I want to see a photo of Steve in the lounge using the loupe to find that thin wire…

The lights nothing special to be honest. It has parts in it given to me by Dale. Thanks Dale, but he wont want to know that its only powered by a humble 1 or 2 NiMH batteries.
The pictures, I was trying to take in the lounge room and the louping was carried out in the shed.
I’m pretty sure I would be shot if I brought the soldering iron into the lounge room. I’m not as lucky as some others here.

Put a 5D XPL into my Olight S1 Ti. Found Turkeylords video on how to get the head open and it went well. Got the old board out, filed down a new one for the XPL, reflowed and tested the board, all is good. Got the board in well, soldered to the correct wires and it works… Sort of . The XPL in the XM-L2 TIR is a square kind of beam, not too bad, I can live with it… The XPL must have a different enough Vf to affect the UI, now the single click to off goes to moonlight and sometimes it cuts off after a few seconds, most of the time it does not. It stays in ML…

Now I need to find a high CRI XM-L2 and do this again…

ReManG wrote:

Put a 5D XPL into my Olight S1 Ti. Found Turkeylords video on how to get the head open and it went well. Got the old board out, filed down a new one for the XPL, reflowed and tested the board, all is good. Got the board in well, soldered to the correct wires and it works… Sort of . The XPL in the XM-L2 TIR is a square kind of beam, not too bad, I can live with it… The XPL must have a different enough Vf to affect the UI, now the single click to off goes to moonlight and sometimes it cuts off after a few seconds, most of the time it does not. It stays in ML…

Now I need to find a high CRI XM-L2 and do this again…

X3, it was your post… I had the 5D ready for this light, a nice halogen yellow color… I have a 7A3 XPL, but not XML… I thought about the 3000K XPL HI too..

i must be getting really bored waiting for toys to arrive from china, lol.

i had spent a couple dollars on a clearance china-tastic zoomie at BIG LOTS. 3xAA zoomie, and a little bit thicker and longer body than i am used to seeing in a low budget zoomie. My several 1504’s are the only 26650 zoomies i own until the china boat gets here, so…

i am looking at the thicer than normal tube, and its a little longer than it needs to be,.. opened it up.

aluminum pill thank god, hollow no doubt but i didnt check. the driver has the LONGEST SPRING i have EVER seen on anything. really long… they made the light longer, then filled the space up between driver and 3xAA battery carrier with this long spring. Its SO long? it had a little plastic insert to keep the long spring from going catty-wumpus and missing the top of the + of the battery carrier. Hmmmm.

the plastic insert is sandwiched between 2 circlips riding grooves above and below the plastic tube insert… got them out, and now had more than enough room to put in a spare 26650 for long runtimes on the 630ma cheap driver. It works with the 26650 now, and i can bang it around it stays okay. for now anyways. Taking out the circlips allowed the plastic spacer to move forwards giving room for the 26650 to easily fit into it.

i have to make a little plastic tube and drill it out for the long spring to go thru it and hold it so it cant slip off the 26650, and, so i dont have to “know” how to put the battery in, i can just slap the 26650 in and go.

emitter looks like a little CREE with 3 bars on it… wonder what driver would be appropriate for it…i guess it looks like it might be an XP-E…

I have a copper S1 coming in the mail (thanks, X3), and am copying his mod soon, ordered a XM-L2 T4 7A3 for it, Simon has the led for cheap! Using a XM-L2 over XP-L (hi) saves you solving centering problems and beam artifacts.

djozz wrote:

I have a copper S1 coming in the mail (thanks, X3), and am copying his mod soon, ordered a XM-L2 T4 7A3 for it, Simon has the led for cheap! Using a XM-L2 over XP-L (hi) saves you solving centering problems and beam artifacts.

I, uh, sorta couldn’t help myself.

Thorfire S70 came yesterday just before dark. It was making 3657 lumens out of the box. Tried a couple of different cells, at 4.18V and 4.22V, really nice output but, well, you know.

So I cracked her open. Dropped an R070 on top of the 2 R091’s. Added a 18 ga spring bypass to the driver spring, 20ga through pcb on the switch.

WOW! Not having time after midnight to fully recharge the cells I’d used earlier in the evening I did manage to see 5700+ lumens… on cells at ~4.12V.

Over 2000 lumens increase! Stock sense resistance was 0.045, it is now 0.027. Fresh cells will show me over 6000 lumens on this big heavy deeply finned light. But the regulation was far better straight out of the box, only dropping 21 lumens in 30 seconds. With it resistor modded the hit is much bigger as it runs, dropping numbers very fast on the light box. Gotta charge up cells today and give it another look…

DB Custom wrote:

I, uh, sorta couldn’t help myself.

Thorfire S70 came yesterday just before dark. It was making 3657 lumens out of the box. Tried a couple of different cells, at 4.18V and 4.22V, really nice output but, well, you know.

So I cracked her open. Dropped an R070 on top of the 2 R091’s. Added a 18 ga spring bypass to the driver spring, 20ga through pcb on the switch.

WOW! Not having time after midnight to fully recharge the cells I’d used earlier in the evening I did manage to see 5700+ lumens… on cells at ~4.12V.

Over 2000 lumens increase! Stock sense resistance was 0.045, it is now 0.027. Fresh cells will show me over 6000 lumens on this big heavy deeply finned light. But the regulation was far better straight out of the box, only dropping 21 lumens in 30 seconds. With it resistor modded the hit is much bigger as it runs, dropping numbers very fast on the light box. Gotta charge up cells today and give it another look…

I got the Basen’s charged up and ran 9.05A at the tail for 5971.95 lumens. BUT it dropped 545 lumens in 30 seconds to 5426.85.

Stock, right out of the box, it only suffered a 21 lumen drop in 30 seconds. Obviously the heat sink and/or the cells can’t handle the bump all that well. In the few seconds of testing the cells went down to 4.11V. These are new Basen’s, only ran the few minutes of testing in this and another light.

Wonder if the Efest 4200’s will break that 6000 mark? Hmmmm……

Edit: Am I sure about the R091? Well I was til you said that! Yeah, I’m pretty sure, because I ran the parallel resistor calculator on em last night to figure if I wanted to run the R070 or R068 that I had on hand.

To be honest, one said R091 and they put the other one on upside down, ugh. So no, I didn’t remove it to make absolutely certain that the second one had the same value.

I re-did the short tube for my A6. The “old version” did not quite fit on the tail side even after modding. Since now I had three shot tubes from BG, that were all wrong I went full retard – removed anodizing and used dremel cut disk on the threads Also added lighted tail tailcap. It draws 0.2mA right now, but it’s still a bit too bright. I might redo it later on

djozz wrote:

I have a copper S1 coming in the mail (thanks, X3), and am copying his mod soon, ordered a XM-L2 T4 7A3 for it, Simon has the led for cheap! Using a XM-L2 over XP-L (hi) saves you solving centering problems and beam artifacts.

In the process of pulling the S70 back down some. The calculator says adding 4 1ohm resistors will give me .0385 in place of the stock .0455 so I’m going with that and keeping it within it’s abilities. About to double check that the mcpcb is DTP….

Edit: Mcpcb is indeed direct thermal path…

With the 4 1 ohm resistors in place I get 5.03A on the Basen’s for 4264.2 lumens at start, 4153.8 at 30 seconds. More reasonable losses I’d say.

While I was back in it I put 16ga Turnigy wire on the driver, removed the mcpcb and cleaned up the thermal paste to get a look at the shelf, it’s machined decently but has lines from the bit fairly pronounced. Used Arctic Alumina Thermal Paste putting it back in.

DB Custom wrote:

In the process of pulling the S70 back down some. The calculator says adding 4 1ohm resistors will give me .0385 in place of the stock .0455 so I’m going with that and keeping it within it’s abilities. About to double check that the mcpcb is DTP….

Edit: Mcpcb is indeed direct thermal path…

With the 4 1 ohm resistors in place I get 5.03A on the Basen’s for 4264.2 lumens at start, 4153.8 at 30 seconds. More reasonable losses I’d say.

While I was back in it I put 16ga Turnigy wire on the driver, removed the mcpcb and cleaned up the thermal paste to get a look at the shelf, it’s machined decently but has lines from the bit fairly pronounced. Used Arctic Alumina Thermal Paste putting it back in.

Dale, anybody, where are you getting your resistors of those values? I can only find .05 and .015.

Olight TC10 Ti rebuilt today.

You’re making a lot of beautiful lights lately, CRX. This is another piece of art!

(are those 4 squares neodynium magnets?)

CRX wrote:

Olight TC10 Ti rebuilt today.

That is some bad-a** light, but I could never do that to my baby

djozz wrote:

You’re making a lot of beautiful lights lately, CRX. This is another piece of art!

(are those 4 squares neodynium magnets?)

Thanks, yes those are magnets.

Henk4U2 wrote:

CRX wrote:

Olight TC10 Ti rebuilt today.

That is some bad-a** light, but I could never do that to my baby

Haha. Yeah i’m just . Do you ever use the AA body for yours?

CRX wrote:

Henk4U2 wrote:

CRX wrote:

Olight TC10 Ti rebuilt today.

That is some bad-a** light, but I could never do that to my baby

Haha. Yeah i’m just . Do you ever use the AA body for yours?

No. I keep this light in the glove compartement of my car, loaded with a lithium primary.

Really great idea CRX, that’s a very appealing look! I like it!

Hey, thanks.

The second red S2+ mod (the first mod is in post#400 in this thread), unrestrained this time, this one is for me! As a few have done before (firelight2 did a very nice S2+ 18500 mod before), I wanted the S2+ shorty to become a 18500 light, for the power and the runtime, and to see if I can do it . I ordered a light purple 18500 for it, and it appeared to be 49.8mm long, I had to find the length for that inside the light! After measuring and calculation this is how the extra length could be created:

*making it a triple

*make springs very short or absent

*shrink the pill

I wanted to be able to close the light fully, as it should, and keep the glass lens before the optic for easy cleaning and waterproofing, and I reckoned if I went for maximum pill chopping I could just do all this. I decided to chop the led-side of the pill and therefore keep the driver slot plus retaining ring, and solder a 1mm thick copper disk between pill and ledboard to keep enough thickness (total 2.7mm copper) under the leds for heat-disposal. So after sawing and sanding, this was leftover of the pill (picture with a BLF-A6 driver to test the maximum component height)

I just clears the Attiny! Also the underside of the pill was slightly shaved off to make another -say- 0.2mm.

A piece of copper was made to size.

Then everything was reflowed on my heatblock: pill->copper disk->ledboard->leds. I chose Nichia 219C warm white (3000K) for this mod: nice low Vf and good output even at 3000K. I made a video of the reflow to show-off my reflow skills , but unfortunately I had a more shaky day than usual, and further, despite several times soldering and unsoldering stuff, I got no light when testing, sofar for my skills .

But after some more fiddling and failures outside the camera, I found that I had rotated one led 90degrees wrong . It finally worked, and that is what counts :

For a driver I scavenged the driver from my aluminium Kronos X6 (that was the one flashlight of the four I bought that I have little love for, I hope that BG will sell this driver sometime to repair it again), it has great output and modes, and the firmware is tuned for the use of tail-lighting. Used very short 22 AWG silicon wires, and after assembly of the pill, I made the ledboard black with a marker to make it better looking through the optic.

The white balance of the camera makes the leds look like cool white, but they are very warm in real.

The driver spring was replaced by one of the short Intl-outdoor BE-Cu springs because they do not have to be bypassed and they can be compressed all the way to wire thickness. Still I removed a bit from the spring to make it fit the batt+ pad of the driver.

Now the head and battery tube were assembled (a somewhat thinner and clear o-ring was used before the lens to gain an extra fraction of a mm, and I did not like the green o-ring that the light came with). The 18500 was tested for length and when pressed in, and it just came out under the edge of the battery-tube: succes!!

I rebuilt the tail as well but I made no pictures. It still has the stock switch and board but everything better soldered, and it has a lighted washer now with six red leds, the equivalent resistance the same as the Kronos lights. For the time being the black silicon waterproofing insert is still in place so I can not see the red leds at all, I can not even check if they are actually lighted inside there . I hope to recieive a new transparant insert one of these days to make the lighted tail work! The spring was replaced with another of the BE-Cu intl-outdoor springs that compresses so nicely compact. Now the light closes neatly with the 18500 battery pressed in firmly without crushing it.

I'm glad that this mod went exactly to plan, now did the light work alright? I'm happy that it does, but of course on Turbo it gets hot very fast (it even does not have a chunk of copper to delay the warming up of the shell). I could only check the turbo output up to 20 seconds before the light got too hot to hold. But the X6/X5 driver has all those nice efficient lower modes too . Here's what I measured on the fully charged Efest lila 18500 sofar:

Turbo 2 seconds after switch-on: 2150 lumen, turbo 20 seconds: 1570 lumen, moon: 0.41 lumen, the other 5 modes of the 7+moon mode group resp.: 4.4-41-163-580-1230 lumen. At mode 4 (163 lumen) I measured an efficiency of 74 lumen/Watt which is pretty good.

The UI mode-change times are very sensitive to heating up, like all lights with this driver, but a bit extra because this is a quite over-the-top hotrod build. So turbo is just for showing-off, even one mode down is not sustainable. I get a sustainable (heatsink to my hand) output up to ~500 lumen, which is fine for a small light like this.

I tried to do a current measurement on Turbo, but with the tail removed and the current going trough the small copper loop that I use for clamp meter measurement, the driver immediately drops to moon everytime I  start turbo (even mode 5 and 6 do this too). I found this behaviour before with my AK-47 converted FET-drivers and do still not know what happens. With the tail in place everything works fine. Well, judged from the output, the current must be a bit over 10A.

And the beam is great, the Nichia 219C works lovely with this Carclo narrow optic. The tint is rosy, as expected from Nichia, a bit too much to my liking, but it gives a comfy feeling .

I have a couple of Sanyo 1600mAh 18500 cells on order for this light: more runtime and a more sustainable (but lower) turbo.

Sweet mod djozz. I felt your pain in the first video when power was applied and nothing. Nice comeback though and what a pocket warmer.

Wow. Really cool mod djozz.
Thanks for the future modding idea as I have a couple S2 shorty hosts and at least one 18500 laying around. I’d like to ponder the heat sinking a little.

nice work,djozz

the threads in the s2+ must be deeper into the housing than the plain s2
as ive had to remove the threads from pill to do mine and hold in place with a sleeve
under the battery tube .

Great job there.

I always like watching your re-flow videos. It’s like a thriller movie where you are just waiting for something to jump out at you. It must be the shaking