Who’s the madman getting 10?!?
I would like to add 1 lantern, to make 2 total (#1372)
He might regret it if the v2 is even better ,better keep another $500+ spare :money_mouth_face:
A person that has a big mansion and wants a lantern for each room in case of a black out.
Since the names and quantities are easily gotten from the Interest list by user names.xlsx file linked in post #1 I can tell you it’s osb40000.
Better hurry and sign up for more if you want to beat him out 
Yes, and the lantern already has one. It’s the lighted button. The expected runtime per charge is over a decade. Basically, the power draw is lower than the cells’ self-discharge rate.
Not really. It has one power channel for warm, and one for cool, using PWM to control brightness. The efficiency shouldn’t change much at all at different brightness levels. However, the cooler tint emitters produce more lumens per Watt than the warmer ones, so if you want to maximize efficiency, use the coldest tint it can make.
No, I measured that it gets dimmer at 4000K… and I artificially increased it to keep the output relatively flat as tint changes.
The output is controlled on two axes — brightness and tint. The brightness is a value from 0 to 100%, and the tint is a value from 3000K (or 2700K) to 5000K. At either end, that ends up being 100x3000K + 0x5000K, or 0x3000K + 100x5000K. And in the middle, it would be 50x3000K + 50x5000K.
But this made it quite a bit less bright in the middle. So it increases the total brightness by as much as 150% in the middle, to compensate. That means, instead of 50+50, it actually runs the middle tint at 75+75 PWM. The overall power usage is still about the same as 100+0 or 0+100 though.
Internally, the actual PWM curves look like this:
… and the before/after for this correction looks like this. Before, it sagged in the middle… and now it’s not totally flat but it’s at least reasonably close.
Yes, actually. I adjusted candle mode specifically for this light… and also have been putting it on a table full of actual candles for comparison. It’s not perfect, but it’s not bad either. The overall brightness pattern is random, but it looks like this on a graph:
I also gave it a mode where it can change tint automatically with brightness, which affects all other modes. When used in candle mode, it should give at least a subtle coloration effect, but it’s not very noticeable.
Basically, at each end of the tint ramp, there’s an “auto” mode. One is warm when dim and cool when bright, and the other is cool when dim and warm when bright.
I think the auto-tint thing looks best when using lightning storm mode.
I am on the list for two, and would like to add 2 for a total of 4 please!
Please count me in for one. Thanks!
I think leftdisconnected is correct about the protected cells not being a safeguard.
Regarding the prior question, I don’t know the TP5100’s algorithm, but a typical lithium ion charge function would transition from constant current to constant voltage charging once the charge voltage (which is higher than the rest voltage) reaches 4.2V.
At that point, the current the battery draws from the charger starts to slowly decline. Once it falls below the threshold, the charger terminates, and does not restart until either the battery is removed and reinserted, or the voltage falls below an initiate threshold. One datasheet I found for the TP5100 (mostly in Chinese), seems to indicate the termination current is 80mA, and the restart voltage is 0.080V below the charge voltage.
As you speculate, that would mean the LED’s drawing continuous power would cause the charger to stay on indefinitely if left plugged in with the LED’s on. However, the charge current absorbed by the batteries should slowly decline, asymptotically towards 0.
It would likely accelerate battery capacity loss, but I wouldn’t expect it to actually overcharge the cells and create a hazard, as the voltage should stay at 4.2V.
I could probably test that using a single cell, (one that is well used i have) by connecting the USB charger, turn the lantern on with the cell connected to the lantern as it would normally be, but with a volt & amp meter connected to the cell to see what happen on the meters over time while the lantern is running when the cell reaches its full 4.2 volt charge, (say with the lantern running on a medium mode setting.
Honestly, this lantern sounds like it will be really helpful in nations that have limited electricity, where there are sometimes daily brownouts and blackouts. If I could afford to, I would buy one (plus batteries) for each of my family’s households that are overseas.
Great point! I guess we sometimes forget the luxury a lot of us live in. Thanks for reminder I should be much more grateful.
That’s most excellent TK! Thanks!
This lantern is breaking new ground in so many ways! :+1:
Interested for 2
I’m in on 1!
First BLF group buy, already psyched! 
Friends and family need these, I’ll get some V2 too, it’s only money.
•
•
BLF’s mantra 
Just bumping up here quickly, after some time AFK. Is the V2 a real thing, or still a fantasy?
Please add me for 1 more lantern for a total of 2.
Thank you!