Convoy L6 UV Rock hunter build 10W+ LED -need build advice

Flashlight Modding and DIY Parts
1

Tons of info on this forum, thanks to all who post it.

I’m hoping to build a10w+ flood for rock hunting based on a L6 host
I’m hoping for 2hr+ run on high.

please suggest the best 365nm 10W+ bulb, energy efficient driver 10W+ capable, long life batteries and reflector choice for this L6 build. I will be using a ZWB2 lens filter.

thank you for any help/advice you can offer.

2

That sounds like something that will put the recently popular Fyrfly C8 build (C8 + LG led + buck driver + two cells) to shame.

A triple or quad would probably be better if enough heat dissipation but I don’t think they make a L6 sized triple reflector or an extension tube for a Sofirn C8F or Convoy S12 to use a buck driver with 18650s or just a boost driver suitable for UV leds. That’s a niche market that needs to be filled.

3

I’m not sure a super thrower like an L6 (when paired with a small die 3535 led) will be very good for mineral hunting, it still throws even with a XHP70, put a small die emitter in it and it’s a pensil beam… Is there a reason you want a thrower for mineral hunting? If you’re searching at your feet Id go with a flooder, maybe a Q8 with 4 of the “new” large die LG 3535 365nm emitters.

4

yes. specifically Yooperlites. ill be searching at my feet or directly in front of me. they react very brightly. I was hoping for an 8’ circle at my feet.

5

I have 2 Fyrflys. they work great. I’m hoping for a better light for my application.

I am a machinist and can produce a head out of copper if heat is all that’s stopping this build.

The L6 can hold the more powerful 26650 cell, sort of the drive for the whole project and a bit more of a flood style light.

6

There are not many choices for an emitter that powerful

one is the Nichia NVSU333A which is not cheap nor easy to buy, but fits the XHP70 footprint fine

you can expect from it about 4.5W UV output

https://www.nichia.co.jp/de/product/uvled.html#NVSU333A
https://www.nichia.co.jp/specification/products/led/NVSU333A-E.pdf

you need a buck driver that runs it at 4.5A, you may have a look on mine

on the other hand if you go to a light like Sofirn C8 you can use 4 3535 LEDs that output more UV with comparable throw
just here you need a driver with many AMCs as the original is only FET+1, while most UV LEDs don’t like DD

7

Any cost estimate in USD? I really would like to push the L6. My current flashlight says 2.7W of UV. 4.5W UV is perfect. at what watt level do the LEDs start being expensive?

8

Tabao list it for 500 Yuan —> 70$

9

is there a better multi led flashlight that uses one or two 26650 batteries? or just do a Q8 with a filter and 365nm LEDs?

10

With a linear (or FET) driver with a single battery you won’t have full output for the entire battery duration, that’s why the FyrFly build uses a pair of batteries and a buck driver. Still with 3x18650 you have 9000-10000mah of total capacity, but output will likely decrease below 1A per emitter after about 4000mAh even if you use the lower Vf LG emitter and 4.35v batteries like the LG E1. Also I don’t think the Q8 is available in host format, so it’s going to be a bit more expensive that needed.

11

8 feets = 2.4 meters at your feets. That’s a lot of flood! I’m not sure you actually need a reflector for that, I think even a mule would work. That likely opens the way for other hosts.

Look here and check if a mule beam pattern is suitable for you. Of course, without a reflector it won’t be very useful for longer distances. Most TIRs are not suitable because they will block UV.

Edit: Still I think a triple/quad reflector would be more versatile. A mule may be too floody. If you have the ability to make an extension tube for 21700 I think a Convoy S11 or Sofirn C8F with a buck driver would be suitable, 21700 batteries top at 5000mah. The C8F should handle heat better, uses an e-switch in addition to a tail switch, but surely you can find a way to seal the hole.

12

There are UV 365nm LEDs that have at 1A only 3.4-3.5V

you can also choose a light like Acebeam TN40S but that is quite pricey and you would need a driver mod to lower from 4A to 1-1.5A
you got here a 2S Boost so it will always give 100% power

good 365nm emitters don’t need an additional UV filter like Nichia or Luminus SST-10
they produce no waste light above 400nm

13

are there any higher output UV LEDs at 365nm that would swap directly in to the q8?

14

8 feet or around that. its more important to eliminate the hot spot.

15

I ordered a Sofirn BLF Q8. What is a high output 365nm LED I can put in there, even if it takes some modifications ie; soldering, fitting, machining? What can I do to get it to flood a bit more and eliminate the hotspot?

16

“__good 365nm emitters don’t need an additional UV filter like Nichia or Luminus SST-10
they produce no waste light above 400nm”

This is incorrect. There are no UV LEDs which are suitable for mineral fluorescence without a filter (unless you don’t care about the visible light they emit polluting the fluorescence). All 365nm LEDs (or even 265nm LEDs) generate tons of visible light. You can easily prove this by taking a UV blocking filter (camera filter, glasses which block below 400nm, etc) and look at the LED.
Worse, 365nm LEDs generate tons of 380nm to 400nm UV which most people can see. This is the dim purple light that you see even with the filter.

17

I’ve built a couple of S11’s with the quad LG part (around 7w radiant). 16Vf, 1.5A - gets hot real quick, I don’t leave it on long. Very floody beam with the stock reflector. (I posted a pic on the FB group)

18

take a Nichia 233B LED and then add a ZWB3 filter, you will see no dramatically change in the light purple glow
to really block all visible light you need a narrow bandwidth filter, which has at least OD4 or OD5 above 380nm
the downside of this is you lose a lot of brightness, and they are expensive, you also narrow in the absorption bandwidth which can prevent fluorescence at 375-390nm

An SST-10 or 233 Nichia can output above 1W of UV and the purple glow is extremely low, so its easily overcome by fluorescence,
while LG or Chinese LEDs produce a lot visible waste light above 400nm which seems white

19

ZWB2 actually better for 365nm. ZWB3 is for shorter wavelengths and transmits much higher into the visible range: .
ZWB1 even better for blocking visible but worse for UVA transmission.

Agreed the Nichia parts have less visible output, but for mineral use they are still unacceptable. This visible light must be blocked. Only the brightest fluorescent minerals will “overcome” the visible light, and non fluorescent areas of the rock will appear blue/purple. Rocks with multiple fluorescent minerals will really look crappy, especially if one is dimmer than the other. Dimly fluorescent minerals will not be noticed when collecting in the field because the visible light will overcome the fluorescence. This is really important because we use these lights to find any rock which is fluorescent; if dimly fluorescent odds are pretty good they will “shine” when you get home and put under SW UV.

LG LEDs are the best choice for power and cost. When filtered they are better than Nichia because they are more powerful.
We have a saying in the fluorescent mineral hobby. I have flashlights with the Nichia 233/276 parts, LEDEngin, and LG parts - “Friends don’t let friends use unfiltered lights” -it’s really true.

But - if FLM ain’t your thing, then it might not matter….

20

yes but still the ZWB2 filter is not really blocking everything above 380nm so always that purple glow visible,
sure it gets better but far from good 20nm wide 365nm bandwidth filters that reflect >99.99% of all above 375nm back into the light so it gets absorbed by a few passes on the reflector