To be clear, none of the options are bad. All five are great choices, even the old but venerable XP-G2. I’m pretty happy with my triples in XP-G2, XP-G3, 219b, and XP-L HI. I don’t have any triples with 219c or LH351D, but my 219c quad is nice, and everyone with a LH351D says it’s awesome.
If we can get more than one flavor made, I plan on getting every flavor.
In that case, it’s probably especially important to maintain a practice of humility and remembering what it’s like to be at the bottom. They say (and studies show) power corrupts… and, as Spiderman says, with great power comes great responsibility. Gotta be extra careful to avoid bad habits.
Any decent cell should work, though higher-amp cells like the 30Q and VTC6/VTC5 should make turbo a little brighter. On regulated modes, the highest output is only 3 amps, so most cells should be fine most of the time.
This light isn’t optimized for maximum turbo performance. The heat sink isn’t massive, the springs aren’t fancy low-resistance alloys, the wires aren’t extra-thick, etc… It’s not particularly hot-rodded.
The main thing is to make sure the cell is small enough, so probably an unprotected battery.
There isn’t a lot left to do, and progress has sped up again… so hopefully release will be soon.
The 10507 has the biggest numbers on paper, but…
- It makes an ugly beam. I hear it looks okay with LH351D, but on every emitter I’ve tried in one, I end up hating the beam. It produces a rainbow effect and lots of artifacts.
- The 10511 can be polished to increase throw, if desired, and from what I hear, it can throw better than the 10507 after being polished. Afterward it has more visible artifacts, but the beam still ends up looking nicer than a 10507.
- Swapping the optic is easy and the parts are only about $1.50 each, for anyone who wants a different optic.
Yes, having data helps with convincing the manufacturer that a market exists for both. ![:slight_smile: :slight_smile:](https://budgetlightforum.com/images/emoji/google/slight_smile.png?v=12)
I kind of prefer the Ranked Pairs algorithm over Schulze/beatpath, but RP hasn’t been working lately on the CIVS site and the results are almost never any different, so I went with the more stable more popular option. I’m using RP locally though. It’s simpler, and it was easy to find existing code for so I wouldn’t have to reinvent another wheel. I just had to mod the input parser a little to accept the format I get when pasting data from the CIVS site.
For XP-L HI, more CCTs are available… so that would probably be 4A (4650K). But for LH351D I need to look at the data more closely to see whether people want 4000K or 5000K.