NightWatch Avaritia NA40

I found the triple XHP50.2 version on a Chinese site but not sure where to order it for the US. Any idea if the triple 50.2 version uses a different driver than the SST40 version? The 50.2 is 6v right?

Oh… it makes sense and Amutorch wouldn’t be the first one to use BLF drivers in their flashlights…

It should be a different driver, as there are 7135s in the sst40 version, meaning the emitters are in parallel.

Would the xhp 50 have more lumens or the sst40?

Every second time I read the name of this light it’s “Avaria” instead of “Avarita”.
And “Avaria” is a homophone of a Polish word for “failure”……

Same thing in Portuguese :smiley:

yup, xhp50.2 would do maybe over 4000 lumen per led, so 3 led would be over 10000 lumen in theory
but finding or making the driver is pain in the @ss
so SST40 is more suitable for this light

Thank you! I went ahead and pulled the trigger on this one.

Try an “Arius” knife. I always register it as “Anus”. :person_facepalming:

I’m thinking swapping the 3x50.2 with 3x Nichia 144a 90CRI would have great results.

Maybe 3x LH351D 90CRI would perform almost as good too.

Anyone got one of these?

From my perspective a far more interesting question about this light is the power source. It is shown a single 26650 battery, and to deliver 6000 lumens, is going take 20 amps if not a little more, and there are not a lot 26650’s out there that can sustain 20 amps for a long period. The more output you want, such as from xhp50.2, the more power you need. 10,000 lumens is going to require 30+ amps, and if you thought 20 amps for a sustained period was difficult, 30 amps is much much worse, and in either case, the heat dissipation is going to be very big problem, very quickly. The Keeppower IMR 5200mah according to HKJ’s testing isn’t good for much more than 20 amps.

What diameter is the driver? I wonder if there is a Narsil driver that could fit…

Is the switch lighted?

This is not such a good cell for high drain due to its heavy voltage sag. Something like a Liitokala blue or black or a Shockli 5500 would be a lot better choice. The really hot ticket would be the Samsung 30T 21700 which can do 30A continous. Depending on the driver design, a too powerful battery might burn up the driver or emitter. More testing is needed.

Any pictures of the driver? If the button is seperate from the driver then I'm sure Lexel has a FET driver with Narsil that will fit.

There is a 6500K version and 7500K version. :confounded:

Well better than a 2500k and 3500k I reckon. At least we would be able to tell somethings lit up.

Yes, 2400 lumen would be it’s output in raw form. Output drops a bit once installed in a light.

In the video
Chris does not have an integrating sphere to measure with. He is using a ceiling bounce measurement and his correction numbers might be off a little.

I imagine 1800 to 1900 lumen is more realistic for OTF lumens. So this NA40 could in theory be doing 5500 lumen with a FET driver and good battery (not so much the KeepPower). 5000+ lumen is a realistic number to me. I’m curious to see output tests with a good battery.

Is it a triple xhp50.2?

For that light you would need a boost driver for sure and current designs are limited in output. The Acebeam EC65 is a quad xhp35-HI running a single 21700 cell. It can put out 3750 lumen. So right around the 45 watt range (Same as Lexels and Loneoceans boost driver limit). If you have 3 xhp50.2 in 6v parallel on a 45w driver and a good high drain cell you can expect a bit less output than the NA40, maybe 4000 lumen, but have it as steady output, where the FET driver immediately starts dropping voltage and output.

IDK, this xhp50.2 version sounds strange and there’s a reason we have not seen that style of light. Further investigation is needed.

I could only find info on the Nitewatch EC21 which is a single xhp50.2 rated at 2200 lumen at 3A.

It’s also made with the sst 40 or xpl-hi.

Then Nitewatch also has a xhp70 light that is rated at 3600 lumen at 4.8A.

I rcv'd mine last night. Measured 13.2A at the tail w/clamp meter on turbo (LK 5000 cell with a magnet on the Batt+ end), 5.45A on hi. There's 16 7135's and two FET's in parallel. FET's are used for turbo, max 7135's used for Hi mode. On turbo I measured 4,640 @start, 4,280 @30 secs, 36.5 kcd taken at 5M. It's a 14 pin MCU, ~21 ohm resistor from Batt+, then to a diode, then to pin #1 (PWR), pin #14 is GRND. Driver is 32.2 mm wide - I don't have any of that size, but enough cavity room to sandwhich one in.

Looks like a copper DTP, though I didn't verify it's a true DTP. Thermal grease pattern is a bit strange. Screw was attempting to hold it down or simply in position, but the screw wasn't threaded in the hole - odd. LED- wire is also oddly soldered but probably no loss's. I got 7,890-6,530 lumens on a triple SST-40 light, but used 4P 18650 35E cells in a bigger host (SP03), so could be the lumen #'s make sense. My modded C8F does about 5000 lumens @30 secs using XPL2 V6's and 19.1 amps using a VTC5D VapCell.

Springs are very stiff and copper look - suspect good quality so bypasses probably won't help much. Tail has no spring, but a simple ring (no pic). Stock LK cells rattle, but adding a thin magnet on the Batt+ end fixed it. Battery options are limited because of the weird tail end - circular contact area is too big for 18650's, batt length must be just right, etc.

It's a nice quality light - easy UI, nice FET/7135 driver design, thick shelf, copper DTP (probably), red and green LED's under switch (4 wires). The head has some weight to it which is a good thing. It seems like SS bezel - has weight.

Not liking the LED alignment pieces - little slop/play results in not centered though not sure if it cause any problems in beam pattern or output. 20 AWG wires, long to be able to pull out the driver. Clearance of PCMCB wire soldering is tight, so upgrading to 18 AWG could be tricky.

It seems like more amps should be possible with an LK - not sure if wires, springs, and batt magnet account for it all. Batt+ must pass thru the 2 spring charging board setup. In theory, triple SST-40's should out do triple XPL2 V6's. Could be restrictions with those FET's, driver layout, etc., as well, or maybe the LED alignment pieces could not be helping. The C8F is really a top shelf host, about perfect in many ways.

NOTE: driver retaining ring threads oppositely - CW to loosen

I hate to bring this up, and I don’t mean any ill will, but your using the Manxbuggy1 lumen tube, right? An educated guess looks like it could be reading 10% to 15% high. Some other Manxbuggy1 tubes seem to read about the same. This is a whole other topic, though. Sorry.

Back to this light, it’s quite possible that @30s it’s only doing roughly 3800 to 3900 lumen. This is much lower than spec. I was hoping for 5000lm at least, but I think that would require about 20A at the tail. (Maybe a Samsung 30T would get the job done?)

I have no idea on that driver spring, maybe it’s low resistance, maybe not. I would swap it with a longer known type of spring and do a bypass on it. That could eliminate 2 potential problem areas.