Actually, I would have to disagree with you there (...) If I am wrong in my thinking then I would like to know from someone with a greater knowledge base of battery performance who can explain why this reasoning would be incorrect
Damn! I knew I shouldn't have taken such a shortcut and made a two-lines post. Apparently you misunderstood my point, and no, your reasoning doesn't seem to be incorrect. I think I mostly agree with what you said but with a few caveats:
- These sites are fine but often have few data points, depending on which one you read you can find the 25R slightly better or slightly worse than the HE4 (the truth is, they are pretty close in terms of performance)
- On Dampfakkus.de (as well as another site I forgot) the 25R is tested superior to the HE4 (see http://dampfakkus.de/akkutest.php?id=569 @ 5A) and this kinda contradicts the reports here
- That site has no datapoint for the newer 25R (green), only for the older (blue)
- It's also lacking some of the newer 30A cells of possible interest (ex: HB4/HB6)
- The HE2 and HE4 are indeed different and one should prefer the newer HE4 but I group them as they have the same specs on the label and the differences are small in practice
- There are so many fake cells that you never know if even reputable testers sometimes don't get B-grade cells repackaged to look like A-grade from the real manufacturer
Those sites are nice but they can be a little daunting for the non scientifically minded folks (and I think you lost some of our readers with your post heh!) I tend to favor simple and to the point graphical material, so I will repost Mooch's infographic that
@TheDudeNextDoor posted in the 18650 safety thread:
As of today I fully agree with his recommendations: all 18650 cells listed here are compatible with the Zi/Mi excepted the Panasonic/Sanyo and the Samsung 30Q (*) If you get anything else it will be *at best* equal to one of these cells but in most cases it will be of lower quality, and in nearly all cases it will not be what's stated on the (rewraped) label.
(*) The 30Q is indeed borderline and will illustrate our first misunderstanding well: it should work in practice, not only because it's under-rated but because the amp draw on the Mi isn't that high, RBT used a good safety margin to help cope with cell performance degradation over time. As the cell ages its internal resistance will rise, lowering its discharge rate (C rating) So if you already are operating at the listed continuous discharge with a new cell, over time you will exceed what the cell can safely output and enter the red zone. The closer to the limit you operate the more strain you put on the cell and it will sag more and more, it's a vicious cycle.
I don't recommend trying to get larger capacity cells as the results will be mixed: it will stay in the usable range for longer, so you will get somewhat more run time per charge. But the larger the capacity the lower the C rating and the higher the internal resistance. The cell will sag way more under load and you will spend all that usable range at a lower voltage. You'll get more seconds of use but they will always be a lower power, the device will last longer but will be weaker during that time.
So my initial post was not about using higher capacity / lower discharge rate cells, it was about the inverse. I was thinking about the HB2/HB4/HB6/VTC3/VTC4 on the picture above. All these cells have a higher continuous discharge than the currently recommended 25R/HE2/HE4 at the cost of a shorter run time. Under a given load they will stay a shorter time in the usable range of interest but they will indeed have less sag (lower internal resistance) and thus the device will operate at a higher power for less time.
Those will works yes, and they might give more ooomph, but it's a matter of trade-offs and my remark was more: we don't need more than 20A continuous discharge as it's already overkill and includes a good safety margin. It's overkill for the Zi, and sorry I don't have a Mi to give you the figures you requested but I know it should draw even less amps and it operates at a lower voltage (the Zi is regulated around 5V)
To illustrate that point have a look at the HB2 5A graph here:
http://dampfakkus.de/akkutest.php?id=375 and note how the HB2 reaches the 3.6V point more than twice as fast as the reference 25R (btw the HB2 is only listed as 22A on this site for some reason....)
On this one we can see the HE4 performing marginally worse than the 25R:
http://dampfakkus.de/akkutest.php?id=569
Now a few controversial data points, here we can see the HE4 performing better than the VTC5 that you considered, under a heavy load (20A). Note how all these cells almost instantly reach 3.6V (*under load*, not the voltage reading you get when you remove them)
http://dampfakkus.de/highamps/569-at 20A.png
Here we have a comparison of the HE2 and HE4 under same heavy load and they peform practically the same:
http://dampfakkus.de/highamps/569-vs HE2 at 20.png
Here we see the HE2 performing better than the 25R @ 5A :
http://dampfakkus.de/akkutest.php?id=507 when the HE4 was slightly worse above. Tbh I think these 3 cells perform almost the same.
This one illustrates the point you were making about getting higher voltage through the usable range, where we see the 20R performing better for the first part of the curve compared to the 25R but then ultimately reaching the 3.6V point faster due to the lower capacity:
http://dampfakkus.de/akkutest.php?id=411
Then this graph is interesting as it shows the HE2 performing better than the VTC5 at first under a pulsed heavy load (closer to the Mi reality) but then the VTC5 lower internal resistance translates pretty fast to less sag and more performance:
http://dampfakkus.de/highamps/523-vs lg+samsung.png
Finally I ran a little comparison using his tool and the better cell for his data set seems to be what he calls VTC5A, the normal VTC5 being about the same as the HE2/25R:
http://dampfakkus.de/akkuvergleich.php?akku1=569&akku2=610&akku3=490&akku4=523&akku5=&akku6=&a=5
To conclude, I think the 25R is the cell being the most counterfeit. At some point Sony halted production so most VTC's around were also fake. Now provided you get genuine ones, yes the VTC5 are surely a good bet, or any of the 30A cells if you want more power at the expense of run time. I've been using VapPower 2500mAh 35A cells with my Zion and they peform as good as the HE4 I got afterwards (I think VapPower are rebadged HE2's and clearly only 20A continuous, but the HE2 is also slightly under-rated and closer to 25A in some tests...)