Go wider on the CCT spread.
My IF25A has LEDs from IF25: SST20s. 2x 2700k and 2x 6500k.
Result : CC= 4250k. duv = -0.0016. CRI= 82 (all on Opple 3)
Appearance is actually pinker than duv suggests. My opple seems to read higher duv with low to med CRI LEDs. Both the 2700k and 6500k LEDs had positive duv on there own when measured in the original IF25 host.
I actually like these hosts. The IF25 is a very good tool for testing different tint mixes. I have one with a mix of high CRI LH351d’s and the other is still SST20s, but may soon get 519As.
I haven’t tried the SST12s yet because of reports of positive duv.
I would try a wider spread but I only bought 3000k, 4000k, and 5000k. I don’t think I’ll be buying any SST12 again.
Though I wasn’t happy with the results, I suppose it was a good experiment. Maybe soon I’ll try an XPP tint mix. I have a couple yellowish 3000k, a 5000k, and I think a 6000k.
I stripped the anodizing off of a Convoy S8 host and got a kind of neat slightly iridescent deep silver/gray finish on it by putting it through the dishwasher.
I stripped the anodizing with powdered lye, and just kept adding more until it got hot and stripped the finish. (if you feel like trying this as well please be very careful and use eye protection because that is very nasty stuff). After that I put it through the dishwasher for a few cycles (the high heat setting seemed to give it a darker color) and got this result that I’m quite happy with. I didn’t want to strip the anodizing off the threads so I left the tailswitch boot on as well as the lens, but I cut a piece of plastic to cover the lens which only partially worked at keeping the lye solution out so there’s a ring on the reflector that lost the shiny coating. I might end up replacing the reflector but it doesn’t have any obvious negative effects that I’ve noticed.
It has a 519a 5000k with a Convoy 5a buck driver. And the lighted switch is also from Convoy. I really like the S8 as a host since the deeper reflector throws enough to be significantly more useful outside than an S2+ or my D4K with the 519a.
I take all of the internals out so it’s just the bare aluminum host, and aluminum going through the dishwasher isn’t going to be harmful. The way an aluminum baking sheet changed color in the dishwasher gave me the idea in the first place, so it seems quite safe as best I can tell.
When aluminum is run through a dishwasher with the cleaning detergent the alkalinity of the detergent can corrode and cause surface color change on the aluminum.
The lye solution used to remove anodizing is just a (much) stronger alkaline than the dishwasher detergent.
I used a heat gun and strap wrenches to disassemble the head of the light. Fitting the SFT-40/centering gasket into the reflector required the hole in the reflector to be enlarged.
This isn’t a very good comparison, but it does at least show how much less bad the tint/tint-shift is. In the ‘New’ photo it may look like there’s a “donut hole” in the hotspot, but it’s not present as long as you’re shining it at something more than, like, one inch away from the light.
aux emitters (I don’t know if I will actually be able to assemble those, but I have aquired a hotplate in order to try it )
new emitter, SFT-40 3000k (if they ever come back in stock)
I have a few questions, though:
Does anyone know if a lighted or RGB switch will fit into this host? I suspect maybe it wouldn’t, since the “1st gen D4” switch is listed separately, but I’m not sure.
Any advice on alternatives to the SFT-40 3000k? Obviously this is a subjective thing, but I just wondered if there are some other options I should consider. I do like warmer tints — not sure what other warm/high-CRI throw-ey emitters are out there.
That sounds like a fun project, I really like the D1S host but they’re so hard to get.
It works, though you will need to modify the switch hole with a dremel to fit the new switch. I have one that was upgraded by ArtieT59 with an SFT40 and an amber switch.
There’s a new customized emitter being sold by Fireflylite, it is 3500k, 95cri, and very rosy. 5050 footprint just like the SFT40, but a round die. It should be fine up to 5-6A, so make sure to get the right driver.
Another option is the Cree XP-P, it is available in 3000k 90cri and supposedly has decent tint, though the beam will be much narrower than SFT40, it’s not very bright but the die is very intense. You’ll also want the 5A driver for it, but if Hank has a 4A driver it’d be better.
Also, remember that Hank will sell nearly all parts to you if you ask, and then you don’t need to worry about being out of stock.
I see. I’ll probably wait and see if I can succeed in building the flex aux PCB that was mentioned — if I do, I will probably skip the lighted switch, but if not, perhaps I will go that route.
That may be interesting; thanks for the link. Would I be correct in thinking that if I were to use that LED, I should flash “nofet” firmware to a 5A driver? Or just be careful which battery I use, I suppose.
A related sanity check: if I do get an SFT-40, it would handle the 9A driver w/FET firmware, correct?
I’ve ordered 10, but they haven’t shipped yet. I don’t think they will run out instantly, since presumably the MOQ is a few reels they’ll have at least a few thousand emitters and reddit is still kind of wary of FFlite.
I managed to cram a couple of 7135 regulators into the stock D10 driver and get a regulated D10 headlamp. It runs very cool and has stable output throughout the entire runtime. Next I will mod the D25LR when it arrives with similar technique. Gotta order a bag of 7135s now.
Interesting, I ordered 5 of them, just for curiosity. Sanan chip likely, like ‘egg yolk Yinding 5050’, but equipped with special phosphor mix according to customer spec (if it is really a custom emitter).
Emitter swap on the Sofirn SC01 keychain light. This has been weighing down my keys for over a year with the battery being dead for most of that time. I think the greenish SST20, micro USB, and EDC-ing a TS10 had given me little motivation to charge it (not sure why I left it on my keys though).
I had been researching emitter swaps on the similar olight keychain lights and was just about to forcefully push the guts out, when I decided to first try unscrewing the pill. Well sure enough, it was screwed-in, not press-fit!
This is a pretty neat light and I’m glad I got one before it was discontinued. Sofirn, how about re-releasing it in a few different colors of aluminum, 519A, USB-C, and giving the option for a 10400 tube extension? I think it would be one of the better tube-style, twisty, keychain lights on the market.
Here’s the pill. Hey, a 7135 chip. There’s a little button-like contact soldered to the bottom of the pill that gives it the two-stage UI. I used an alligator clip, held in a vise, to keep the button in place. I used a heat gun for the reflow. Luckily everything stayed in place while the solder melted and it amazingly turned out to be an easy mod. I should have done this a long time ago!
Went from a greenish SST20 4000k to a dedomed 519A 5000k. Huge improvement.
I was definitely worried about it. The heat gun melted all the solder on all the components, but everything stayed in place. My only other experience with something like this is doing TS10 emitter swaps - same deal, solder on everthing melts, aux leds included, but it basically stays in place.
)
