how wide is the housing? you might be able to fit the driver, depending on width and mounting needs, in between the 2 LEDs. Given the limited airflow it’s going to see, I think you’d be better off focusing on a driver with good thermal monitoring instead of more heatsinking. As much as I might dislike the guy at the moment, Taskled drivers do have that very well covered.
If you wire those 2 LEDs in parallel (frowned upon, but fine in practice) you’d be able to run the light off a 1S parallel battery pack. Stick a 3 cell holder inside the rest of that case and you’ll be able to choose between 1, 2 and 3 cells, depending on how long you expect to be under. Plus you can keep a spare cell in your backpack, just in case.
the first housing should be fine - it’ll give you 18mm between stars if you use 20mm stars (enough for a KD AMC7135 driver) or 20+mm if you use smaller stars (enough for a Taskled Lflex).
I figured it would be relatively chilly underground, but you still need airflow if you want to drive LEDs hard. My helmet light (2x XM-L @1.5A) will step down even at 10C if I’m only riding at walking pace. More surface area helps, but it’s really airflow that drives heat dissipation. Another plus of programmable drivers like the Lflex is that you can find the right balance between output, heat and runtime without taking the thing to pieces. They’re really worth the investment.
With 3x18650 you would certainly be able to fit an Lflex in the battery case. That would leave room for 3xXML in the head using 20mm optics(different beam patterns with tir?) The longest switch wire I’ve run with a “flex” driver is 16”.
trouble is, you’d then lose the thermal monitoring, which is one of the huge pluses to using the Taskled drivers. Given the restrictions on airflow and surface area, that’d be a trade off I wouldn’t recommend.
Thanks mattthemuppet, your’re a great source of knowledge. Do you think I’d have better heat-sinking with easy2led or maybe something else? I was thinking about how with 2 easy2leds I could have two independent lights, which would give me all sorts of combinations and I could also have fallback if one of the lights dies. I could have two wires running to the battery pack, each connected to one 18650 or a pair independently (I’d need to find a bit wider housing for 4 batteries).
you’re welcome! I had a lot of help when I built my first LED light a few years ago, so it’s nice to do the same now
I think 2 Easy2led housings would be too heavy - the one I built for a friend was ~90g for the head unit, including helmet mount and adapter. Another housing would 60-70g to that, which is getting to the uncomfortable side, especially if you have a 200g battery on the back. Almost a pound/ 1/2kg on your helmet is going to get noticeable after a little while.
If you build the light right, there’s no reason why it would fail - it’s all solid state. You might need a couple of test runs to iron out any kinks, but after that you should be fine. All of my lights have been 100% reliable for several years, although I guess I’m tempting fate by saying that!
I’d work backwards from how much runtime you think you’ll need (always assume you’ll need the light on full all the time). If you have 2 XM-L in parallel @2A (so 1A each) you’ll need around 2.2Ah for every hour of light you want. Decent 2.6Ah cells are $7/cell, top of the line 3.1Ah+ cells are$12/cell, plus ~$10 or so for a PCB (or buy protected cells for more :money_mouth_face:. Honestly I really wouldn’t go for a 3 cell battery - I spend every night ride wishing I had a 2 cell strapped to my helmet instead (I have another light with a similar weight head and 2 cell battery and the difference is noticeable). Once you figure out runtime and current draw, then you can work out if you have enough surface area on your housing.
ah, it was pretty petty. I had a great deal of regard for George as he really knows his electronics and provides fantastic CS, but I recently suggested that someone use a piece of red plastic and a white LED to make a tail light and he completely jumped down my throat and was incredibly patronising to boot, saying that I’m delusional if I think I’m getting more than 20lm out of my light, I can’t read spec sheets, I’m spreading misinformation, red XP-Es are the only way etc. I learnt an awful lot about relative spectral power and the differences in that between different tints, but it did leave a sour feeling and has certainly diminished my respect for him. I’ll still recommend his drivers as they are excellent and there aren’t many alternatives at that level, but I won’t go out of my way to do so.
The irony is that, although he is right, that red XP-Es are more efficient than even WW white LEDs+red plastic, the guy that originally asked the question can’t make the light he wants right now as he can’t run 2 red XP-Es of a 1S battery using the driver he has due to the funky Vf of the LEDs.
Why not make up a head unit with the led’s, optics, driver/driver’s and switch+dimmer pot, and then run a lead out to a plug and socket with a curly cable down to a belt mounted battery pack? You could then run more cells in parallel if need be (run time) and have a much less bulky head unit, or two lamp housings either side of your lid, with a control box (driver + switch gear) on the back.
My bike lights light head is not much bigger than a 22mm half inch drive socket at the most, has the on/off/mode control on the back and the pack is that small, I Velcro it to my swan neck. It even comes with a head strap.
Don’t forget the momentary switch ( normally open) for the Lflex. The UI for Flex drivers is pretty extensive and a bit of a pain but the tradeoff is fantastic info and control. Let me know if you have any trouble programming it.
