Hi:
I have been working a bit with NiMH AA cells lately in powering various gadgets such as LED flashlights, HID flashlights, and a photoflash unit.
I have been using the highly regarded Sanyo and Energizer 2500mAh cells. I use a LaCrosse BC-900 to pre-condition and charge the cells.
The Energizer and Sanyo cells seem to be fairly well matched when conditioned well, with a spread of perhaps 2.4 to 2.6 Ah.
However, when used in series packs of 4 to 8 cells, always it is the case that one or two cells collapse before the other cells are depleted. They never all go down together. Thus, the evidence of cell depletion as perceived in the operation of the device is very ambiguous, making it hard to tell when to turn it off to avoid cell damage.
Of course I have no way to be sure if a cell has experienced reverse voltage while installed in most devices. If I have the chance, I quickly remove the cells and measure with a DMM when the device appears to be running low on power. What I find is one or two cells with a very low voltage of 0.9V or so, rapidly rising. Thus, I suspect that these cells likely had reversed voltage when current was flowing, and are now recovering their positive voltage due to material diffusion.
Inserting the cells in the BC-900 on discharge mode reveals that sure enough the low cells have effectively zero capacity and the other cells might have 5-10% or so of remaining capacity. Just enough to keep the gadget going long enough to reverse the voltage on the weaker cells for a considerable time.
Now I am fortunate enough to have test instruments, know a little about the need to quickly turn off a gadget once declining power begins to be apparent, and I have one of the most sophisticated non-laboratory grade battery chargers on the market (which I think is still rather deficient in many quality and feature aspects).
The best things about the BC-900 is it allows one to condition new cells, and measure their capacities. I would never be satisfied with a charger that just said "has a conditioning cycle with 'done' indicator LED" because one really needs a quantitative measure to be certain the cells have been conditioned. Even the BC-900 typically fails to fully condition on the first run. A second run usually does the trick. This can take 2-3 days.
New cells take quite a few cycles to approach full capacity. This coupled with the cell voltage reversal problem leads me to suspect that the average consumer might have mostly disappointing experiences with NiMH cells sold in stores. Especially since the chargers sold with them don't provide diagnostic info nor have the ability to condition cells with multiple cycles before first use. Also, the capacities are VERY poorly matched on the first few cycles. Finally, unless they buy very good cells, many cheap cells out their never have well balanced capacities even if one attempts to condition them. (The batteries that come with the charger are a case in point--chuck them and buy Energizers.) So if the new cells are used in devices rather than first in a conditioning cycle in a charger (with *independent* cell charging channels) then the cells are likely to experience the most prolonged and damaging voltage reversals in the first few uses!
I have noticed that the rate of self-discharge is very slightly faster on the cells which I suspect have reversed. Though, they only reversed for a very brief time. If consumers leave their gadgets running until they basically "don't go anymore" ie, deep discharge the packs, then they likely are damaging their cells quite a bit, leading to a continually degenerating performance of the pack.
My experiences with NiMH cells have convinced me that they take a lot of care and effort to use effectively. Ideally a device engineered to use them will shut down before cells collapse. But this is very difficult to ensure unless circuitry can actually monitor every cell independently. Of course, no designer would spare the expense of such a scheme. Thus I doubt that consumers are ever realizing the full potential of these cells, which truly have remarkable capacities, and the current to make high-drain devices perform very well. But without the proper care which I suspect most folks never give these cells, they are likely to seem like junk, with very poor lifetimes and performance in gadgets.
Any thoughts?
P.S. I have also recently tried rechargable Li-ion CR123 shaped cells. Just a pair of them in an LED flashlight. Sure enough one cell collapses before the other and now that cell has slightly higher self-discharge rate than the other.
It seems cells need to have a diode built in to limit reverse voltage to a non-damaging or at least a very minimally damaging level (maybe a Shottky).