Drop-in Module Headlamp Prototype

[potholer]

MCU -> microcontroller.

Hand-soldering SMDs on stripboard is fairly simple - just apply a spot of solder on a strip where one end/pin is going to be, then hold the device in tweezers while melting the solder and positioning the device, then solder the other end/other pins.
It does help to have a reasonably fine soldering iron.

I tend to use 0803 size devices now, though I started with 1206 size which are probably slightly less fiddly.

One definite other tip for limited-space work is to consider PTFE insulated wire.
I tend to use it now for all fixed connections, and even for some movable ones.
PFTE 7/0.12 wire can carry 3 Amps, but is only ~0.7mm diameter. PVC insulated 7/0.2 wire might be rated for half the current while being close to twice the diameter.
PTFE may be more fiddly to strip, but the insulation doesn't melt back on soldering like PVC, and the wire is much less springy and easier to prod into a final position.
Cost could be an issue, but sometimes it might make the difference between something being feasible or not, or something with exposed wiring looking good rather than passable, or passable rather than rough.

[Tenzin Beck]

Ah, yes, MCU -> microcontroller was my guess, but I wanted to be sure.
I would like to get some PTFE wire at some point. The theory of soldering SMDs makes sense, I have a pretty good iron, and a decent amount of practice with through-hole components.
Any idea how much PCB prototyping costs?

[snoboy]

A great supplier of high end LEDs and associated supplies (refelectors etc) is:

http://photonfanatic.com/ComponentsFS.html

[Tenzin Beck]

A great supplier of high end LEDs and associated supplies (refelectors etc) is:

http://photonfanatic.com/ComponentsFS.html

All the tech there is a few years old. LED efficiency moves fast.

[rlboyce]

I'm pretty sure I won't be pursuing the drop-in idea any longer. Using drop-ins has the advantage of easy replaceability, but has the disadvantages of poor heat sinking options and larger size.

Instead, I'm going to do it from simpler components. I'll still be purchasing drivers though. Turns out I have a machining resource which will allow me to build a housing for dirt cheap. I have designed a rectangular housing, but I'm ditching that idea too and going with a Scurion-esque design. Thinking about making levels inside of the light so that leds and drivers can be inline with one another, but both will have a surface to heat sink off of.

If you're curious, below was the design for the rectangular housing. The design needs some fine tuning, but the basic concept is still there. The groove in the top face is for an o-ring.



For this go around, I'll probably be using some cheap DX flashlight drivers, as well as some XP-G R5s and a tight eight degree optic. Think I'll make it a blade mount. Should have the first iteration of the new design to show you guys in two weeks or so.

[LukeM]

Looks like you're well on your way with the design phase. You're so lucky that you have access to cheap machining! That's pretty much the only thing keeping me from taking my light designs further at this point. Have you decided what you're doing for a power source?

[rlboyce]

I think I'm probably going to go with li-ion 18650s. Seems to be what most diy-ers and cave light manufacturers use, and it seems that most people claim they do better with high power LEDs than NiMH AAs.

I'm not sure about the quantity of 18650s to use though. I've never worn a light that uses 18650s, so I don't know how heavy 4 batteries feels on your head compared to 2 batteries. If it's not too heavy and the drivers will allow it, I think I'm probably going to go with 4 of them. Wear them on my waist if need be, like a ceiling burner generator or something. Unfortunately, the drivers I want to use have a max input of 8.4v each, so that means I can only put two 18650s in series if I wire the drivers in parallel. However, I'm curious if having a parallel set of two 18650s in series will work (A = batt #1 + #2 in series; B = batt #3 + #4 in series; Put A + B in parallel). Do you know if this is inadvisable? I'd rather not fry the electronics if I can help it. : )

I've done some runtime estimates using different battery options and light modes. I'm very happy with the calculations. Using two independently operated XP-G R5s, approximately 80% efficiency drivers, 1.2/.55/.1 amp modes and 2x18650s I will be getting 1.65/3.9/24 hours with both leds running at the same time. If I only use one led at a time with 2 batteries I will be getting 3.3/7.8/48.1 hours. Not bad. If it turns out I can use 4 batteries it would double the run times. Considering the conservative approximate lumens per mode for one led are 360/180/34 I can't complain.

I will probably be rubber dipping the batteries together and using a connector between the housing and battery pack.

[LukeM]

I'm currently using a DIY 18650 battery pack with my light. It uses 3 cells (11.1v) and I find it to be light enough and the size is fine. I think 4 cells all parallel with each other would be fine, but I wouldn't want more than that on my head.

What you're talking about with putting batteries in parallel is referred to as 2s2p (2 series, 2 parallel) and by doing that you're effectively doubling your capacity while keeping the voltage the same. It shouldn't affect considerations when it comes to your drivers.

If you're making a homemade li-ion battery pack don't forget to include a protection pcb!

[rlboyce]

It shouldn't affect considerations when it comes to your drivers.

Excellent! I am happy to know that I won't be limited in the number of batteries I can use.

If you're making a homemade li-ion battery pack don't forget to include a protection pcb!

Are you saying I should use protected 18650s, or are you talking about some sort of separate electronic component? I was advised to use protected cells, but I don't remember hearing anything about a pcb. Something to do with voltage dropping too low and ruining the cell.

[LukeM]

Something like this. I'm inclined to pass up protected cells and rely on a protection circuit pcb since you're left with only 1 possible circuit failure point rather than several. Besides, that's also how pre-made li-ion battery packs are put together. Speaking of which, you can save a lot of time and headache for a bit more money spent by buying a pre-made battery pack. All-Battery is pretty good.

Edit: Also BatterySpace.com

[rlboyce]

Ah, I see. Thanks for the suggestion! I see what you mean by added failure points, and I'm definitely going to strongly consider using a pcb with unprotected cells. However, I still might make my own as it looks like I'd save a lot by doing so. I'm finding 2s2p 18650 packs somewhere around 40 and 50 bucks. Some much higher. However, I can buy 4 DX trustfire unprotected cells for 13 bucks. Throw in the pcb I should still be under 20. I think the savings are worth it, but maybe this is just my cheap side talking. Don't forget, the more I make on my own the more I can brag!

[LukeM]

Don't forget, the more I make on my own the more I can brag!

True! You also learn a lot more that way. Don't count out any hidden costs though. I don't know what you have but plasti-dip, solder, shrink tubing, etc are all minor costs that add up.

[E-man]

You have inspired me sir...
Been thinking about making a high power headlamp for years, but finally decided I needed to get it done. Spent a couple days on http://www.candlepowerforums.com/vb/forum.php and got up to date on the latest technology. Already ordered the parts I need for 3 headlamps and a couple extra battery packs. Going to make one, figure out what I want to change, sell the first one, then make another with the modifications. Here are the parts I ordered:

6 Cree XM-L cool white T6 bin
3 20mm 15 degree optics
3 20mm 45 degree optics
20 18650 3800mAh unprotected cells
5 7.4V protection circuits (planning on 2P2S packs)
3 regulation PCBs - 3-mode (3000mA, 1000mA, 150mA)

I think the first prototype will use a mechanical switch. 2nd and 3rd prototypes will likely use a reed switch or a magnetic pcb switch for complete waterproofing. Going to use some heat-sinks I found at the local metal recycling place, and embed everything in epoxy for complete waterproofing. Future prototypes might include a red LED (along with a separate switch and regulation circuit). I might also check out what heat-sinks I can find at the computer recycling place to use as housings. If I really get into this, I might try my hand at welding Al or find someone with a lathe/mill to help me machine some housings, but for now it's going to be pretty ghetto... but freaking bright... cue dancing banana

Doing the math, this thing should be about 4x as bright as the current Stenlight or 1.3x as bright as the current Scurion (using the same LEDs), and cost me about $50 in parts (including the battery pack). Just for kicks, I will use the same wire connection as the Sten (Molex 0.062" - http://www.stenlight.com/other-questions.htm) so I can lend my battery packs to my stenlight using friends. I will post pictures when I have all the parts and get around to making the damn thing. Wish me luick.

Peace, Eli

[LukeM]

Good for you Eli! Nice to see more people jumping on the DIY bandwagon.

Just a few things I'd note about your plans. First, do some experimenting, but 45 degree optics might be too narrow for a widest beam. This depends a lot on your preferences. Personally, I like to have a very wide flood. The main difficulty is balancing that with a narrow beam if you're going to run both at the same time. If the transition isn't very smooth it can be annoying. Many high end headlamps are using a bare led for the wide beam these days, and there are reflectors that only slightly limit the width if you don't want any wasted flood beyond your field of vision.

2nd, how bright are XM-L's at 150mA? Would that be too bright for a lowest setting?

Here's a mechanical pushbutton switch that I can recommend having used it in my headlamp. It includes an LED, which I opted not to use, and you'll see that you can buy a waterproof silicone cap to go with it. (I epoxied the cap in place to be safe) It's insanely easy to operate with gloved hands. Here's my post on what I made if you want to check it out.

How many mAh did you say those 18650 cells are?!

Sorry to the OP pushing this slightly OT! (but really still on topic :P)

[rlboyce]

No apology necessary LukeM! You are right, it's still related.

E-man: Wow! 6 xm-l leds is a lot of potential there. Be careful you don't start catching people's clothes on fire! "Hey Jim, is it just me or is this bat I'm looking at smoking??" Remember, people won't notice the ghetto appearance because you'll have blinded them, and if they haven't been blinded they will be too jealous of the output to be concerned with appearances. The fact that your light cost half of what they paid for theirs won't help.

Unfortunately, I haven't been able to progress much further with my own design. Machining a complex housing for free with built-in heatsinking seems to have become impossible, but luckily I've come across a CPF thread where a guy used a Hammond project box for a housing (which turns out to be an ideal, amazingly cheap and waterproof alternative). I think the outer dimensions are somewhere around 2" x 2.5"... I can deal with that. Another thing hanging me up is that I'm having trouble finding small enough waterproof SPST momentary pushbuttons. I've found larger ones that will work with my design, but I still want to find something smaller. I know they're out there, but I haven't had time to look yet. With a new baby, my free time usually goes to caving, not building an admittedly unnecessary light.

Good luck on your project E-man!