EAGLE EYE X9 XHP70LED inside Hard Light flashlight 26650/18650 3800LM

Sorry no pics yet. I ordered the N4 3C, but it's hard to believe it's an N4 bin, but the tint looks to be a great NW - milky white consistent tint through all zones of the beam.

Tail amps on LK 5000 cells: 4.7A, lumens: 2,850 @start, 2,770 @30 secs, 36 kcd (380m) measured at 5 m

This is well under the 3,800 lumens rating. I'm not familiar enough with what 4.7A should get you with an N4 bin, but for it to be 73% in my testing of what it's rated is pretty far off. For example, last night, along with the X9 I measured Lumintop SD26 and SDMini lights and both measured higher than their ratings: ~1,060 lumens on each with both rated at 1,000 lumens.

Those are bad news… What does EagleEye have to say about that ?

Usually the NW are noticably not as bright as the CW. I didn’t even know Cree offered a N4 3C. The highest I remember seeing was N2.

I believe EE only rates the CW version at 3800 lumen.

Even the CW N4 versions need more than 4.7 amp to get 3800 lumen. Closer to 5.25-5.5 amp is needed.

I guess it’s time to see if it can be resistor modded.

It's right here: http://www.mtnelectronics.com/index.php?route=product/product&path=60_114&product_id=597

So if the info posted by Richard is correct, there is a N4 3C. I know NW's can measure lower than CW's on our cheapo light meters, but for 5000K, mine has measured top output levels in the BLF Q8's for example 3D tint in both XM-L2's and XPL's.

I'm thinking it's not a true N4 though. But again, would love to know what others are seeing on XHP70's and at what amps.

I tend to go by the cree data sheets. It’s the 4500k and below that only get up to N2. My bad.

If the driver is only capable of 4.8 amps, then I wonder why Eagle Eye rated the lumens at 3800? That’s a bit optimistic.

Have you looked at the driver yet? Does it look like it can be modded?

Djozz has plotted the xhp70 output so that gives you a good idea how many amps you need to get a certain lumen output.

djozz test of an N2 CW is here: https://budgetlightforum.com/t/-/30860/576

So at 4.5A, it does ~3,400 lumens naked, so allow 10%-20% optics drop, for 20% it's 2,720 lumens - right smack dab in the ballpark.

Some explanations:

  • I'm using 4.5A and not 4.7A because I measured with the tail cap removed using a reliable clamp meter. Tail cap may add resistance.
  • 20% may be a bit high for the reflector and glass, but again, in the ballpark

I'm thinking the NW XHP70 is an N2 bin and not N4 based on this quick & dirty analysis.

So maybe the CW is a true N4 so will be bumped and maybe I'm getting somewhat lower amps for whatever reason, or maybe the 3,800 is a bit optimistic. Thinking Dale has a CW so hoping he can post his numbers. Our PVC light boxes are very close to the same calibration from past experiences.

For me I still really like this design and the light - it sure is a compact 2 26650 cell light, and can certainly double for a pretty effective weapon with the thick and somewhat sharp SS bezel. If I want more power, easy enough to do, and could replace the LED if I want. I'd probably not do resistor modding but prefer to do driver replacement - NarsilM of course.

Would you compare it to a 2*D maglite in terms of physical dimensions ?

He already did. See post 39.

I got 4.84A for 3212 lumens out the front, using a pair of Efest 18650’s. Have yet to test it with LiitoKala 5000mAh cells.

Edit: Okay, so I just got done (9:50AM 5-10-17) charging and testing two pairs of 26650’s, Efest 4200mA and LiitoKala 5000mA. I have sliced the dome on my XHP-70 in the X9, that said, on the Efest cells I got 4.32A for 3084.3 lumens and on the LiitoKala cells I got 4.30A for 3015.3 lumens. Two different 365nm UV lights show this XHP-70 in the NW range.

Oh dang, forgot or missed your #'s posted (my bad). You should be able to see in your Ali order history what tint was ordered. On mine in the order history summary it says "Product properties: N43C".

The CW option is N41A. The differences is interesting - could be in the realm of one bin difference, though amps is slightly different.

I didn't measure amps at each mode but did measure lumens:

lo: 29 lumens, med: 561 lumens, hi: 1,309 lumens

So I'm lower across the board, consistent with a lower binned LED, or consistent lower amps... Boy, hard to tell though but still I think we measure with our PVC light boxes pretty close.

Yep, my AliExpress order history shows N41A.

Of interest, I’m using an XTAR Rocket charger (SV2) with 2 bays and 2A charge rate capability. It seems that when using the 2A rate the cells don’t really get topped off properly. So I charged the LiitoKala’s again at 1A and tried them in my Convoy L6 with ramping firmware on an FET driver.

At 2A rate they did 11.21A for 6610.2 lumens
At 1A rate they did 11.98A for 6796.5 lumens

So this charger has different performance when using the higher 2A rate to charge 26650’s.

Dale, you got half an amp more on the smaller batteries? That’s pretty interesting. In theory the 18650’s would have more voltage sag and produce less amps.

This reminds me of a weird amp reading I got on my Convoy L6 with a stock driver. My protected KeepPower 5200’s gave me 0.5 more amps on turbo compared to my Liitokalas. After a resistor mod and after a direct wire mod the Liitokalas gave me 1.3 and 1.6 more amps compared to the KeepPowers. There was something about the stock driver that liked the “weaker” Keeppowers. IDK, it’s a mystery.

Sure, I can understand the charging effecting the results. Did both charge methods result in the cells settling to the same voltage level? Maybe at 2A, it settles/drops a bit?

Ok, I posted to Eagle Eye (Mr Fang?) detailing my tests and methods, and if he can confirm the N43C option is a true N4 bin LED.

Ohh - took apart the head assembly and I like the setup - pretty large MCPCB, copper, real thermal grease under it, neat on the wire soldering, nice setup for the LED centering piece. That SS bezel does have good weight to it. Confirmed it's a little thin around the area of the MCPCB (heat source), but figured that. Looks like the pill is quite deep - didn't look there yet though, but seems they could have saved some mm's in the length of the light there. The driver retaining ring is a nice thick brass one.

Really like the overall quality.

Remember to take note that I sliced the dome off my XHP-70 in this X9, so the fact that it’s still doing just over 3000 lumens with the sliced dome is pretty good, the tighter beam pattern works well with this reflector. :wink:

The Efest 4200mAh cells, recharged at 1A, give 4.36A for 3087.75 lumens. According to my initial readings when I first took the light out of the box, that’s only a little over 100 lumens lost to the slice, I’ll take that every day!

With the very nice opaque 18650 sleeve that came with the X9, I used a pair of freshly charged Sony VTC6’s.

They showed 4.35A for 3087.75 lumens.

With 3000mAh the run time will still be quite good and it’s so light it feels like there’s no cells in it. :slight_smile: (This is a relative thing, of course, all about perception. My Triple XHP-50.2 weighs almost 3 pounds. :wink: )

I got access to the driver. It's a tight fit for sure, but the body around the pill area is pretty thick.

The R010 resistor soldered to the LED- wire is the obvious candidate to resistor mod - but this one can be very risky!

For driver replacement with e-switch support, maybe best option is to work out a switch PCB because a standard size (like SRK) would fit, then you can get rid of the vertical board.

The space is tight in the pill, maybe 17 mm wide hole, the top of the toroid inductor is very close to the the top, so that's it for space.

Eagle Eye replied and confirmed the NW is a N4 bin.

Nice work Tom, good pics on the driver and corresponding components as well.

With that R010 resistor there’s not much headroom for modding, in my experience it’s this kind of driver that produces the obnoxious magic smoke when tampered with. Makes it a likely candidate for driver swap…

Ouch, ok. Noted above now.

I still can't get my head around the readings - somewhat lower than expected output based in the tail amps. Maybe the driver has lower than normal efficiency - could be part of the issue. I would need more time - look at LED amps for example. Maybe try another LED, etc. Figured I'd post the pics I took the other day.

No time now - world is wait'n on my firmware updates/fixes for the Q8 . I'm probably holding up 1,300 lights or so for shipping approval .

Yeah, don’t know what you were thinkin messin around with this light when so much of the global community is chewing their fingernails to the quick waiting on you over at The Miller’s…

:stuck_out_tongue: