Great work, but maybe worth investigating with the Li-ion cells is that they put out a lot of radio interference due to the circuitry. I'm not sure if that will matter for flashlights, but for people looking for an all-around replacement, that might be a deal breaker (for instance, you can't really use them in an emergency radio).
On that subject, here's some stuff to keep in mind.
One word y'all: NOISE. It's the elephant in the room that few want to address. It's a word that will make any experienced electronics professional cringe - especially me.
Here's a quote from a 'review' of some Xtars:
EDIT: I neglected to include acknowlegment of and thanx to 'sammyshp' for this quote:
************************************************************************************
"Another device I used for testing is a radio clock with external temperature sensors. At about 1.3V the display starts to fade, at 1.2V it is almost unreadable. And that's the normal...
Keep in mind that the RFI issues are both 'conducted' and 'radiated', and are not simply limited to affecting devices which use RF to communicate (as with the example above involving a 'radio-clock with external temperature sensors', but also digital devices which do not involve any external RF communication functions at all, (such as with the self-contained 'clock-thermometer' in the other example). Also, 'RF' isn't just about 'radios', but also A/V RF remotes, Bluetooth, Wi-fi, cellular, etc. In the digital realm, if the SNR (signal to noise ratio) gets too bad, it becomes more and more difficult to tell a one from a zero, and bit error rates go up (which you may or not be aware of / notice). If it gets bad enough, things shut down. Somewhere in between those 2 things one can go from being unaware of the problems / degraded operation, to going crazy trying to figure out the cause of the failure(s). It can be a nightmare scenario for the user or technician.
Xtar does document the issue with "radios", but without mention of other digital devices, and $100 says even this probably doesn't appear on the product packaging:
"...They are also not suitable for radios due to the noise caused by electromagnetic interference."
I also doubt it is possible to design a cell operating on these principles which does not have these issues, and would currently assume that all similar devices on the market suffer from the same deficiency, and apparently Xtar agrees with both those statements.:
"Can the electromagnetic wave interference be solved?
Not yet resolved. Because the rechargeable 1.5V lithium battery has a built-in charging/discharging and voltage step-down circuit. It will cause electromagnetic effect from the inductance in the circuit. All the 1.5V lithium batteries on the market have such characteristic."
Both of those XTAR quotes are found here:
The frequent questions and answers to the XTAR 4150mWh AA Li-ion battery. Is it safe, which devices is it suitable for, etc.
www.xtar.cc
Further, in theory: Also keep in mind that while these cells produce a potential interference 'signal' at a specific frequency, their output waveform is a square wave, as seen in the time domain with an oscilloscope. A square wave is essentially the sum of a single sine wave signal at a specific 'fundamental' frequency, plus all of the odd harmonics (3,5,7...) of that frequency, and when viewed in the frequency domain with a spectrum analyzer, those signals would be displayed as secondary signals appearing at frequencies which are odd multiples of the fundamental frequency, albeit at decreasing amplitude as they get further from that fundamental frequency. So rather than a single offending frequency, this results in a number of potentially interfering signals across a range of the spectrum. Simplified, that looks something like this:
I don't currently have access to either a scope or a spec-A, so I can't do measurements of these signals on samples, or do the type of testing that's called for with these cells, and apparently most of those doing cell testing know very little of such things.
In my opinion / speculation: These things should be subject to a requirement for FCC certification under FCC CFR Title 47, Part 15C, which applies to 'Unintentional Radiators'). The problem is that these are not just 'cells', but are switching power supplies with an internal power source. Most lithium battery assemblies are tested as part of the cert testing of the equipment they are used / installed in, and these separate, standalone cells have never existed until now(?), and thus they likely fall through a giant loophole that allows them to be sold as a standalone product without FCC certification. I'm not an attorney who works with FCC regs., so I'm just speculating about what has created this problem / allowed it to exist. Clearly these things should be subject to testing / FCC certification IMO. Currently, it appears to me they fall into a loophole which should be closed, as some are taking advantage of this. I might call it 'clever', but that would be far too kind;-)
I'm not going to carry on and bore people to tears, but if you have a question about anything in this post or the one I linked to, let me know and I'll try to clarify.
cc:
@DaveTheDude (I owe him some info;-)