I have a host that runs 3 x 26650 and is a triple LED setup. Called a Lustfire, although looks very similar to some of these ‘EDC’ branded lights that have been on the forum recently.
Stock it ran XM-L’s (bought it quite a while back) with an ok’ish driver.
But I want to “improve” it.
I’ll only ever use this on High and want max performance, so was planning a Direct Drive setup. Using 3 x XP-L HI wired in series.
However, I got to thinking, would direct drive XHP35 HI wired in parallel work and give more lumens and probably more throw too?
The XHP35’s seem to cost about double the XP-L’s, but I can live with that. My main concern is, direct drive is there risk of burning the XPH35’s up? I want it to be usable and durable and I’ve heard the XHP35’s don’t like too much current.
Somewhere it is posted a chart of current vs voltage for the XHP35 and at 2A, I believe voltage is in the range of 14V. My guess is that 3X lithium ion cells will be safe and maybe a little underwhelming.
Well it’s only a 3S host, so 4S won’t be an option. My biggest worry was too much current and nuking the emitters. But if the volts are too low to worry, then I think XP-L HI is the way to go.
Not really interested in buying an expensive driver. As by the time you’ve done this and the LEDs, plus the time and other tweaks, I might as well just put the money to a new better light anyhow.
I’ve got to admit; I don’t understand this lust for DD lights. Wouldn’t you rather have a regulated light with consistent performance that won’t be a fleeting burst of output and will not subject the emitter to insane temps?
Well, for most LEDs, a FET driver is the simplest and cheapest way to get maximum brightness. Regulation is great, but it almost always means less output.
But you don’t get a fleeting burst, you actually get a nice discharge profile, which not only offers maximum output pretty much. But also the longest runtime, as it’ll never fall out of regulation. (Runtime for high output that is).
Plus, lights like this I would only use for bursts at a time anyhow and would be wanting the highest output. And it’s a large host, I highly doubt heat will really be an issue.
I think we all want consistent high performance, but like I said there is usually a compromise between output and regulation. The ideal is a high power, efficient boost driver, but this doesn’t really exist for flashlights. The next option is a buck driver, but then you need more cells in series to maintain that high output as the cell voltage drops, and very high power buck drivers don’t really exist for flashlights. The FET+ linear drivers PD68 mentioned are probably the best compromise.
The FET driver output vs time curve doesn’t have to be very steep; it depends on the cell discharge curve. Cells like the LG D1 have a nice plateau. If only the LiFePO4 cells with their very flat discharge curves had higher voltages and capacities; it would turn FET drivers into nice regulated drivers.
I’m certainly not knocking it if you don’t want direct drive. But curious, do you have any direct drive lights, or have you used any?
The consistency is generally fine, buy the time they noticeably dim to the eye, chances are they would have fallen out of regulation on a linear driver anyhow.
I have a number of lights that are direct drive. Some have a current limiter, some don’t. But none really appear as inconsistent in output. Not to the degree my eyes can tell anyway.
I’ll be honest and say I never turn my lights on at full charge and leave them running until they are flat. But I’m not convinced regulated lights are always as constant output as many think. And will usually nose dive when they fall out of regulation. DD generally has a gradual decline.
Personally I like, own and use both kinds of lights. But there honestly isn’t such a void in usability as can sometimes be claimed.
Thanks for the info. But in practice I don’t think it works like that. I have a good number of single cell direct drive lights with no driver at all. And many with FET drivers, which in Turbo amount to the same thing really. None of them get to 8amps draw from a Samsung 30Q across various different types of emitter. And none have burned up, some I’ve been running for years like this. And in hosts much smaller. Some don’t even have DTP copper stars, yet still perform after much abuse and hours of on time.
I admit I’ve never run a multi cell, multi led setup before. But 3 cells and 3 LEDs should essentially mean 1 cell per LED, so my expectation is each LED will perform similar to how it would in a single cell host.
Although I’m happy to be corrected on this. I guess the only real difference is, as there are 3 cells, each cell is pushed less hard. So maybe I’ll see more amps than on a single cell/single emitter setup?
It does, if you supply the LED (in this case XP-L HI) with 4.2v, it will draw 7.5A.
If you supply the XP-L HI with 4v it will draw 7A.
That's what the Vf - If graph is for.
I also run an XP-G2 direct drive, that's because it has much higher Vf and only uses about 5A at 4v.
I'm guessing these other lights that you run direct drive are not XP-L HIs, they are probably other LEDs right?
Because older LEDs like the XP-G2, XM-l, etc. will easily run direct drive without consuming too much current.
And yes you are correct, 3 LEDs in series with 3 batteries in series would the the same as one LED and one battery.
The problem is that with the new Cree LEDs like the XP-L HI, XP-G3, etc.... they all have much lower Vf, which means if you give them 4 or 4.5v they will draw much more current that one of the older LEDs.
Older LEDs would need much higher voltage to run at 7A, 8A, 9A, etc. which is more than what a single 4.2v lipo cell can deliver.
So you either need to limit the voltage to lower levels (because lipo does 4.2v at full charge) or you need to use a current limiting driver to keep it at lower currents than it would otherwise be at in direct drive.
I rarely use turbo but it is nice to know it is there. The vast majority of the time the 2.5A regulated mode is more then enough.
That too^
There is also voltage lost in the springs, wires, circuit traces ect. It all adds up to me also seeing about 5-6A max from an XP-L with all the tricks done to it with an HE2 (the highest preforming cell I own, even more then 30Q’s).
