No-name '6XL2' dissection

There's a lip around the ID of the rings that covers not only the substrate but even part of the dome. Way way too tall and that inward lip isn't needed, at all. I'd guess that only the front 1/3 of the reflectors are getting hit with any significant amount of light.

Would be nice to see the difference in lux (ceiling bounce) after you've `helped` it a little bit.

they use the raised ring to block the light that hits the deepest part of the reflector, and which causes the messy halo around the hotspot. So the hotspot is nice and clean with those rings, people want that. And it does not matter that you loose some spot brightness in the process, people can judge a clean hotspot, they can not judge brightness by eye. (In practice for most people: a nice distinct hotspot equals a bright beam)

Well this thing has rings like Saturn. If that's what they were trying to prevent they failed miserably.

The 'Securitying' lights have too-tall rings too, but they don't have the inward-facing lip at the opening. They start out stock with a pretty good beam, but get better when the centering pieces are corrected.

Before and after:

I haven't pulled it apart yet but I'm pretty sure the 6XL2 is using these: http://www.fasttech.com/product/1581400

my srk are from fasttech too and the driver are diferent




!all leds are in paralel :smiley:



I have add micro usb chargue









As-delivered numbers from the calibrated I.T. (Integrating Toilet™) with others for comparison, since 'calibrated' means 'not actually calibrated':

93.5 - stock 'Securitying' SRK clone, 3x XML cool white (unknown actual bin/tint)

100 - stock CNQG SkyRay King, 3x XML T6 3C

245(!!) - modified 'Securitying' SRK clone, 3x XML2 T6 3B

117 - stock 6XL2

So tell me if I've got this right:

There's nothing between the MCU and the FET's gate, right? So if I short the gate to BAT+ and output current goes up, then there's PWM on high limiting the current (meaning, the FET is capable of more, but the MCU is commanding it to do less than 100% all the time). If I do the same and it doesn't increase, then the output is being limited by the FET itself and a better FET will crank it up.

Would there be any benefit by sticking a toroid inline between BAT+ and LED+?

What is the lumens of this? and runtime? Does it beat SRK?

See the lux numbers from various SRKs/clones in post #7 - it's a little brighter than a normal SRK but not by enough considering it's six XML2s up against only three old XMLs. But the ugly beam makes the little bit of extra light not really worth it.

I haven't checked the current yet but since it's the same basic setup as a SRK (4x 18650 in parallel, LEDs in parallel, linear driver) and close to the same light output the runtimes should be similar to any other SRK.

Yep you are right with the Pwm test. But normally you can check it if you shine on your hairs cloth or something near your ear, or I front of a water faucet and watch the shadows….

To test if the FET is limiting you have to do an additional bypass test, connect the battery- to the led minus. Direct drive gives the max current which can flow.


I really like the charger model in the back, I wish I could unscrew the back of my m6.
—-
Anyway which colour temperature does these xml2s have?
6xml2 are worth 30$ on its own, so the host costs only 10$…

If the PWM is high enough it won't be visible. Not seeing it doesn't mean it's not there. There's no visible PWM on this one's low mode but it's definitely there, the singing springs are a dead giveaway.

The tint is somewhere slightly warmer than 1C, cooler than 3B.

edit: I already know the output current is being limited, because it ain't bright enough... I only care whether it's due to the FET or the firmware. One I can fix, the other I can't.

my skr don’t have pwm in high mode, I have removed the coil

the way to get the fullpower is “repair” the spring resistance and use at least 2 samsumg inr 18650 20r

I never had a light with so high Pwm that I couldn’t spot it.
For example the nanjgs with 9kHz are easy hearable for me if I point the light on a fleece pullover or my hairs next to my ear. Also swinging the light shows it or the faucet trick….
This FET sucks anyway I have one on a spare 12T6 driver from kaidomain and even on 10V it has 30mOhm…and we all know in direct drive every volt counts.

6parallel xml2 will get serious current if direct driven…


If you decide to stack another FET and get rid of the old controller with Pwm on high…and if you are willing to be a test object ……I could provide you my srkfirmware…
You just would use a nanjg with the flashed firmware instead of the stock controller.
momentary soldered to the nanjg, ± to nanjg and tiny output to the FET gate…

I just received this noname flashlight from fasttech. It had some issues with the driver (sometimes it was hard to turn on, then hard to turn off. The amount of light had been switching itself from a glow mode to very high etc.) now it seems to work fine. Im just wondering … is there a driver which i could place in this flashlight so that the output is higher? Id like the LED’s to receive like 6A on high mode. Any suggestions?

There's this: http://wallbuys.com/Product/3-7-4-2V-3-Mode-CREE-T6-Driver-for-SkyRay-King-3T6-LED-Flashlight-Golden-Black-46-2mm--10684

Since this light has a 1-piece MCPCB and only uses two wires to run all the LEDs, on that old-style SRK driver you'd have to tie the driver's 3 sectors together to one common point, then connect the LED- wire to that.

what batteries are you use? the Panasonic ncr18650 3400 have 110 mohm and the samsumg inr 18650 20r have 40 mohm

Im using Samsung 26F M

That is what I want to do with my SRK as well

Was told the batteries might loose contact and no longer charge evenly...I definitely like it though...most awesome!

I posted an update here yesterday and it's now gone, what happened?

Machined the base of the reflector flat (not shown, I also milled out two little pockets for wire clearance):

Switched to different style insulators (http://www.fasttech.com/product/1581402)

Before:
93.5 - stock 'Securitying' SRK clone
117 - stock 6XL2
After:
94 - stock 'Securitying' SRK clone
174 - 6XL2 w/insulator mods
Those numbers are with both lights using Samsung ICR18650-28As charged to 4.20v, on INR 20Rs the Securitying goes from 94 to 120, the 6XL2 from 174 to 215.