What did you mod today?

[INeedMoreLumens]

@Jeffgoldblum

how much can you get out of one 18650 really

With a FET driver, FET Switch and Excellent springs you can achieve 20+ Amps.

In my opinion the switch would be the weakest link so the FET Switch should help in that regard.

[Jeffgoldblum]

@INeedMoreLumens

With a FET driver, FET Switch and Excellent springs you can achieve 20+ Amps.

In my opinion the switch would be the weakest link so the FET Switch should help in that regard.

We need to steal more tech from vapes. Vapes used have a FET as the switch sometimes. I think they called them Clickfets lol. They’d be rated for like 40-100a

I took apart an old vape today for parts, cuz I haven’t used them in years. It was 100w off a single built in 2500mah “20600”. It was boost and buck, you could set output voltage on the fly from 0.8v to 6v with buttons and an LCD screen. It had a FET bypass mode. Thermal protection. LVP. All the contacts are thick chunks of brass and copper and aluminum. I think I paid like $50 for the thing. Why don’t we have those

[Quadrupel]

You say vapes have stabilized output?

[INeedMoreLumens]

I haven’t made any yet but loneoceans designed this FET Tailswitch for the S2+:

GFS16 - Battery Indicator Tailcap Light & 1mR FET Tailswitch! (Rev B)

Hello BLF, TL;DR: I built a tailswitch system comprised of two PCBs, called the GFS16 System. First is a 18mm Taillight board which provides the regular taillight, and also features a battery sensing system to change the LEDs colours when the battery voltage is low (Rev B features 5uA standby current!). Second is a 16mm Tailspring/FET board, which allows replacement of the switch for a super low resistance FET, allow for the best performance in high current flashlights. All PCBs are up on OSHP…

[Quadrupel]

Yes, yes and that switch must to be powered with extra power source to operate. Its just nonsense. Its more simple to use “e-switch” design

[gravelmonkey]

Back in the day, Overready improved switch resistance with their zero-res twisty. Basically a clicky bypass relying on the thread anodising.

[INeedMoreLumens]

That would work too

[Jeffgoldblum]

@Quadrupel

You say vapes have stabilized output?

Idk lol. Now I want to know what happens if you soldered the leads from one of these right to an mcpcb and set it to ~3v. Somebody has tried that right? This had to be a thing at some point

[Halflight]

(I’m sure many of you know this already) Given the whole vape ‘mod’ (literally referring to ‘modified’ flashlights) community was pretty much born from flashlight tech I’m surprised there isn’t more apparent crosstalk these days. I know many early modders were also into torches, this inspired them to harness the power of high drain batteries to create an alternative to the ‘plastic cigarettes’ coming from oriental sources.

[Jeffgoldblum]

Makes sense. I quit vaping years before I bought my first flashlight, guess I missed that whole era.

So then why aren’t there more buck+boost+fet flashlights??

[Halflight]

Indeed, perhaps the market just isn’t big enough for more complex systems outside high end collectors items?

[Jeffgoldblum]

Had to take the L70 hi out of my Z1 because it cooked the spring on the convoy 12a FET driver somehow. Wasn’t a short, just a few turbos. Noticed it was flickering and not as bright as it used to be, checked the spring and it no longer sprung.

So took the opportunity to replace it with the glass L90 hi. Switched the 22awg wires on the driver to 18awg because I cut them too short and had to replace them anyways, so why not, and bypassed the driver spring this time.

It’s a little cold…

I thought for sure my opple was broken, I had to check some 519a’s right after to see if it was off, and no, its working fine.

On the plus side, it’s a lightsaber even indoors

Yes, I hit the chandelier

Feels like a round, 10,000k sbt90

[TUIG101]

There’s another L90 on AliExpress, equally cold but it has twelve bond wires instead of only four. Don’t know if it’s really important but it should handle more current on direct drive.

I popped the lens off on mine and the projection got better, there’s less projection of the region around the die and the focus got better.

The L90 performed way better than the L70, despite using the same die, the difference was visible even at low current. Too bad it is so cold.

[Jeffgoldblum]

Ooo that’s definitely worth looking into. I don’t know how important the extra bond wires are either but you have to assume they must help current handling right? They can’t hurt.

Hey the on KD has 12 bond wires in the picture too. Or it does now. I wonder if it did when I ordered it…

[TUIG101]

There might be different batches of this LED, each one with varying bond wire count, just like the 5050 had a lot of variation on tint.

I brought mine from this listing, but there’s the possibility it comes with only four bond wires just like your did bcuz these LEDs certainly came from the same unknown place and were made by the same unknown manufacturer.

[Jeffgoldblum]

True, I mean the CCT was also off by about 4000K lol. Seems significant though. 4 bond wires instead of 12.

[James_C]

Wurkkos FC13S - Sliced the XHP 50.2. I did a little “dicing” (removing phosphor around the perimeter). I also added some minus green. The beam is better now but could probably stand to do a little more dicing. I’m always afraid of removing too much phosphor and letting out the angry blue. Perhaps blackening the gasket will help, too.

[merlot]

I have been experimenting with a reliable way to get a decent tint out of the xhp50.3 series. I have found that dedoming the HD version and making a little gasket out of foil (& superglue) to stop side emission produces great results. I put one in a convoy s21a with one of their 4A boost drivers and it works very well. Unfortunately the s21a OP reflector produced a very ugly blue spot in the middle of the beam, so I replaced it with an s2+ OP reflector.

I will be replacing the driver with the 5A driver from convoy when it arrives (maybe adjusting the sense resistor for more output if it can handle it), and the magenta AR glass with green AR. I’m curious to see what chip the 5A driver uses, and what the input current limit is. There are a few with a 10A limit which won’t allow an increase in output but we’ll see.

The emitters are the 4500K 80CRI HD from mouser:
XHP50D-00-0000-0D0HH445G

[Jeffgoldblum]

@merlot

I’m curious to see what chip the 5A driver uses, and what the input current limit is. There are a few with a 10A limit which won’t allow an increase in output but we’ll see.

Do these help at all?

[thefreeman]

@merlot

I’m curious to see what chip the 5A driver uses, and what the input current limit is. There are a few with a 10A limit which won’t allow an increase in output but we’ll see.

Yes it’s the usual MP3431, in the 2.5A 12V drivers the Input limit is set at 10A average, I assume it’s the same here, it can be increased by modifiying the ILIM resistor (the actual limit is set higher because it’s the peak inductor limit).