【convoy】H4 and B35AM options are updated

Any chance we will see just the host for sale soon, since that’s not dependent on driver supply? And will there be a non-e-switch version. Thanks!

How’d you manage to get that shipped to the US? Pretty slick light. I can’t even see the listing for it unless I follow a google link with no ali in cookies/history…otherwise it blanks it out because he’s honoring that patent and won’t ship here (same happens on Banggood lately).

Might look to see if that emitter was covered in Djozz’s tests but a quick peek at the data sheet for it says a max of 1.4A and with UV emitters that max is usually accurate (as opposed to white emitters which we happily overdrive the heck out of). That driver would optimistically give 2.5A per emitter which I suspect would fry them in short order unless legit tests have shown otherwise. Not sure I’d push 2A on them if they’re anything like the ones from Nichia and LG in terms of respecting that current max on the sheets. I think his 6A drivers would be fine, maybe use smaller wires and remove any spring mods if present. Seems like that driver puts out something around 5.6A or so based on other comments here.

I got it through a group buy on r/flashlight. A couple users in & outside the US imported a bunch of them.

I guess I need to visit reddit once in awhile, sounds like a great buy.

Guess you are speaking about the “8A driver for CULPM1.TG”, doesn't it? That is a buck driver, it will deliver the full output current for as long as its input voltage is high enough. In this sense it behaves as the linear variable load drivers like the 6A units you mention, if only with higher efficiency as it is a switching driver.

Correllux says “optimistically 2.5A per emitter”, I guess you are talking about a triple. But 8A / 3 = 2.6̅ = ≈2.667A, which is slightly higher. Don't presume, but measure.

If you want to make 100% sure that the driver doesn't frys your emitters the key is to change its onboard sense resistor. For this you first take a close look at the driver board, find the sense resistor and note down its value. The sense resistor is a bigass, low value precision resistor which should be near the edge of the board. I think it is a 10mΩ unit in such driver. Then, you calculate the driver's sense voltage: Vsense = I × R, and so Vsense = 8A × 10mΩ = 80mV. With this in mind, you can now calculate the maximum regulated current for other sense resistor values: for 20mΩ, I = Vsense / R = 80mV / 20mΩ = 4A; for 18mΩ, I = 80mV / 18mΩ = 4.4̅A = ≈4.444A; for 16mΩ, 5A; for 15mΩ, 5.3̅A = ≈5.333A…

Just in case it is of help had this link with 1206 R015s for sale in my wishlist, which is nice enough I think. However, I do not know 100% sure if the onboard sense resistor is 1206… O:)

By the way, I wouldn't build triples in parallel, it's waay better to set them in series and use a boost driver, because:

  • With leds in series the current is exactly the same for each emitter.
  • A boost driver (like this “XHP35” one) ensures full output drive current and power regardless of input voltage or battery state of charge (as long as it meets the bare minimum, of course).

Of course you need a compatible size and arrangement triple MCPCB (KD has this thing, but no idea if it would fit).

wow, thank you for this great info, Barkuti! I won't use it for UV anything (no interest in uv atm), but I will use this for other 3x/4x emitter builds + adjusting driver outputs! Appreciate it!

Hello simon is green 18350 short tube coming back for s2+ anytime soon?

The VF value of UV LED is much higher than that of ordinary white LED. I use a 6A drive and the test result is 2.4A current.
I have not try 8A driver, The current should be higher, maybe 3-4A

The mechanical switch cannot withstand such a large current.

Yes, in production, but it is rectangular thread, I don’t think it is suitable for your flashlight.

Simon, I see you’ve released a remote switch for L21B. Is it possible to remove the remote switch from the L21B tailcap and install it in any other Convoy 21700 flashlights with a tailswitch? Perhaps S12 IR?

Are you speaking about the 12-groups switching buck driver with up to 8A output?

Despite what Simon says, using a high current driver means that if the conditions are met (this is, enough effective input to output voltage difference, and low enough emitter Vf), the current at the emitter could go above the maximum it can tolerate. When I say “effective” I say it because the input voltage at the driver diminishes as the input current goes up, because of current path resistances (springs, switch, etc.).

The correct solution is always to adjust the output current at the driver, this requires adjusting the sense resistor value in it. The advantage of this approach is that all the driver modes will upscale or downscale proportionally. For the 12-groups high current buck driver, swapping the stock R010 with a 40mΩ unit (R040) nets a maximum of 2A at the output, and with a 33mΩ one (R033) the maximum output current would be 2.4̅2̅A = ≈2.425A.

Mon, 10/18/2021 - 18:15

Can I ask which S2+ UV light would be best for photography illumination (such as a model with UV body paint). What is a “safe” on-time?

I sold it as you said before.
But the sale was not good, so I gave up.

All ultraviolet rays will cause certain harm to the human body.
Among UVC UVB and UVA, the damage of UVA is the least.

Any info about the availability of the SFT-40?

So, once again… I have to sing the praises of Simon and Convoy. My S11 FC40 developed an odd driver issue, on mode option 1.2 the 2 low modes worked fine and the three upper modes were all the same dim brightness and the light was always indicating low battery voltage. Contacted Simon about the issue since I wanted to be sure to get the correct driver ordered… he just sent one.

Put it in today and the S11 is as good as new. I really like this one and the M21B with the FC40, would hate to not have either one of them.

Thanks for the excellent service Simon!

Simon Will you be getting any more green 18350 body for s2+ with trapezoidal thread?

the threads will be unified into rectangular shape, so I have no plans to continue to produce 18350 tubes with trapezoidal threads,sorry

I’d be a lot more interested in how these outputs are holding up after a few minutes.