I put some DC-FIX frosted plastic film on the inside surface of the glass. It works a treat. Of course this is not easily removed, so it’s only really suitable if you want it to be diffused all of the time. I do!
I did some measurements yesterday and it got very hot. By 1 hour is was dim, I forgot to measure after 35 minutes. After 35 minuttes it was down to 550 units from 780 at the start. Today I repeated the tests and this time it did not even get warm, and the regulation was much worse, in fact it was appalling:
5:10 = 939
5:15 = 745
5:20 = 695
5:25 = 6:50
5:30 = 605
5:35 = 562
5:40 = 623
5:45 = 487
5:50 = 43*
5:55 = 400
6:00 = 366
6:05 = 330
6:10 = 300
6:15 = 268
I then got bored ….
The units are arbitrary since I cannot accurately measure output, only brightness in lux.
So it looks as if the previous run on high has burnt something out in the circuitry or the LED. My feelings are that this light is best avoided. It has insufficient cooling, and poor regulation. I noticed from other reviews that Crelant lights seem to have poor regulation. What a shame.
Your results above seem to match my own fairly well, though I didn’t record details like you did. So, I don’t think you burned anything out. It’s odd that you said it didn’t get hot on this round, though. Did you measure or feel the heat during the first 20 or 30 minutes? I find that’s when it got the hottest, because that’s when it is brightest. As the maximum output drops, so does the heat.
Yes, it’s not regulated at all. It seems to be a simple direct-drive light, so depends entirely on what voltage your batteries can supply under load. By the time I ended my test (boredom), the high mode was down to about 100 lumens, but the lowest mode was down to moonlight-levels (instead of about 20 lumens when the batteries were fresh). So low is equally affected by low battery voltage.
I wouldn’t say to avoid this light. Just know what to expect. It has impressive throw for a 4xAA light when your batteries are fresh. But understand you’re only going to get that for a few minutes. After a few minutes, it’s still bright, and throws decent, but it’s no longer that impressive. This light is about a third the price of high-quality 4xAA lights (which are regulated), so don’t expect the same performance.
Yeah, I was hoping it would be regulated, but that was probably unrealistic.
I have been using mine for a couple of days and I still like it. Tint is very acceptable for a neutral white, seems right on the 3B that was specified.
Just a small PSA, you may want to redo the mcpcb fit and thermal paste. I took mine apart for inspection and the emitter wire holes have not been countersunk/de-burred on my sample, causing the mcpcb to not sit flush. Probably the reason for the excessive thermal paste.
Before:
After:
As discussed above the switch wires are somewhat tricky to manage when disassembling, but they are manageable with fine tweezers.
I also tried to lower the current, considering the poor thermal interface between the pill and head and to go easier on the eneloops. Changed the R056 to R100, but it is still pushing >3A. The driver seems to force the ground current through 300 degrees of the ground ring and then through a 0 ohm jumper, so possibly the R056 is only half of the equation. It also means my R100 will not survive long.
The driver is somewhat different from the typical ‘east-92’ style. Two parallel FETs driven by a mystery PIC-pinout MCU. But with a large (1 k) gate resistor, hinting it is operating the FETs at least partially in linear mode. There is some switching happening, but not full blown PWM. Looks like a driver Crelant has used in other lights as well. Should have looked a little closer before I put it back together.
Another light with this driver:
(Unfortunately that thread did not succeed in modifying the emitter current either.)
FWIW, I started a thread on the light in the appropriate forum so this group buy thread doesn’t get totally sidetracked with reviews and the like.
The following is quoted from a slaes page:
2. High efficient constant current circuit and output-luminance
3. Durable,custom-designed Microcontroller drive circuit
So presumably the above is not true then?
Well, (3) might be true. But (2), yeah, not so much. It’s not even a regulated output on low mode.
Well today I finally used the light for more than a few seconds & unfortunately the wonky driver reared its head again :confounded:
Within being on for under 5 minutes it started to sporadically blink until it ultimately just started to do that “slow strobe” mode again. I was hoping that I could just live with a very slightly wonky light as I didn’t have a positive experience with BG’s CS a few years ago (my one & only time in contacting their CS) but I had just submitted a ticket with them because this light is now too wonky for me to keep even if it was meant to serves as a non critical around the house kinda light.
I hope their CS has gotten better over the years b/c I recall my experience w/ them back then kinda sucked. Well, wish me luck!
Trying not derail this thread so here’s my last post on my light w/ the wonky driver. Here’s a short clip of that I set to BG’s CS. Please excuse the shoddy vid. It was made quick & dirty.
^
Says, “Please sign in to view this video.”
Has BG CS got back you yet?
I see this video is private…
Gearbest had the Nitecore EA45s for $36.59 with code so got that instead. Extra $10 but should be better light.
There wasn’t a neutral option though was there?
Don’t think so. Has the dedomed led.
Yeah, if you don’t mind cool white (I do), the Nitecore on sale sounds like a better deal.
Ooops!
Sorry guys. I must have made it private by accident.
Here it is again.
Initially they said that it goes into strobe when the batteries are low. But I tried more than 1 set of freshly charged eneloops w/ the same result & that’s what I initially told them.
However during the day I thought more about it & -> realized that as soon as I inserted fresh cells the light didn’t go into strobe (I think) BUT an hour or so later it goes into strobe when I go to turn it on. I don’t have my DDM to confirm but I am wondering if the driver in my light has an extraordinarily high level of parasite drain. & thus the cells are being heavily drained while being in the light in the off state. I haven’t had the chance to tell BG’s CS my hypothesis yet but I will. What do you guys think? THX!
EDIT - it seems that BG’s CS has improved over a few years ago. After viewing the video they have decided to send me a replacement. + my hypothesis seems incorrect. Yesterday I placed another set of freshly charged cells & from the get go the light started strobing on startup. Yep, my light was officially messed up. Thankfully BG is sending me a replacement. As the coast is clear I guess I can resume buying all these BG GBs then… 
Second attempt posting this because the devs have turned firefox for android into an unreliable pile of crap. I am impressed by this light. Though the ui is a bit simplistic, after about 1.5 hours on high, 3 hours on low, and tonight the output on high is still comparable to my L5 on medium using the same un re-charged 2700mah tenergy blue (L5 is running on a Lii 26650). It was originally intended as a host but due to its performance though lacking features it is vying for position as my preferred aa with the ea41w (thanks again 264). Looking forward to a driver offering more customizable levels. To the individuals who organized this buy I say ‘thank you!’.
I see this in the specs list.
Lens: Toughened ultra-clear glass. It means that it is coated both sides?
If it is not,can I put another coated?
Mine certainly does not have any anti-reflective coating. I’m not sure what they mean by “ultra-clear” glass. It’s transparent, sure, but that’s about it.
You could probably replace the glass with a lens that has anti-reflective coating. Might be a nice upgrade. I’m not sure where you’d order the lens, though. Looks like about a 40mm diameter.