On a whim I pulled parts out of the bin and made a new bait fish fishing light.

Sc02
Green XP-E2
BLF SK68 driver populated with a baby FET on a 7135 pad. E-switch 13 (TomE) firmware.
Horrible tail cap setup replaced with my typical Cu disk / spring press fit in place of the stock crap.

Total spent: zero

Tried an upgrade of the BLF Q8 from the XM-L 4500K to Samsungs LH351D 90 CRI 4000K.

Reflowed those new LEDs

Shiny new screws for the MCPCB

100uF 1206 capacitor to reduce problems with bumps:

And an AR-coated lens

But sadly, it looks like I damaged one emitter somehow.

Strange, since one emitter definately made some troubles when reflowing

My question is: would you replace that LED or just live with it?

EDIT: I forgot I ordered a spare LH351D on a 16mm MCPCB. Everything fixed now.

Interestingly, in the tests with ceilingbounce, it went from around 5500 lm to 6500lm, so that partially damaged LED clearly didn't pull it's weight.

My big advice for everyone trying to reflow such a big MCPCB with a hot air rework station:

Remove any nozzle attachment so you can heat the biggest area and hit it with maximum airflow from below.

Works like a charm.

Also, don't use too much solderpaste, it will only short-circuit.

NovaTac SPA Defense SPL-120 – Nichia 219C 4000K & 18350 cell mod.

Since my D4 didn’t get much love. I swapped the 219C for SST-20 4000K and put Anduril on it. Much better now!

P.S.: crx nice and clean as always!

I put SST-20 4000K LEDs in an E2L triple (again). Using a FET+N+1 from Lexel with Bistro HD.

This time they seem to be surviving. The last SST-20’s turned yellow on me after testing a high drain battery on turbo. Maybe my reflow was bad, guess time will tell.

Does about 2,400lm at 0s with VTC6

I also added a clear switch boot and drilled a tiny hole in it for an ice blue trit. Thanks to Dale for showing me this. Just enough glow to see it in a dark room but not enough to bother you and it doesn’t drain your battery.

LichtAn wrote:

Since my D4 didn’t get much love. I swapped the 219C for SST-20 4000K and put Anduril on it. Much better now!

P.S.: crx nice and clean as always!

I kind of like how pointy the beam looks in my triple with these LEDs. What do you think? Between these and the LH351D you really have both extremes of beam shape covered.

Any idea what the output is on your D4?

vinte77 wrote:

Newlumen wrote:

Very nice.. is this light take 26650 too?

Thanks, yes it does. I have the light running on a liitokala 26650.

You mean using a liitokala 26650 on the Fenix HM50R? How come?

Does emitter swapping count? First I bought a preflowed mcpcb with a 4000k Nichia on it for my H03, did a simple mcpcb swap. Then last night I did my first ever emitter reflowing, I swapped that 4000k Nichia for a 4000k LH351D that pairs better with the TA driver I put in that same light. It was a success!

Also, that TA driver, i got it from Lexel, prebuilt. But I also got some bare boards and all the little bits that go on it, and I built two of those out a couple nights ago, my first time! And they seem to work fine. I haven’t tested them fully in a light, but I flashed anduril and ran through the modes with a test led and it all seemed fine.

I love this place so much! Thank you all for your help. You guys are awesome.

You’re doing great beastly!

Been away for awhile and this thread has gone mad. To many mods to comment on but lots of terrific work going on. Well done guys and keep posting your mods up.

contactcr wrote:

LichtAn wrote:

Since my D4 didn’t get much love. I swapped the 219C for SST-20 4000K and put Anduril on it. Much better now!

P.S.: crx nice and clean as always!

I kind of like how pointy the beam looks in my triple with these LEDs. What do you think? Between these and the LH351D you really have both extremes of beam shape covered.

Any idea what the output is on your D4?

Tbh I’m a bit worried to run it on full turbo. The heat generation feels about the same as the Nichias, which is pretty extreme. I don’t want to fry anything (yet). I might start to take measurements with 4.1V and go up from there.
The beam is still very smooth, no defined hotspot. Not really my type of beam, but better than the even less focused 219Cs.

LichtAn wrote:

contactcr wrote:

LichtAn wrote:

Since my D4 didn’t get much love. I swapped the 219C for SST-20 4000K and put Anduril on it. Much better now!

P.S.: crx nice and clean as always!

I kind of like how pointy the beam looks in my triple with these LEDs. What do you think? Between these and the LH351D you really have both extremes of beam shape covered.

Any idea what the output is on your D4?

Tbh I’m a bit worried to run it on full turbo. The heat generation feels about the same as the Nichias, which is pretty extreme. I don’t want to fry anything (yet). I might start to take measurements with 4.1V and go up from there.
The beam is still very smooth, no defined hotspot. Not really my type of beam, but better than the even less focused 219Cs.

As long as your reflow is good you should be fine (especially with something like a VTC6/30Q). My triple has less output but higher amps per emitter I would guess and it seems to be doing fine on the 2nd try.

If you are using the smooth narrow optic that is probably as tight as it will ever be with a multi emitter TIR. Maybe try a C8F or something if you want more hotspot. The LEDs definitely are capable of making a nice tight beam though.

I have LH351D, 219C, XP-L HI and SST-20 all in triple carclo optics and you can see the difference in beam shape pretty noticeably between them. I listed in order of most floody to least.

I can only measure Lux though. But when I know the current I can probably estimate the OTF lumens pretty well with an output chart.

Edit: 4.16V VTC5D gives 18.7A. 10 seconds until its too hot to touch. That’s 4.7A and 933lumen per emitter according to maukka’s output chart. Let’s say 80% leave the front, so roughly 3000 Lumen OTF.
It’s such a nice 4000K tint.

I finally built the Ultimate HiCRI BLF Q8

Replaced the tailcap with dual Beryllium Copper springs for maximum conductivity and brightness.

On the stock XP-L HDs, with an AR lens and brass screws, along with the new spring arrays, that got me 6764 lumens at 30 seconds. The lumen number compared to others might be a bit higher since I’m using a more efficient lens(6500 without the AR lens), but I’m still extremely impressed, considering I didn’t do a spring bypass.

Now for the real star of the show: the 4000k 90CRI LH351D. This LED almost has beautiful color reproduction, and is more efficient than most other 3V 3535 HiCRI LEDs, meaning not as much brightness is sacrificed in exchange for the higher CRI.

The tint is also better than the 219C, with less green than it. It’s also quite a bit floodier, meaning it gets a more balanced beam.

All in all, this resulted in an amazing beam, and a renewed interest in the light. I now use it absolutely everyday, and I now use it as a soft light for photography and night hunting, and as my big EDC

Parting shot

I have to say. I really like this mod, as it made the Q8 my new favorite light

Nice! I plan to do a very similar mod. I may combine 5000K and 4000K LH351D emitters, though.

I put three 4000k 219c’s and one 3000k 219c in my D4, I like the blend.

Hmmm I didn’t think about tint blending.

That would have made it even better, since dynamic range of the beam would be boosted.

6,764 lm @ 30s is quite a bit higher than what djozz reported for his similar mod. What batteries are you using?

@contactcr, I mentionned that this was for the stock XP-L HDs with all the mods possible in the light, which include the new springs, AR lens and brass screws. I’m probably getting a nice boost just with the AR lens itself, about +/- 200 lumens.

The 4000k LH351Ds 90CRI are less efficient than the XP-L HDs, and they get around the stock numbers, so around 5100 lumens, which was to be expected. The light is getting much hotter than stock though at similar power levels due to the much lower VF at the same current, meaning they pull quite a bit more current than XPLs. They are quite close to their max brightness. I wouldn’t do any other mods, so no fancy 16AWG power wires, no wire bypass, as the difference would still be minimal, no FET change, etc.

LH351D: 3,53V at 5A

XP-L HD : 3,81V at 5A

Consequently, even lower modes get hotter.

Edit: I’m always testing with Samsung 30Qs at 4,19V.

contactcr wrote:

6,764 lm @ 30s is quite a bit higher than what djozz reported for his similar mod. What batteries are you using?

Unless you have the privilege to work with top-of-the-bill instruments, that will have to be calibrated in a regular interval, all measuring is pretty much a relative thing. If Blue SwordM says that a certain mod will improve things by 5%, I will believe him. If djozz says the same thing, I believe him too. What I don’t believe, is that both of them get identical results when they unbox their lights for the first time. The difference may be in the lights itself, it may be the measuring equipment they use. And the batteries, of course.
If you or me perform the same mod, we will have an identical outcome. As a percentage. Not in the actual readings of the meter, before or after the mod.

@BlueSwordM that explains it much more clearly i missed the XPL
HD part since the rest was about the Samsung LEDs. I have done this exact mod and got similar results as you did.

I need help on how to install driver from .mtnelectronics in c8f with firmware like the emisar d4

You could start by telling us which driver from mountain electronics.

The 17mm fet driver for electronic switch

Sal415 wrote:

The 17mm fet driver for electronic switch

You have to reflow LEDs to the MCPCB they included first then attach LONG LED wires to that. Before you screw the MCPCB to the base of the reflector be sure to line up the wires with the milled out holes in bottom of reflector and screw hole. Here you can optionally cover the wires in kapton tape if you have it. Not necessary though. Screw MCPCB to reflector making sure to keep your LED centering rings in place. Apply solder paste to the bottom of the MCPCB. Next, feed the wires through the shelf into driver cavity.

There is a long screw that goes in via the driver cavity which secures the reflector to the shelf from the under side. You can put this in now while you can push the wires out of the way easily.

Next, get about 5-6 hands or clips or clamps or whatever that will allow you to hold the driver close to the body so you can solder the 2x LED wires and 2x switch wires. You can trim the LED wires (dont trim switch wires they dont carry current) as close as you feel comfortable being able to hold the driver and still be able to solder it comfortably.

Next use your tweezers or whatever and gently tuck the wires into the darkness and secure the driver with the retaining ring.

Here is a diagram of where to connect switch wires. They are tiny and hard to solder so dont spend too much time with the heat. If you dont get it let the parts cool back down and wait to try again.

Connect switch leads to OTC pads or MCU Pin #2 and ground. Picture below:

Completed my BLF D80 UV Mod (Seoul Viosys Z5). I salvaged a old AMC based driver (removed 3) now working with 5 AMC for 1,75Amps.
Its bright! On your skin you can feel the heat after a few seconds. To bad my phone is not able to get a good picture of a blacklight picture. It is so bright when i use the light that my phone is not giving a usable picture.
So i took a picture of a box with tailcap rubber boots and some GITD o-rings. One with the light 1,5m away and one a few second after i turned of the light.

That D80 looks good with the black filter and the ss bezel!

It fits nice together. And the light can handle the heat of the led much better than a S2+. Even after a few minutes its not getting hot (+ you get more output). But i did not notice the interesting pattern in the middle yesterday.

I’m glad that you regard that pattern as interesting
Is it just that the light is very close to the piece of wood so that you see the middle of the reflector and further away you have a normal spot? Or does that spot remain at distance, in that case you could try to project the led with a projection lens and see what is wrong with the led.