SS is a good package, but in real world conditions running in a 2S2P configuration like the Lux V, it is competitive, but not a blow away winner. I could be wrong of course, but those were my results.

The main downside of the Lux V package is that you cannot overdrive it or light output drops. People got in the habit of pushings Lux Is up to Lux III power levels which worked fine, and this very much not true for the Lux V package. I will admit wishing that the Lux V was available in a K2-like thermal and ceramic phos package.

Light management is sort of like tuning up a car for gas mileage. At one time we had a 1993 BMW 525i. It is a very nice, good size, 4 door sedan with decent performance. After about 50 K miles, we were still averaging 24 MPG highway at 70 MPH, a very competitive, but not mind blowing number. I started focusing on improving this with improved tires, tire pressure, syn rear grease, syn oil, etc. Nothing was done with software or tune up chips. Guess what - same car - 30 + MPG under similar conditions.

The point is that getting the light where you want it can be just as important as the theoretical LED outut.

Last but not least, I strongly perfer to buy products made in North America or EU than far east copies. I make this decision every day on every product I can. Just my part to reduce the trade deficit and revive America's economy.
 
Last edited:
Sorry for breaking up the reply into multiple posts - didn't want to have to retype a long reply.

Spencer - yes, I might make more than just lights for me. That is one reason for picking the 18650 format - I want to use a cell that is available in protected format from a variety of suppliers and is unlikely to go away - like the protected RCR2 cell did.

Obviously, you noticed that I reserved a few reply locations, which is typical of someone intending to make more than one.

Darkzero - actually, you can get X bin Lux V still from Future, but you have to buy a reel to get them and be patient. I actually helped Fred Pilon get the last batch.

I see that I have made a lot of spelling errors in these posts - maybe I need some coffee. :)
 
Last edited:
I was wondering why you aren't using the quad die Cree MC-E LEDs. You can wire it in 2s2p and it can hit 700 lumen or maybe more(not overdriven). Tint and flux binned but not Vf binned.

But a Luxeon V might work better for you and you are much more familiar with it too. The X bins seem to be fun too.

Either way, good luck and have fun with your project!
 
I was wondering why you aren't using the quad die Cree MC-E LEDs. You can wire it in 2s2p and it can hit 700 lumen or maybe more(not overdriven). Tint and flux binned but not Vf binned.

But a Luxeon V might work better for you and you are much more familiar with it too. The X bins seem to be fun too.

Either way, good luck and have fun with your project!

Thank you for that input. It figures that someone would point out that LED - I guess I have some more homework to do.

I really prefer to build a light using LEDs with a known, low Vf, as my preferred configuration is to use resistors A, B, and C to manage brightness. (similar to the "Breeze") The reason is not cost, and this light probably has room for a driver board, but instead I just don't like the electrical noise of a driver board - and frankly, lower reliability. The resistors I have for this project are similar to space qualified, so any driver is going to reduce ruggedness.

Let's see how far I get on the first one and what LEDs are available by then - probably LL will have something new by then as well. :ohgeez:

Worst case, I can just build them up to a certain level, and home modders can make their "adjustments" from there. There is a lot that can be done once people get their hands on a basic light body form factor they like.
 
Well, I see that I need to "eat some crow" on this. You guys are dead on about Lux Vs - even X bins, availability, etc. A year is a long time in the LED world.

I guess I need to figure out how to use a handful of Rebel-TFFCs in there. At least they are easy to obtain in reasonable bins. Looks interesting, but makes the project that much more work.
 
You can try the Cree MC-Es. You can wire the dies in them in 2S2P like the Luxeon V.

But I'm not sure how different the emitting pattern is.

Check the Marketplace and see if you can find some. IIRC $20-$30 for each, more $20 I think.
 
Hi Gunner,

Thanks for the input on the MC-E.

Loosing the Lux V general availability is a setback for my "Breeze" project that will be difficult to recove from. I have not (yet) found in the market an LED package with similar optical, electrical, and mechanical characteristics. This is especially true if you need a highly binned Vf LED.

For this new project, in theory I can start over. In practice, I wasn't counting on that. It appears that the SS and Cree MC-E are quite bright, but optically a pain - even more than the Lux V.

Lumileds goes through a lot of sorting to get matching sets of LED die in the LuxV - for output and Vf binning. I have not read enough yet about the Cree part to see how well these are matched up in the same package, but it might be close enough to run 2S2P, or I might have to run 4S, and then 2 packages in series.

Since the high power lights on here are mostly Cree or SS, maybe I should do something entirely different. After all, what fun is following the crowd ? :laughing:
 
Top