Review: FandyFire K2 1x26650 XM-L U2 Flashlight

Thanks for the info RaceR86. Nice pictures. So using the DrJones driver you would use the OEM switch?

@RaceR86, are you doing a mod thread when you install that driver? I'd like to see how it's done with this type of switch.

I can't find the right size o-ring to swap. The closest match I could find was 29x2.5mm.

Thanks for the great info R86! My measurements were tail current, and I think they match well with your emitter current. Needs more drive on high, for sure.

I was not planning to, but since you asked...

Mod thread here.

It should answer most questions... :)

Just got an FF K2 from WallBuys, think it was an instant kill, paid $22. Really liking it so far though. Same excellent packing, the o-ring seems ok so far - doesn't seem loose, UI is the same, exactly as described, a real AR lens (wow - unusual for a FandyFire).

I measured:

-- On A KK ICR 4000 button top at 4.14v, at the tail 2.36A:

lumens: 799 @start, 785 @30 secs, throw: 22 kcd

I got better #'s than Relic38 across the board. Not sure, but maybe I just got CREE lucky and/or it likes the KK cell better. Got similar results on a Samsung 20Q (low resistance). Really liking this light though - the size, feel, side switch, and the modes aren't so bad. This is the 3rd side switch light I got in the past week (SupFire M6, Convoy L4) and it's sure making tailcap switch lights feel awkward and cumbersome.

The XP-11, K2, and M6 make an outstanding set of comparable light in 3 distinct sizes. I'd say the K2 fits perfectly in between in width and length.

I received a solarstorm k2 from blackshadow a week ago. I have zero gripes about this light. The ui does not bother me at all after becoming annoyed with the supfire m6. As long as the strobe and sos are out of my way then I am happy, I will never use them. And 90% of my other budget lights contain annoying disco modes that you have to cycle through sometimes. I don’t mind the low start, it only takes a second to click up to high.

No issues with the o-ring. This light is so solid and feels great in the hand. The machining is perfect and the anodizing looks high end. I love the side switch and glowing green button.

I almost gave up on 26650 lights after buying the trustfire a8 and ultrafire version. Those two lights were awkward with their size and tail switch. Plus the overall quality did not compare to the k2.

I’m glad i purchased this light.
I have to agree with Tom. After handling this light for a few days. I don’t feel the urge to handle my
tail switch lights anymore. For some reason my solarstorm k2 appears to be just as bright as a convoy m2 running nanjg at 2.8ma.

785 lumens is sort of amazing for an XML U2 at 2.4A, and on aluminum? Something fishy there - Got to open mine up and check, but maybe the alum star has a direct thermal path, or they upgraded to copper. The lens and reflector look perfect though, and the low profile SS bezel helps too - from other posted test results on black vs SS bezels and their size. Actually looking at it compared to an A8 is night and day - A8's output has got to be suffering from the infringing deep bezel while this one is wide open. This light competes well with bigger, better driven stock C8's.

I think the A8 could be as bright as the sun and I still wouldn’t care for it. Just an awkward size with a tail switch. I’m becoming a side switch snob now.

Is there any differences in the solarstorm and fandyfire k2? My solarstorm came with the little plastic ring removed from the driver/battery + contact board. And the O-ring fits perfect. Maybe the workers stretched a few o-rings in the past during installation, causing the loose fit.

I’m fine with the output of the K2 as is. I love the feel, look and anodizing. And that little glowing button.

I bought one of these a few months ago and love it.
Seems to last long enough on a panasonic 18650 3400mah and it throws really well considering the price. I can easily light up houses at 200m and the battery indicator light is really helpful/neat.

The finish on it is on par with my supbeam k40 and 100x better than the cheap cree c8’s i have. I have dropped my k2, 3 or 4 times out my pocket whilst on my motorbike ( ie. from a few feet up at speed on to tarmac ) and it has survived well, althuogh there are one or two chips in the finish now :frowning:

Overall fantastic light and worth every penny I paid. ( £15 gbp )
For the price i cant find a torch as bright, that throws as far or with as nice a finish for that price here in the uk.

Mine was listed as solarstorm but seems 100% identical to the fandy fire.

Rex

Pretty sure the Solar Storm and FandyFire K2's are identical. Dunno, but somehow they get the max out of this XML emitter in this light for the amps it uses. Yes - 22 kcd I measured is 296 meters for 1/4 lux (moonlight).

Update: This evening, I decided to go ahead and take a looksy, which turned into a full take apart. I can confirm my FF K2 does have the same LED star and pill design, including the hole in the middle (wuz up with that?). I was able to pop out the driver with pressure through space by the switch using a solder pic tool. It was glued in, probably by thermal adhesive. I was also able to pop out the aluminum pill easily with a little force from a hard push from a finger. It could only come out one way based on the cut out for the switch, so it was pushed towards the battery end.

I believe the R010 resistor is a current sensing resistor because it's not wired directly to a LED wire, but goes thru a couple of FET's (3 legged things...). I'm going to go a little mild on a resistor mod that would get maybe 30-40% increase. I think 4A+ is too much for this little guy - the reflector is heavier than the pill, and the body is quite thin. All the weight of this thing is in the battery holder end and the reflector -- quite weird... I'd like to keep the stock driver - the UI isn't awful awful, and nice to keep the green and red LED's active for a battery status.

Thinking a R020 would give a 33% boost on top of the R010, so will give that a try, if I got one...

Tom E, I think the light may have been updated. The driver seems to drive more current on hours. 2.4A at the tailcap on a fully charged cell is closer to 3.2A at the emitter. That would explain the bump in output. I think the reviews do get back to the manufacturer and that may be why many of my negatives (O-ring, drive current) are addressed. DX sponsored the review, and I have no doubt they would read it and want to address the problems. If I am right with the assumption, the one thing they missed was posting here that changes were made and sending someone an updated light. Opportunity missed, IMO.
Anyway, glad to hear that the light is improved.

Ohhh, good point, thanx Relic38!! I did notice one resistor value that differed from the pic RaceR86 has in post #8 -- it's the 103 resistor just below the yellow cap - can't recall the new value but it's much smaller (from 10K). Also all the epoxy Racer86 ran into around the switch is not on mine - easier to remove the driver and pill because of that, maybe.

Also that white disc on the button shown in his pics fell off on mine, but think I could just epoxy it back on easily.

Looking at the thickness of the stock MCPCB, I'm going to add a copper disc under the Noctigon to re-produce the height as close as possible. I'll reflow the disc to the star, and then can re-use the hold down screws, so this should really help with thermal issues, at least in the most critical transition area.

Also will definitely do an amp measurement to the LED while I have it all apart to confirm this, before doing any resistor modding. There really isn't much heat sinking in this light in the critical areas of the pill and MCPCB and I'm not that good with metal work to do much with it.

I have spoken with the manufacture before and mentioned that they should build a 4xAA light with similar looks and build quality to the Solarstorm/fandfire K2. Add XM-L2, Keep the similar looks and glowing button. But make it a little fatter, machine channels in the tube for 4xAA without a dumb carrier.

They said that they might start the designing process after their new year holiday. I told them to pick up a sunwayman d40a as inspiration. I guess we will see.

Stacked 2, then 3 R075 resistors (lowest I had) on the R010 and it had no effect whatsoever. Under a loop, it looks like the solder connections are good - dunno why this didn't work, because it should get a 20-35% bump. I'm reading about 2.9A at the LED, open on the bench, basically same value before adding any resistors, might have gone up from 2.8A - not sure. Gotta take another look - wish there was some way to measure the resistance of these resistor stacks - always seems too low for my DMM, even a fluke. The R010 just seems to stand out as a current limiting or sensing resistor - nothing else looks even close on this driver.

Update: All Done!

I re-did the resistor soldering and seemed to be getting 3.1A to the LED with the 3 R075's. So, left it at that. Upgraded the wires to 22 gauge, added some AS5 in the body where the pill slides tightly in to, added 2 dabs of Fujik epoxy to hold the driver in. Also added a dab of solder on the driver for the + battery connection. Tried the same KK cell - now read 3.1A, hhhmmmm. Buttoned up, looks good and bright.

Note: Got a XM-L2 U2 1A on a 20mm Noctigon, reflowed on a 7/8" 22 gauge copper disc. Sanded down the star and disc to minimize height, maybe 0.1mm higher than stock. Re-used the screws, used AS5 on sanded surfaces to 2000 GRIT.

Measured:

lumens: 1,088 @start, 1,064 @30 secs, throw: 34 kcd, 369 meters (measured at 5 meters)

Not bad for being moderately driven at 3.1A, wish I could have gone higher of course Smile. The button green/red lights still work of course because it's the stock driver. Putting in a e-switch based Nanjg driver would lose the switch lights I would think.

Great work Tom E! Maybe I will get doing something with this light someday. I think the current path could be adding up to additional resistance. Getting over 1000 lumens OTF is quite good as you have it.

There's more...

MNKE IMR 26650: 3.4A tail, so better #'s but losing 100 lumens at the tail. So, tried my best battery: Powerizer:

Powerizer 26650 @4.18v, 3.99A tail, and that did the trick: 1,248 @start, 1,214 @30 secs

Thoughts

  • the driver doesn't seem to regulate well because low resistance cells get a higher tailcap amps reading, higher lumens output (maybe because of my resistor mods?)
  • weird, but the Powerizer seems to compensate for the resistance of the tailcap spring pretty well (not much if any drop). I assume though as the Powerizer drains and loses voltage, droppage is more significant becuase the spring's resistance will kick in
  • Run this light on a Powerizer!! Gotta order more - I have only two
  • The glued tail really sucks on this light, as so on others. If I could get the tail loose, I could copper braid the spring, though the tail spring is an odd ball in that it seems held in by the tail cap tightened (and glued) to the body.

Pieces:

New setup of the LED/star. Noctigon reflowed on the copper disc, then disc and pill top sanded smooth, AS5, and secured tight with screws.

I wonder where @Relic38 has been, haven’t seen him post since last year.

good

Tom and Relic38: How do you guys measure the lumen values? I built an simple IS. It works pretty stable to compare different flashlights but I just can’t determine the absolute values.
After I used the sun method to calibrate, the measurement results are very low (~250 lm for U2 at 1.5A). It could be my lux meter not accurate but cannot be that much off, isn’t it?