While I agree those are crap I can’t ALL **fire are junk. I have a set of protected Trust fire flames 26650 that are rather good. Yes it is a crap shoot and nearly all are junk but from the correct sellers and with research you can find some that IMHO are worth the risk.
There are a lot of resons, not just one. You mention some. And the bias is self reinforcing. 18650 dominate, so many devices use them to benefit from cheaper batteries with higher energy density. Many devices drive demand. The volume makes development pay for itself if it helps capture a larger market. More development means higher energy density. Higher volume makes cheaper batteries and/or higher profits. This makes the 18650 dominate the market. And round and round...
If you’d like to compare amp/volt graphs from 26650s and 18650s you may consider dampfakkus.de - pretty self explaining. They have a good selection of high amp 18650 and 26650 with 2A 3A and 5A discharge graphs. It just shows the huge benefits you get from 26650 in single cell aplications.
Interesting conversation. I like both 18650 and 26650 cells. As mentioned 26650 can hold higher voltage under load. Great for high power single cell applications. Though, for power to size ratio 18650 is king.
A123 are not so good for flashlights. They were very good tech when they came out and had (and still have) kick ass output. Though voltage (3.6v) and runtime (2300mah) are at the low end of the scale. You kinda need to run them in a 2s configuration because of the lower voltage. Even for direct drive on an MT-G2 you get more output from regular 26650 because they are higher voltage.
I’d like to know why the tesla (and other stuff that uses shite loads of batteries) didn’t end up with lipo cells? They are the highest energy density cell and come in prismatic shape so less space would be wasted. I know you can get prismatic l-ion but the output is poor (last time I looked).
Maybe part of the reason 18650 got such a good hold of the market is because the revolution of led making everything smaller, 26650 seemed to be the wrong direction.
Math says no, but you don’t even have to do math. Get some pennies. Pretend they are the tops of 18650s. Lay them out as efficiently as you like. Now pretend they are 26650s. No difference, the gap area scale with the cells.
Someone mentioned the contribution of the can thickness, but that’s not all that significant, really.
Agreed. My guess is that the 18650 originally had the best mix of characteristics for the mid-range of the laptop market. The diameter allowed a reasonable ratio of guts to skin and worked well with the overall thickness of other components and the length was about right for 3-4 end to end, a number laid side by side, or various other configurations. The market volume drove competition among suppliers, who invested in automation to lower costs, and other improvements. They were able to utilize that equipment to produce cells with different characteristics (power vs energy density vs durability, etc). Moreover, the form factor was also flexible enough for other applications, like tool packs, etc.
These days, I think the laptop market isn’t as important to the 18650, and vice versa. The volume of smartphones is crazy, and probably on its own, enough to drive down the premium formerly commanded by pouch cells. Tablets, and a shift to thinner laptops have pushed things even further. The 18650 continues on though. Right now Tesla alone is buying an huge number. Whether or not they switch to another form factor in the future remains to be seen, but the 18650 still has a lot to recommend it for their application. The ratio of surface to volume is better than with thicker cells, which is important for temperature control. Longer cells would have the same advantage, but they’d make for a taller battery pack, which doesn’t really make sense given its position under the car.
Another other issue with 22650 is that because of their lower volumes, the incentive to make multiple variants (high energy vs high power, for example) is smaller.
Yes the gap area scales with the cells, but that does mean that as the cells get bigger the gaps get bigger. If the goal is to have the highest energy density in a multiple cell pack with fixed external dimensions, you want the smaller gaps
……………………
A better test would be pretend the pennies are 18650’s as you said, but then lay next to them some quarters representing 26650’s. Which gap is bigger?
The RATIO of cell space to total space is constant, regardless of cell size.
That made me think of pi, and now I want pie.
Yes, the gap at the margins when fitting batteries into a space of fixed dimensions will vary depending on cell size, with the smaller cells being more efficient since less space will be wasted. Within a unit cell though, the percentage of filled vs empty space is the same.
Or, take a 234x234mm space that would accommodate an integer multiple of either cell diameter and arrange them in a simple grid for ease of calculation. That works out to 9*9 26650s or 13*13 18650s.
The total area is 54,756mm^2.
The area occupied by the 18650s is (13*13)3.14159(18mm/2)^2 = 43,005.22551
The area occupied by the 26650s is: (9*9)3.14159(26mm/2)^2 = 43,005.22551
Packing on a simple grid isn’t always the most efficient scheme, but the results are basically the same. The differences in density between cell sizes are going to come down to any wasted space at the edges, not in between cells.
I am surprised that you saw the same, or even within 10% of the same, direct-drive current with an XM-L2 as you did an XM-L. I suspect that there was a lot of extra resistance in this test setup. 3.18A is very low for a MKE 26650 direct drive to an XM-L.
It was tested at the tail of each light. Ideally they would have been tested without a host at all. I am also using a crappy DMM until my new one arrives. I figured it was fair since all batteries used the same setup. But maybe the DMM was limiting my higher end currents.
How much more current pull do you normally see on a XM-L2 over XM-L?
If the gap size when packing, and the cell size as single are both almost equally efficient between 18650 and 22650, or in other words, there is an almost 1:1 ratio of size to capacity, than why aren’t 22650s used when the 18650’s don’t offer enough capacity? Why don’t they make more 22650 flashlights?
Aren’t they more efficient than stacking 18650’s when the double voltage is not needed?
As I stated earlier it is due do the fact that 18650’s were the original laptop batteries and industry doesn’t feel like changing over.
Your DMM leads are most likely the culprit here.
Usually an XM-L will pull 7A-7.5A from a hot cell; the older XM-L2 were usually 6A-6.5A; the newer XM-L2 are usually around 5A. If I didn't see at least 6A from that MKE 26650 into an XM-L I would be surprised.
I compared my EVVA 26650 batteries to 3 of my better (Panasonic/Xtar/Keeppower) 18650 batteries, and carefully recorded the results.
No **fires or any junk batts were used.
The results were consistent. The 26650’s fared better.
While I do agree that the Ultra**** I have seen have been crap, actually crap has more value to me, but all the Trustfire flames that I have, 22 in total, have been excellent in capacity and mAh/$$.....especially the last dozen 26650's I have bought for my J18's and others. ALL of these TF's have tested out over the promised 5000mAh and for less than $8/ea, look to be about the best value out there in any Li-Ion battery.
The 18650s have no unique advantage over the 26650s. It comes down to form factor. If you can handle the larger size than you should use the 22650s.
Maybe I’m picky but it’s 2*6*650 not 22650. 
26650 I stand corrected.
It’s not surprising at all. Going the other way, you can only get 2-3A out of smaller cells like 18350/14500 or less with even smaller ones like 10440.