Review: NAGE 3xXM-L 4x18650 ~~~~Drivers Out~~~Adding Chips~~~~Trouble!!! ~~~Fixed~~~Finished (post #31) It's Bright!!!

I felt compelled to post this general overview of the Nage 3 x XM-L in case anyone has been looking at it. Let me start by stating that there are two versions on the market. There are Nage and an Ultrafire 3 x XM-L and 1 x XM-L. They are the same lights with different branding. I have already posted my build of the 1 x XM-L version under the Ultrafire JS-8054 brand. I will not be getting around to modding this one for a bit, so if you want to have a look at the driver, you can refer my Ultrafire post. !

Another piece of the head unscrews, but it seems to be glued on the Nage. I didn’t really try to twist it very hard, so it might just need a bit of elbow grease. Removing it isn’t necessary as it is only a separate component in order to convert it to the 1 x XM-L version. This is a truly ruggedly built light. Nothing is thin in its construction. Heavy walls and pill. I was surprised at how thick it was when I had to drill through the top of the pill on my Ultrafire. I believe these hosts can absorb all the heat you can throw at them. Here is a comparison with the 1 x XM-L version (under the Ultrafire logo). I’ve had this light sitting on my table for a week and never noticed that it is shorter than single emitter version. It’s just a matter of a lower profile head component that accommodates the triple reflector. !

This is one roomy light. Whether it’s the tailcap, or the pill, there is just a ton of room work with. I think you could fit any driver on the market inside the pill.

The driver is like a giant NANJG split into three sectors in parallel with the emitters. Each sector has spots for six AMC7135 chips. I haven’t wrestled the driver out to see if all the spots are occupied, but judging by my tail readings I believe they are. If you want, you can stack another level of chips for increased current. Because of the size of the board you are not as constricted as when stacking on a 17mm driver.

My single emitter version only had seven chips. That left empty slots for eleven more chips and link all three sectors with a short bridge. I’m expecting a Nage 1 x XM-L in the mail tomorrow, so I’m curious to see how many chips Nage uses as compared to the Ultrafire. I’ll post some photos of all three together when it arrives.

These are fast becoming my favorite hosts for modding, as you can pretty well do anything you want with them. I have plans for MT-G2 build, and want to convert one to 4 x 18650 in series just so I can say I did it. Perhaps I’ll become a Nage specialist the way OL is a MagLite specialist.

While its on my mind I just want to mention that I’m not sure if the lens is plastic or glass. When I removed it to take photos it felt kind of plasticky to me. I didn’t want to do a lot of tapping on it in case of scratches, but when I dropped it on some metal tools, it had made that glass clicky sound. I think I might have an AR coated lens here from CNQG that might fit.

Also you might have noticed a charging port beside the handle. I didn’t want to mention it, because it isn’t mentioned in any of the specs at any store. There is no charging circuitry on the driver, and it doesn’t come with any charging cables. I wouldn’t mess with it, it could lead to a big kaboom. I removed it entirely from my other build to make room for extra heat sinking. The plug that seals it has its own o-ring and I wouldn’t worry about water getting in. In a future build, I’d like to put a second switch there that would send it into DD or something. I’m going to have to send a close up to OldLumens to see if he knows what would fit there.

Seeing as the triple reflector isn’t really a reflector at all, but three small cups set into a rubber gasket. I was surprised at how bright this light is. A nice even flood. Actually, it is pretty close to my MT-G2 builds In its output and beam profile. I wonder how it will be after I stack a bunch of extra chips. But then again, I’m waiting on a 5 x XM-L reflector to see if it fits. Hell, I may have to order another host.

I guess I should mention the UI of this triple version. There are seven modes on this thing, non of them hidden. Kind of unique, or a pain, depending on how you feel about it.

Low=1xXML
Med.=2xXML
High=3xXML
Sequence=123-123-123
Med.=3xXML
Low =3xXML
SOS……………
I like the idea of the 1-2-3 emitter start up as it would be great to start in low with a single warm hi CRI emitter, then add a neutral on med. and a cool white on high.
The bad part is it seems to have next mode memory…well…kind of. I found if I count to ten, it will go back to the single emitter low mode.
The slow rotating sequence of 1-2-3 is kind of neat. Just to please those people that don’t like the one emitter at a time Low-Med-Hi, the added another Med and Low that utilize all three. The SOS is very slow and spaced out. At least there’s no strobe!

I shot this little YouTube video for your viewing pleasure. If it seems that the output is not that bright, it’s because bright light shining into it just turns my iPad screen into a white hole. To compensate, I had to shine two MT-G2 flashlights directly at the front of the Nage in order to obtain balance.

Tail current readings with 4 x UR18650FM
Low(1xXML) - 1.52A
Med(2xXML) - 3.05A
High(3xXML) - 4.60A
Sequence - 1.53 - 1.53 - 1.53A
Med(3xXML) - 1.35A
Low(3xXML) - 0.44A
SOS - … —…—

These are great hosts to work on. Happy modding.

Thanks for the review. I was just looking at these on ebay and I thought I might try one. I passed till the cost came down or the seller decided to send one for review, but ebay sellers usually just laugh when I say that to them.

Thanks for the review, Ouchy Foot. This FL’s insides look good.

Oh boy. I got another Nage in the mail today. This ones a single emitter model.

Nice gift box. My Ultrafire just came wrapped with bubble wrap.

I got a tail reading of 2.6A, so it is pretty much the same as the Ultrafire with 6 or 7 AMC7135 chips. That means there’s 11 or 12 empty spots to fill with chips.
I don’t know what I’ll do with this one. I may change the batteries to series and run an MT-G2 in it. I also have a 5 x XML reflector coming. Maybe it’ll fit.

The middle section on both of the Nages seem to be glued down. The Ultrafire had no glue.

That's a sharp set there. I really like the look of that model. Especially, the single emitter version. The charge port cap and chain seems like a gas cap.

Best of luck on the mods to come.

Here’s a video of the 1xXM-L Nage with factory 2.6A output, and the same torch after more AMC7135 chips were added to the empty spots on the driver. The modded one is pushing about 5.0A on copper.

NAGE: 2.6A vs 5.0A

Wow, what a difference. No dedome?

wow, that looks really good, and with that reflector size and 5 amps on copper, I wouldn't mind knowing the actual throw number :-)

No De-dome. This big reflector really concentrates the hotspot.

Actually, the Nage single emitter model seems to be a bit wimpier than its twin Ultrafire was out of the box. It has a hotspot, but the spill is almost indiscernible like it doesn’t exist. Even on a white wall. It’s like nothing is hitting the sides of the reflector. Everything is exactly the same, right down to the centering disc. I’ll have to take an emitter reading. Perhaps they installed fewer 7135 chips in the Nage, or I got a weak emitter. I’ll let you know when I pull the driver.

…or maybe I’m so used to my modded version, that I’ve forgotten how much improved it is.

Well, I didn’t have any plans tonight, so I pulled the driver. It was pretty much what I expected, but a bit of a surprise too.

I picked and poked, but the driver didn’t want to come out the back, so, let’s try the front. First I tried removing the big Aluminum MCPCB that the emitters are mounted on, but it’s glued down tight with no point of leverage without wrecking it. Okay, it can stay. I couldn’t believe how much problem I had desoldering the wires. I had my iron cranked to 850°F and it would not melt. I’ve desoldered copper on copper and never had any problem. I wonder if it’s some kind of new fangled Eco solder.

I drilled through the pill utilizing an unused hole already on the edge of the MPPCB.

So far so good. I don’t think I hurt anything. I inserted a hex key in the hole and felt around the edges to make sure I had a clear component free spot. I then took my trusty boat building mallet…and gave it a whack.

Success!

The driver was exactly what I expected since I had already pulled one from a single emitter version. It is made up of three sectors with spots for six AMC7135 in each sector. What surprised me was that they only populated four spots for each emitter, leaving two empty spots in each sector. It also bothered me that they pre-tinned the empty spots. I know from my last experience, that those chips are harder to install properly than one might think. I was hoping for untinned spots so I could try solder paste this time.
Let’s see…4 x 350 = 1.4A per emitter. Not great, but the stock light did seem nice and bright.
I guess I’ll be busy stacking chips. I’m not very good at that either, but there’s lots of room on this board compared to a 17mm driver.

The board basically runs the three emitters in parallel much like relic38s DIY Copper Tripple-Triple

These photos are borrowed from relic38s DIY build thread noted above. I hope he doesn’t mind.

Working on this board has given me a better understanding of relics diagrams. If I end up wrecking the board, I’ll just use his model and build a new one.

I would have mounted XM-L2 on copper, but that board is not coming out without destroying it, and I don’t even want to think about reflowing in situ. I have some 10mm sinkpads coming from Vestureofblood, which might fit around the edges. But there is a big recess underneath where a 22mm MBPCB would rest in the single XM-L version. I already know this from a previous build. I think they must have filled it in with some kind of goop. Everything would have to be built up for a level playing field, and I don’t think I’ll be worth the effort.

This is what the top of the pill looks like. (Pictured from another build)

So as you can see, the center depression must be filled with lots of adhesive. To remount with copper, I’d have to fill that center with a copper disc, then add another full sized plate to mount 10mm sinkpads close to the edge. I think the extra hight would end up interfering with the reflectors.

I guess I’ll pack this up until I’m in the mood to reflow those chips into the empty spots before I start stacking.

Advice Needed: Can I test one LED at a time without hurting the driver. I really don’t want to keep hooking up three emitters every time I want to test my chip stacking in each sector. Can I leave the — LED wires from the other two sectors unhooked while I test the sector I’m working on? Will this hurt the driver?

7135s are fine with being powered up with nothing attached to the outputs, go for it.

Clean off all the original solder with wick (and add some flux, makes it go easier). Then use your paste for mounting the new chips.

Thanks comfy. That makes things easier for testing. Maybe I’ll have a go at it this morning. I’m hoping I’ll be able to control this butane gun without melting anything besides paste.

Just keep the nozzle moving all the time, go around the chip in circles mostly, occasionally go over the top of the chip. Paste will melt before anything burns.

Just a little O/T: This is what I built for my AT-1226/aspheric thing that I'll probably never finish...

I removed everything but the 7135s from the original driver, and scarped off the mask from big unused areas of the ground plane to mount the little copper brackets that hold the Nanjg in place.

Hey. That’s what I was going to ask about next. I was wondering, if you could add chips buy connecting the legs of the AMC7135 chips from another board that has been stripped of everything else. To do something like that, does everything have to be removed, or just that big thingy with all the legs?

I survived my chip reflowing. I had a few false starts, and had to do several over a couple of times until I got the hang of it. It was a matter of learning the right amount of paste to use and how to utilize the butane torch.
The first few I had to test and double test with the DMM and hook up to an emitter to make sure I hadn’t killed the internals.
I did the last three all at the same time. At first I was nervous about applying too much heat and causing damage. I was having problems getting anything to happen. By the end, I had the system down. I just cranked that baby up and let it sit on the table til the nozzle was cherry red before I even bothered to try applying heat. Those chips just snapped to attention. Those little chips really don’t need as much babying as I initially thought they would.
All the chips seem to be functioning fine.
I tested all three sectors that now have six chips apiece and got readings of 2.27A — 2.28A and 2.31A.

This afternoon I’ll try stacking some more chips. I’m not sure how many I’ll add since I’ll be using the original board and emitters. Since everything is mounted on aluminum I don’t want to go with excessive current as it will be a waste due to heat sag.
What do you think would be the highest i should go and still retain efficiency in the emitters? If I add too many I won’t be gaining anything but heat.

8) You still have my attention. love ya work.

Just remove the diode and leave the rest, in case the driver ever gets recycled and needs to go back to stand-alone use. You can strip everything off, but it works exactly the same if you don't.

You do not need to connect any part of the slave board to BAT+. The 1226 driver in my case is the slave board, and is nothing but a holder for 10x 7135s. Whatever you use as the slave only needs GND, LED-, and PWM.

What's the UI on that Nage driver (and it uses a normal clicky switch in the tail, and not a momentary pushbutton like a SRK, right?)? If the UI sucks you could use a Nanjg on yours as well but with no 7135s on it at all, just used as a holder for the MCU, and the PWM output connected in parallel to each of the three banks of chips on the original driver. The one I added to mine was flashed with DrJones' luxdrv, the original driver was only high/low/strobe with next-mode memory... yuck. :Sp

The UI is detailed it the OP along with a short video demonstration.