i ordered xpg3 on 16mm star, some gaskets, and reflectors, will try to see if i can get it focus properly. i like d25 but its bad beam profile is a turn off.
Adafruit makes a breakout board/adaptor available through Digikey. $2.95, but shipping and minimum order will eat you up. If you don’t need the wire attachment through-holes, you could probably design a sheet and have them made somewhere. You can sometimes find QFN sockets with SOIC footprint, but that would only be useful in a prototype board - they’re pretty tall.
FET+1, Zebralight like or Anduril UI . 11$ shipped. If interested PM.
I’m thinking something like this:
It’s a WIP, but shouldn’t take much to complete it. I just need to make sure I optimize the pin mapping (PIC’s ADC pins to ADC pins on the Attiny, etc). Unfortunately, the minimum board size at Oshpark is 0.25” x 0.25”. So some filing may need to be done, but I’m making it such that it won’t harm any traces.
This is not a yes-no question really.
The purpose of my experiment was to find the limits. The baseline. I found it and started adding features. I ended up with something comfortable that still fit 4KB.
It didn’t feature thermal regulation but it did feature LVP (as I said, it’s not immediately possible to remove it anyway).
Would I want it? Yes, having UI tuned for my preferences matters to me more than thermal protection.
Now…there’s something that puzzles me.
Does thermal protection of Andruil really consume nearly 2 KB?
I have my reservations. I think it’s likely that either enabling it pulls some unnecessary stuff or that disabling it removes something crucial. So this 4KB Andruil build may not actually be feasible….
I find your driver interesting. Fair price, nice upgrade. I’m not a fan of Zebralight UI but nevertheless I may ask you to make me one at some point. 
Another great idea. 
This is already a nice budget light family. Fitting it with full Andruil for the price of MCU and a small OSHPark board would be awesome.
Actually…Sofirn upgrades these lights already. This further inexpensive tweak would make the light way better without costing them a lot. I wonder if it would be possible to convince them to make such upgrade…
Well… the easy part should be done now (Oshpark link). Now just to port over the code 
Alternatively, one could make a similar board using the attiny1634 and use the existing D4V2 target since it utilizes the attiny1634. I doubt you could get away with on-board programming though unless you’d manage to squeeze the pads on the underside of the board.
Just to make sure it’s clear: I no longer have a D25 headlamp and I doubt I’ll ever personally install one of these boards in a D25. But once I get this working for other flashlights, I’m sure it could port over to the D10/D25.
spend a day watching you tube videos about 123d design, and seems like got enough skills to make a simplest design, so i made similar holder for zebralight 600 series. first one came a bit crooked, the base was not 100% parallel to the printer bed. but otherwise the holder works, holds the light pretty firm, so today i made another design, hopefully better, will try to print tonight, this one is for zebralight as well.
anyone wants stl, pm i’ll email it, i can’t upload it to imgur. now that i know how, (it was not that hard turns out) i’ll replace all my silicone holders on all headlamps.
Nicely done, sir
Looks like gchart et al. is making this a blf favorite headlamp. I’m thinking of getting another one to mod with combo spot and flood optics. Anybody with ideas how to improve waterproofing on this one without sealing the top cover?
Hey @YuvalS. Have you managed to get your modified caving A6 driver to control 2 leds leds separately to work on this thing? I think this may be a perfect headlamp to test that.
I have noticed a few seemingly identical versions of this headlamp, ranging from as little as $10.76 USD on AliExpress, to a whopping $92.99 CND at Lynx OGT.
Other than the battery and headband, is there any difference from all of these versions to justify such a significant price range?
Only difference I’m aware of is the Sofirn D25S where they changed emitters and UI (I think those were the changes, I haven’t owned one).
Damn that Lynx version is expensive. Looks like they have the other generic boruit models too at 10x the price. A few are even sold out!!. Even if they put zebralight level driver, leds from the best bins and improve waterproofing don’t think the price is justified for this host. At $10 though, it’s one of the best bang for buck headlamps out there.
I doubt the driver is any different from one model to the other. Even the green button is the same across all price variations!
Sorry to ask this question again…. but I plan on getting one. Has anyone officially determined if these XM-L2 LED’s are not a real cree? I played with a wowtac a2s then a sofirn sp40 and I was thinking of getting one of those for myself. I then ran into the D25 and thought it would be within 1-2% difference between them, since they all have XM-L2’s. I would actually expect the D25 is better because it has 2 XM-L2’s? I’m not sure I understand.
1. I don’t know if the LED is genuine, considering low price I suspect it’s not.
2. Regardless if that’s the case, performance is not within 1-2%. Let’s assume for a while that the LED is genuine XM-L2…
- For the same CCT and CRI, the difference between top bin and bottom bin is ~25%
- Within bin there’s about 7% difference
- Any CCT differences add variability
- Driver and optics are not the same, driver has huge effect on performance while optics has quite minor but still way more than 2%
- D25 has unregulated driver
while the other 2 are regulated. This means that D25 output drops as battery deplates which is not the case with the other 2. It may be better at full cell but worse at near-empty
EDIT:
Scratched info that the 2 lights that you’ve seen have regulated drivers as I’ve got notice from RD that this is incorrect w.r.t Sofirn SP40
I see people assume this a lot. Cree lists +–7% which is a potential 14% difference within a single bin. I guess they try to stay within 7, but allow 3.5 extra on each end just to cover themselves.
The bottom line is it’s probably about 7% like you said. Just thought I’d mention it.
I was talking about different 7% and missed this one.
You’re talking about “Cree maintains a tolerance of ±7% on flux and power measurements”.
I was talking about bin width, f.e. T5 is 260-280 lm (7.7% width) and U3 is 320-340 lm (6.3% width).
So…there’s even more variability than I listed above.
They has XP-Footprint and I think they are no genuine Cree LEDs.
They looks similar to XP-L LEDs.
i’m pretty sure those are LB leds, not cree
The Sofirn SP40 driver is also unregulated. Brightness decreases as the battery runs out.