OSRAM CSLNM1.TG & CULNM1.TG 1mm², CSLPM1.TG & CULPM1.TG 2mm²

I mean those are nice drivers he mentioned. It wouldn’t be right to recommend something less than $9,999

Nice 2x21700’s hope they come out with it soon… robo…
Pic stolen from the Maxtoch site… by me…

@BobbyMk: What driver are you using

Qlite 3.04A with added 3 7135 chips and biscotti fw.

Thanks!

I’ve built a C8+ with ld-A4 at 5A and white flat 1mm

I need to tune the centering ring heigh but at the moment I’m getting 211kcd nearly 920m

Is there any similar build? I’d like to compare them and also know how many cd should I expect

I will work on D1S and likely some other sizes CC FET drivers, but this will take 5-6 weeks at least to get to production if first design works out well

I’ll start saving money for you, Lexel.
There are some other drivers i want from you too.

Same, I’ll buy one or two for sure.

Did you measure the beam with a lux meter at what distance then convert to candela?

Yes, 3300lux at 8 meters

What CC FET driver design? Think someone (TA?) did one? Wondering, not sure, if parasitic drain was considered.

I’m very impressed. My C8 (BLF A6 DD, NCR18650B) is easily hitting houses at 800m distance(measured with Google maps). That’s the holy grail of LEDs for throwers.
Can’t imagine how the D1S will perform.

190kcd for the c8 measured at 10m.

Put a white flat 2mm^2 in an Emisar D1S. I wanted to add some resistance to the circuit so I could use the FET driver and limit the current to 6-8A. Switching the springs to high resistance steel would work if you had the right size springs. The battery compartment length is very tight on this light so it would have to be the right spring. I chose to add a separate resistor in the head of the light. I figured I would need to add about 0.1Ohms to limit the current enough with a 30Q cell. The resistance would dissipate around 4W. There is a good amount of room in the head to put the resistor. One could probably use the right combination of 1W resistors in parallel, possibly using thermal adhesive to sink them to the head so they don’t overheat. I chose to use something I had around, some 16 gauge nichrome heater wire. Two inches or so of the nichrome wire has about 0.1Ohms. A difficulty with this wire is that you can’t solder to it. To attach the stranded copper wires I lashed them together using fine gauge magnet wire.

Using a clamp meter I measured 6.0A with a charged (but old) 30Q. Ideally I would have wanted closer to 7.5A, so I might shorten the resistor. I used a XM sized “butterfly” style centering ring that fit the emitter hole of the reflector, then I just fiddled around closing the bezel until it finished centered. I had to add two layers of Kapton tape under the centering ring to get the focus right. I measured 221 kcd. This intensity measurement is consistent with the measurement of this LED at the same current in the EE X6 (111 kcd). This reflector has right around 2x the frontal area (I estimate 1270mm^2) of the EE X6 reflector (~620mm^2). It’s always satisfying when things make sense. :slight_smile:

EasyB,

That all is if you insist on using high current cell with fet and WF1mm… Low current cell like LG 18650D1 even in mod with classic bypasses wont pas 5.7A and without bypasses wont pas 5.5A

But of course very creative idea.

Samsung 30Q and other high current cells are only good for old G2 S4 2B (which is still one of my favorite emitters).

Edit: Just saw that is WF2mm… OK… Not bad idea indeed :+1:

Yikes! EasyB, please be careful with that thing! :open_mouth:

EasyB,

what light meter do you use? Generally I have read about inconsistency of cheap light meters, so I don´t know still what to buy.

IF member A, and member B, and you, and me all had identical lights (model, cell, led, driver) and held this in fromt of the same model luxmeter we would have four different outcomes. Because there are too much variables to consider. Fresh cell’s, topped-off cell’s, waiting 30sec (yes/no), distance from light to meter, etc.
The only outcome that is reasonably accurate is the difference before/after mod of each individual member.
If you say it is +15, that can be 13 as wel as 17%, but it is still more reliable than comparing the outcome of member A after his mod with that of member B before his mod.

Djozz once measured that UNI-T UT383 in comparison with mobilux one and told me that uni-t under reads for about 15 % de domed XP-G2 S4 2B and I think Djozz or someone said that same light meter could easily over read White flat emitters because of tint or hue I am not sure…

If that is true than theoretically except better tint of WF1 (well for mine use kinda greenish warm tint of G2S42B is better), no need for de dome, better run time there is no much difference between them. G2 is still more suitable for FET drivers and has larger die so greater hotspot in aspherics.

But if I just check side by side same light with different emitters(g2 vs wf1) without lux meter it seems that WF1 throws further but I can’t give real values (around 15% for sure).

I just want to say that even cheap lux meter like UNI-T UT383 will serve well to tell you where you are with your build.

Ha, well actually it was not the safest setup. I was not thinking quite enough and put the resistor on the positive side of the LED which left the possibility of shorting the battery if the anodization wore through at the head and at the battery tube-to-head threads.

I shortened the resistor some and put it on the negative side to reduce the shorting risk. Start current with the same charged 30Q is now 7.0A which should get me an extra 8% or so of output. Ceiling bounce test puts it at very approximately 1400 lumens OTF (83% of a dedomed XPL V6 at 6A from a EE X6).

I use a Tondaj LX 1010B lux meter.