Hey guys, this one’s had some interest so here’s a thread on it; replacement TA based (BLF) driver for the Sofirn SP33!
This is a linear driver, it runs a 3v led off 1x26650 like the original SP33 V1 was, it runs all the awesome BLF firmware options like Anduril and NarsilM (and any other FSM based toykeeper UI). It’s a 3Ch with 8+1x7135+FET using the ROT66 config file.
Driver boards available on oshpark here:
And switch board here:
Programming pads on the bottom (they have solder on them cause I just soldered wires to them, no cool adapter or anything but atleast they’re there and you can flash it without taking it apart!)
The most reasonable (easy and cheap to obtain, and with lots of tint choices including several HCRI) choice that I know of would be the LH351D however if you fancy higher output than that you can always go luxeon V.
When I started the driver design I planned on using a LH351D but then I got my hands on the luxeon V2 and the tint won out. An LH351D should do ~1600, I’m seeing 1455 OTF so not a big loss for (what I consider) superior tint. It’s also a tad more throwie with the smaller die and dome (34,800Cd) than an LH351D would make it.
Edit: my reply was in response to PBWilson, lost service in the subway and typed my reply, didn’t refresh before posting.
One of my original design goals was for the main board to work with the factory switch board to save money and it does however it will take two very interesting to run air wires (switch signal and AUX LED signal which you’ll have to pull from tiny tiny MCU pads) and it’ll take scraping mask back to expose ground points (which also serve as physical support). So in short, even though I met my design goals just spend the $3.25 and have such a nicer build. You also wont have to destroy the factory driver this way.
I don’t mean to sound like a dick here at all man but for me to do a project it has to be something I’m interested in and a 6v LED in this light isn’t currently something that appeals to me (heck I couldn’t rip the XHP50.2 out fast enough!)
Since it is able to be locked out at the tail you always have the option to do a zener mod and run 2x26350 to a 6v led that way but what if anything is there to gain over stock? Not runtime going down to 2x26350’s… Note you’ll also loose lvp as there’s no voltage divider on board (voltage is monitored internally by the MCU and thus batt monitoring won’t function at all).
I am more than happy to share the actual eagle files if you or someone you know uses eagle, there’s enough open board space to add whatever you’d like on the switch board.
If there was a Nichia 144 MCPCB available >20mm I would definitely be much more open to the idea to alter it with a LDO then but using a 20mm mcpcb in this light looks like a nightmare (you already have to be attentive to solder joint size to keep the reflector from shorting with a 26mm MCPCB).
Sorry. You could always order my boards, leave them bare and piggyback a Lexel LDO driver onto it!
For me it is rule number one in my hobby and BLF: if I’m not personally interested in doing something I feel completely free not to do it. I’d left BLF unhappy years ago if I would not have kept this rule in mind.
I can import Eagle hopefully in my program,
but with respect I would also want the original switch board measurements and pads kept so that the original can be used
Nice CK! Do you have a source for the switch? Not sure - don't think I have any of those. I have more confidence if I can get a true spec'd switch from Arrow or Mouse for example, as opposed to the originals, but an original would do as well if you transferred it.
The problem with trying to use the original switch board is that the MCU is on the switch board so the switch signal and aux LED signal didn’t need to be broken out at the junction from SW board to main board so there arnt pads for them. The stock driver only has 6 total pads between the two: it has 4 ground pads (to function as ground signal and mechanical stability), one signal pin and VCC.
That’s why i sort of gave up on the goal of compatibility even though my lower blard will physically interface perfectly with the stock upper board it’s actually a useless design feature. To use the stock switch board you’d have to solder air wires to the tiny MCU pads to pickup the switch and aux signals and while possible it’s an ugly solution.
I don’t yet, was going to start looking but THB I was gonna ask you if you knew of one lol. It seems to be a common 5x5 body but the plunger is longer than other 5x5 switches that I have here.
Hhhmmm, I'd have to check at home. Boy, sure looks familiar but probably just from so many lights I've modded. I've bought a few types of the other type surface mount switches, the smaller right angle ones, like in the ZY-T11 clones. Those you can even get in different force activation ratings. I found the higher force ones are too hard to press, but I believe the rubber boot and setup of the rubber boot factors in as well. You can go nuts with all the possibilities and differences, tweaking to get the best feel. I tend though to prefer softer activation because it seems better made for ramping (press&hold) UI's -- actually, that's what's sooo great about the ZY-T11 clones: easy activation switches. Seems like the bigger/wide the rubber boot, the easier it is.
I soooo much prefer the small round PCB mounted side switches, but wayy better when the switch bezel can simply be unscrewed. I hate it when I break open a new e-switch light only to find one of these perpendicular PCB drivers in there... Of course for you, you can simply design a new board or board set like you did here, but for me - comes down to if I can jury rig in a cut down 15 mm driver or something, which I've done in some, but mostly I gotta punt.
Sofirn, unfortunately, is going with the perpendicular driver mount in many/most of their lights. Got a couple at least I'm stuck with, think'n the SP32 and SP32A as 2 examples. Don't have pics, probably because I got sooo frustrated...
