For a show-off light I’m happy to drive a led at the current that gives the maximum output, but for a user-light I play it a bit safer, I go for the current that gives me 80% or for a thrower 90% in my led test. Not only because that probably improves the lifetime, but also because I expect the heat path in a flashlight to be a bit worse than on my fan-cooled copper chunk. The W1 rewards that a bit with a better tint too.
At 4A the W1 in my led test was at 96.6% of its maximum output, 5A gives marginally more output at ideal circumstances, but I expect some costs over time.

The 3amp biscotti driver in a c8+ already gives nice power to w1. Led/host gets pretty hot allready with a full cell. I personaly wouldnt go higher than that with a flat white in A smaller host. L2 can handle more heat. Atleast if you sell them make sure they know it gets hot.

4-4.5A is a good current
not many LEDs will handle 5A in a flashlight, they mostly tint shift and may turn blue after a while

Okay, I’ll leave as is at 4.3A.

I really appreciate the replies,thank you

I told Simon that there’s interest in the option to use OSRAM CSLNM1.TG 1mm and he’s ordering some to test. It would be a great addition

And tell him to finally make some nice old style zoomie flashlight as addition to his reflector lights. He is long time into this business and he never offered zoomie light…

I tried to convince Hank to make aspheric head for Noctigon K1 which is a perfect candidate IMO but no luck. Maybe other people should bring it up :slight_smile:

For those of us who played with them, aspherics can be a major (or minor) PITA. In the past, BLFers offered aspheric lenses for the HD2010, and P60 lights. I bought both, of course, and did get them assembled but both were a PITA, and somewhat disappointing in results - not as throwy as we thought.

Dunno for sure why they are looked down upon, could be the poor quality of parts available for decent prices, or the complexity to get something working well and reliably. For sliders, PITA with the water resistance vs. the air pressure problems, the slide or twist mechanics - just a bunch of nasty issues.

Both Simon and Hank are into quality/reliability, so zoomies might not be a good option for them.

For me I was only suggesting a fixed aspheric but you make good points for zoomies

Then you have your decent zoomies like the B158. I think this LED would be great in a B158. I have one of 2mm2 in a Cometa it performs good.

A build with his ‘5A’ linear bistro Biscotti driver delviering 4.62A and WF1 in an M2 yielded 80Klux. I think there’s more left on the table if there an AR/ultraclear lens and a bit more current. This is a SUPER setup IMO.

BTW, does anyone have access to the Boost HL (4040 1mm WF) or just ONE sample they’re willing to part with? L4P had a batch that sold out and I can’t find them available anywhere. Really wanting to push 100Kcd (turn-on) out of the M2. Not going to get it with the 3030 for sure…

They also over the w1 and w2 in green color. Even higher candella rating. Could be good for tactical lights?

I’ve asked him about a zoomie before as well. Maybe others could join? :wink:


Hmmm… Simon and Hank should not be afraid of their own OEM zoomie…, and regarding quality/reliability reflector lights should not be only option.

Imho in a future (if we will live to see) I would like to place a bet that zoomie lights will prevail… That is already happening with IR spectrum illuminator lights and new VCSEL laser modules which gives ultra high performance greater than any Osram and similar LED IR emitters and of course suitable host for them are only zoomie lights. So for example small 20 mm VCSEL zoomie is equivalent to 69mm Osram 4715AS zoomie. Runing driven at 0,7A vs 2,2A of Osram led. They are running cooler, longer runtime so only 1x18650 cell is more than enough + ultra light host. So it is just a matter of time until laser VCSEL technology will appear in visible spectrum… VCSEL are not that expensive either…

There are waterproof zoomies on a market so it should not be problem to make them waterproof especially if they will use acrylic lenses instead of glass ones. Obviously only Brinyte (and maybe just few manufacturers) figured that /acrylic plastic is more consistent, more resilient (shockproof + lighter) and better for mass production than glass.
Lets say twist zoomie: double o ring at the tail, double oring at lens section, and double oring on throat section… More than waterproof :slight_smile:

My favourite approach to waterproofing is to add a secondary lens, could be a flat one that completely seals the light engine. Effectively a waterproof mule with added zooming aspheric. That’s simple and reliable. And doesn’t add friction. :wink:

As Lexel mentioned above, 4-4.5A is best for reliability.

Bistro? Don’t you mean Biscotti?

Biscotti version uses his temp driver which only pushes 4.3A on my W1, does Biscotti increase the current? I thought it would be the same

I don’t think the firmware changes the current. I measured another driver running an SST20 at 4.28A. I could have had a measurement/zeroing error. Also, please excuse my misnaming of the FW. I meant Biscotti.

As for reliability, I think 4A/mm2 is pushing the limits. I’m obviously chasing some numbers here which is admittedly not practical. Just having fun with it. For a duty/EDC light, I’d run much more conservative drive levels.

Convoy’s new 17 mm 5A driver with biscotti uses temp regulation.

Get rid of the NTC to eliminate the temperature stepdown and allow max current

It’s still outputting max current until the step-down occurs, right?

probably not, since the NTC is an analog element.
It sets the current limit depending on the actual temperature