Samsung 21700 50S unit price has decreased significantly

Samsung 21700 5000mah 50s 20A-25A CDR :+1:

Last year, the price rose rapidly. This year, the new date is 2022, and the production of batteries has decreased

I’m confused, how has the price gone down, if the production has decreased?

It used to be a Korean product, but now it is produced in Tianjin, China

At present, some goods are on sale. It is expected that there will be distributors around the world soon

Here it is on 18650batterystore.com

(I got an advertising email from them today.)

https://www.18650batterystore.com/collections/21700-batteries/products/samsung-50s-21700-5000mah-25a-battery

5000mAh and 20A-25A is pretty impressive!

:+1:

Despite the good specs, this battery has poor cycle life, with capacity dropping rapidly after 200 cycles:

I have been wondering about that since this cell first popped up on the market. Modern Li-ion cells are pretty mature tech so there’s always got to be a compromise, whether it be peak current, capacity or cycle life. This cell has great capacity and peak current, so that leaves cycle life.

This is no surprise, a 5000mAh cell claiming 20-25Ah discharge does not seem realistic, something would have to give and thats probably cycle life.

But how does cycle life improve when not subjected to such rigorous discharge? All that heat takes a toll. If used like most of us might, say 8amps or even 15amps and removed for recharging at 3V…? Datasheet suggests standard life, but falling short quickly with 60% original capacity after just a couple-few hundred cycles when stressed like this. That’s worst case scenario of course (I think it was as little as 100 cycles for that 60% when pushed at higher amps).

Doesn’t seem like a 40T replacement but a nice upgrade from 50E that can be used in a wider variety of flashlights while still enjoying higher capacity and likely run times for most. Hopefully someone will do standard current draw tests on these soon.

Funny thing is that the main reason is not capacity or discharge rate.

It is silicon anode concentration and its properties that are likely causing problems at full charge voltages.

Remember, silicon tends to expand at high charge voltages. Unless you use an exotic expensive nano silicon arrangement, nanowires, or something like Tesla’s innovation to massively increase silicon concentration and therefore energy and power density, you have to give in cycle life.

I would not be surprised if this was a 4700-4800mAh cell instead of 5000mAh, but they gave it an extra boost in silicon, trading cycle life for capacity.

TLDR: If you do not use an innovative techniques to have more silicon in the anode, cycle life will suffer.

In the link I posted above, there is a test at 1C and 90% DoD. It is currently at 400 cycles and presumably still ongoing. Also interesting to note that this cell gets much hotter than other cells at 1C, 3C and 5C, despite internal resistance not being higher than average.

What’s the hive mind thinking on Mooch’s E-Scores? In my limited comparisons it at least doesn’t contradict what I read here or from HKJ’s tests.

https://www.e-cigarette-forum.com/attachments/4bd89b17-2185-42f0-b943-21e6131f9241-jpeg.974739/

Always seems solid…can’t recall one where their respective results were all that different. Mooch focuses more on high drain applications and HKJ is a flashlight and electronics enthusiast…often I get more useful information for myself from HKJ’s graphs and current selections but both of their tests are very valuable and transparent. Thunderheart and the testing linked in this thread also have value but some different approaches. I think Mooch has done the lithium cell communities an incredible service over the years (and he’s weathered the resulting heat from manufacturers and highly opinionated or less-informed users very gracefully and professionally…).

Thanks, Azirine…I missed that link. Helpful and insightful to see all the rest of the data. I’m not sure that testing is quite as useful to us for flashlights but it’s data and observation nonetheless. I noted that he force cools the cells during testing…does show us one thing but then doesn’t show us another thing that could be more applicable. And maybe not so much at the lower currents but that forced cooling could skew results somewhat if the exterior of the cells is appreciably cooler than the core (imbalance of resistance and how that affects the pixie flow). Mooch doesn’t cool, so his results show higher temps.

He tested this cell last year, said in another thread that he hopes to test them again with the newly available cells. May not be any different from someone like Samsung but it could be…we’ve seen that a lot over the years, usually little changes (for better or worse).

His test at the e-cig forum: https://www.e-cigarette-forum.com/threads/bench-test-results-samsung-50s-25a-5000mah-21700…a-great-performer-but-ran-a-little-hot.966587/

In Pajda’s results for 1C (5 amp) the 400 cycle capacity chart shows about the same curve as the other cells. He doesn’t have the 3C (15 amp) results there as he does for some other metrics, which would be nice to see. I wonder what an equal current test(s) for all the cells would show. He’s doing the industry standard C-rates but as flashlight users it might be helpful to have just straight current comparisons (Mooch and HKJ do give us that at equal levels). Comparing and chewing on the data between Pajda and Mooch, it seems to me that this 50S is an incremental improvement over the 50E(3)…enough to make it worthwhile and for typical flashlight amps will put it a notch above the 40T/P42A maybe…in run time anyway, and looks like maybe it won’t suffer cycle life like these high drain torture tests are showing. I think I’ll pick up a couple if I can find them a bit cheaper than 18650 store has them right now. If nothing else, it looks like the 50S can hold its voltage a bit better than the 50E(3) for most of its curve, and that’s a general benefit but for UV aficionados without a boost driver that could be a nice bump.

Also thought it was interesting that his 1C IR graph projects a downward less-resistance curve, but maybe that’s just because of the few initial plot points/slight variation.

Another thought I had was that by running a bit warmer overall (it’s not so bad at the low currents really) it might boost cold weather performance a little.

Interesting cell. Glad to have all the data these guys spend hours of their time to provide for us!

I found this chart helpful (for lights)…additional stuff over the graph curves:

And the 1C / 5Amp 400 cycle graph. Keep in mind that here it is comparing 5Amp to 4Amp or thereabouts, and in the 5C graph first posted, it’s 25A to 20A. Hence why equal current testing might be more helpful.