I’ll lay down the challenge to solve the heat issue. Miniature water cooling set up on the led.
compared to my convoy s2 thats not very good, it can get that temp in well under 5 mins
I’m kind of wondering if they just expect the trapped air inside the head to transmit heat
[quote=Bort]
While the S2 probably has better thermal contact, it also has a lot less thermal mass, which would allow it to heat up faster. It may be driven harder too, not sure though.
its 2.8Amps, likely a few less otf lumens, i’ve been thinking about the thermal mass, i can measure then calculate how much there is in the s2, but don’t have any data for the ea4
thanks for that Relic38
What was your ambient temperature? About 20C? Was the light tail standing were you holding it?
We are just moving into Autumn but summer nights can hover over 30C (86F) here. It seems your temps were still rising about 2C per minute when you ended your test - I don't suppose you could let it run longer and see if they stabilise?
Just askin'
Considering the max heat rating for a Cree LED, I think they have correctly figured out that's it "good enough". Certainly it's obvious that it works and it works in many, many lights made by several Mfgs. I will still always go for larger heat sinks and always make sure there's as much surface contact to the body/head as possible, just to feel safer about what I do, but this sort of disproves the fears about overheating a high powered led, by having too little of a heat sink. If we were to look at (as suggested in this thread), the surface area of the die compared to the surface area of that flimsy heat sink and where it touches the body, then it is "good enough". That coupled with the fact that most users (majority), will never have it on hight long enough to ever do any damage and it looks like although I don't approve, I should not knock them for what they did.
That said, I still think it stinks, LOL.
I usually stop around 10 minutes, as that gives an indication of the internal thermal coupling. My ambient temperature is around 19C in this room. The light is tailstanding, I only touch it to reactivate Turbo mode. When (if) I run the test again, I’ll go to 15 minutes and capture lumen output as well.
its more complicated then that, i read a post about a guy who de-domed his led by pulling the star off the pill and running it for a few mins till it was smoking, so that suggests there is more going in in your ea8 then meets the eye, though i can’t explain how
Nitecore flashlights are criticized in China shoudian forum for their heatsink design either. Someone open his EA4, and the design is similar with EA8 OL has shown us. They all come with a patented heatsink disc.
It is to save cost that they use aluminium tubes instead of aluminium block. Using aluminium block, it will be one piece flashlight head and body acting as heatsink with direct contact. I guess that is why EA4 and EA8 price are so competitive compare to Olight or Fenix.
Even Nitecore is using tubes, they should design it better with more contact with flashlight body. Not that Nitecore flashlights are having heatsink problem, just that most want better heatsink solution and probably can be mod to drive LED harder.
Interesting heat sinking system, they concentrated more on making a complex switch.
Could you please provide a link? I would be curious to read about it (courtesy of Google Translate. :) )
subscribed
Old-lumens, thanks so very much for documenting the tear down of your EA8. Very surprising results. I had expected more from NiteCore. Such a small degree of heat sinking… because they’ve engineered the driver to be so efficient that more is not necessary? Or because they’ve made a trade-off to keep the price low and sacrifice longevity of the internal components? (electronics and high heat don’t go well together). I’m really suspicious of how long these lights will last, including the EA4. I was going to buy another, plus an EA8… but now, I think I’ll hold off.
I would really like to see a long term heat exposure test to the EA4/EA8… see how well the components hold up. But then, maybe the component decay will be harder to test. Maybe instead of the light lasting 10 years, it’ll last 5 years… still well beyond the warranty period and enough time for people to forget about these models in favor of more advanced ones that succeed them.
@ xevious
Good to see you on this forum, I wanted to get the word out on CPF that’s why I was very caution the way I posted about it, so that it doesn’t get easily censored, as they have a habit of doing that on CPF; not the case on BLF 
I'm now very glad you did this, Justin! I will admit my original opinion was wrong. ;) It obviously does work for what they intended but it could have been so much better. I never considered getting an EA8 because of the pile of cells needed to run it but I was considering an EA4. I don't think I will now.
I had a clarification question for Old-Lumens in regard of this photo. You said that the 2 mm thick heat sink board is sitting on a plastic spacer in the photo above instead of the metal shell which would constitute the heat-sink design, but in a later post you said: "I could see it, if the lip inside the body, below the 'heat sink' had been thicker, where the 'heat sink' actually sat on the shelf. I could also see making the 'heat sink' big enough to 'really touch' the body at the heat sink's OD, but they didn't score on either one. What their style suggests is that either..."
So my question is, us the turquoise arrow pointing to the plastic spacer, and the red arrow, the actual metal body which would be the same as the green arrow on the outside? I would personally try to open my own EA8 to confirm but I don't have a heat gun readily available and it won't budge with the strap wrench I'm using.
The main issue, besides not enough contact area, is Nitecore is not using 2 screws to firmly secure base plate onto the heatsink disc. It seems to be using reflector to press base plate on the disc.
this is quite common and not a big deal, especially on this one since the whole thing was held shut with loctite
I am at work and it’s hard to see anything but a small thumbnail of the photo, but I will try to explain. In one of my photos you can see the bare body tube and you can see a thin lip, where the tube has been opened up to a cretain depth. The stock heat sink does touch that lip “somewhat”. The stock heat sink also touches the ID of the body “somewhat”. Unfortunately it seems that from concept to manufacturing, something got missed, as the area of the tube that was “opened up”, was not done correctly in some of the lights (off center), which means a “poor fit” and contact is not good. The plastic under the heatsink, is the plastic housing that holds the driver and switch. The heat sink sits on top of that plastic housing. The plastic housing is hollow, so the heat sink only sits on the outer edge of the plastic housing. The clearances were done so that if made right, the heat sink would sit on the plastic housing and sit on the lip inside the body, so that the heat sink would be touching the body. The reflector does tend to hold everything in place, when the bezel is tight, but the whole “pill assembly” (driver, plastic housing and heat sink) are a “little bit” of a press fit, not loose enough to fall out, but not so tight that the assembly can’t be pushed out easily.
In your photo, it looks like the Red arrow is touching the plastic housing and the Turquoise arrow is pointing to the lip in the body, if that is what you want to know.
I hope this helps explain things better.