Review and Test: Titanium AAA keychain flashlight

Thanks, MascaratumB and gchart.

The brass retaining rings are quite nice indeed. The only things that feel a bit cheap are the TIR lens with a cavity in the middle and the led on the MCPCB. But I need to say that most 10180 keychain flashlights have no glass lens in front of the TIR and the led mounted on a normal PCB.

Thanks for explaining, Agro. I tried to show the cavity (better than saying hole, sorry) in the TIR lens by putting a led in because it is hard to show in the video. On pictures it also looks like a flat TIR lens. The outer ring is clear and the inner cavity is slightly frosted.

patmurris, you are right, for that price the quality is not bad. It is not very bright, has only one mode and most likely I will swap the cold white XPG2 for a nicer led. The titanium body is designed and made nicely, though. It is cheaper than most titanium flashlights and of course I need an AAA titanium flashlight to have all materials in my collection.

Agro, I will add a few pictures and write my impressions down later. I know that this flashlight is more for collecting and a Sofirn C01S or a Lumintop Tool is more usable. That TIR lens could be changed for a similar sized one without cavity, the beam profile is not that good, either. Adding an o-ring at the lens might be a good idea as well.

I tried to remove the custom part of the link but the languages on Banggood always change like crazy. You can try to search for ‘keshun titanium’ on Banggood, it is the first flashlight that is shown.

It is great fun to reveal internal parts of a flashlight that haven’t been shown before.

My Lumintop Tool has a SW45K inside so I think the lumen values might be a little lower. The output is lower than both the Lumintop Tool and Sofirn C01S on high. It is hard to tell the actual lumen output because of the different beam profiles.

It depends on the AAA battery but I would estimate 50-80 lumens at start-up with full battery and maybe 30 lumens when it runs for longer. The output may drop also because of the bad cooling without a metal PCB and with a titanium body.

Waterproofing the switch somehow would be useful as well…

I might have missed it but is there noticeable PWM with this light?

Shouldn’t be any PWM. Based on the pics of the driver, it looks like a single mode boost driver similar to the PAM2803 (likely a knockoff of that driver chip). Output is set with a single resistor which is conveniently out in the open on the battery terminal side. With the assumed 95 mV feedback voltage and the R13 (0.13 Ohm) current sense resistor, it’s dialed in for 0.73 mA to the LED.

I think you mean 730mA or 0.73A? Seems far too high doesn’t it? Something like 73mA might sound right.

Well… you’re right. I meant 0.73 A, not mA.

Like I said, though… a PAM2803 sets amps according to ILED = 95mV / RSET. And an R13 resistor is supposed to be 0.13 Ohms, which would mean 730 mA / 0.73 amps. However, I think the resistor is using non-standard labelling (which I also observed in my Skilhunt E3A with a similar driver). I agree, 730 mA seems too much for a AAA light, especially one with a titanium body. So I think instead of a 0.13 Ohm resistor, it’s probably a 1.3 Ohm which means a much more reasonable 73 mA.

So I finally grabbed one of these. Mine came with a pocket clip instead of a keychain ring. I’m planning to attempt a gentle emitter swap soon. Mine has a 3535 emitter, not the XM-L2 claimed. That’s nicer, I can put something high CRI in there now.

Cute little thing. I don’t have much need for AAA lights but I have four AAA eneloops I never use…

They seem to be gone on Ali, but here is the MANUFACTURER

Ended up with sw45 219b in it. Gives me a good excuse to grab it. Now to find out where I stashed my Thrunite Ti5T… It’s on shelf queen duty, but I think the shelf is a drawer…

what is the UI?

medium and off?

It is one mode. So on/off. It’s a reverse clicky, if that matters.

Alright, so mine has a flicker.

I’ve managed to rule out the cell (freshly charged cells, multiple cells, etc.). I thought it was the tail switch but I bypassed that and still got flickering. It was happening before I reflowed a new LED, and also with each consecutive LED I reflowed.

So, the probably-knockoff boost driver appears to be of knockoff quality, would be my best guess.

This is a pretty simple board, though. Would it be difficult to design a replacement board to use a better boost driver? The output would still have to stay pretty low because of the heat path (or lack thereof).

Strange. Does the retaining ring make good contact with the driver? I also changed the led in mine but it does not flicker.

I would bet on that tail switch were it makes contact with the brass ring. Shine up the brass with find sand paper and then possibly add solder blobs to the switch ‘wing’ contacts.

Thanks to you both it was the driver retaining ring. Got better torque on it and got another probably 120° of rotation on it.-

I’m still interested in an imitation of this “driver” board - would be a neat way to make almost anything into a cheap AA (A) flashlight.

Edit: nevermind, played with it more and it still does it.

A uni-board LED + single mode AA/AAA boost driver? Honestly that’d be relatively simple and inexpensive. Same basic idea, and an even smaller board: Skilhunt E3A (AAA twisty) - First Look - #8 by gchart

I remember that! I had it confused with this driver in my mind, but knew I had seen a similar driver reflowed before.