Hi, thanks for suggestions.
As for me, as an electronics engineer, I tried to define and analyze the problem in theory, assuming that the internal camera voltage measurements are accurate. With CHDK I reveive readings about 2,25V when the red indicator starts to blink. Recently I've build an automated rechargeable battery diagnostics system (including advanced pulse charging methods and reamining charge estimation) and with Eneloop/Infinium batteries (freshly charged) on A570 IS I get discharge up to around 70%, while the same batteries left at least one week unused (fully charged) can be dischargewd only about 50%. On typical GP Ni-MH in any conditions (excluding very low temperatures) the discharge is at least 90%. In my opinion, is is a battery issue, but looking for a better charger won't solve the problem. So, the only way to cope with this problem would be to FORCE the camera software to ignore the voltage measurements and let the user decide, when the batteries are depleted.
I agree, that every battery finally reaches full discharge, and the camera will stop responding or start to do some weird mechanical operations, but that's not the point. I just KNOW, that in most cases (as stated earlier) there is still some charge available in the batteries. This is especially irritating, when i.e. the batteries were charged 2 weeks ago and the camera made only some photos. Everyting looks fine, but after a few shots with flash suddenly everythoing goes dark and You get thew "Replace Batteries" message.
The wole problem is caused by he battery internal resistance, which increses with depth-of-discharge, and is also NON-LINEAR. When the camera is in idle state (no shutter pressed, no zooming/flash) it drains some current (few hundred mA), and the voltage drop is small. When the camera starts to drain higer currents (above 1A) the voltage drop becomes larger, but not proportionally. Increasing battery current causes also internal battery resistance to increase. This is some kind of an avalanche effect - increasing current 2 times makes the internal resistance increase several times. This is mostly observable with high battery current (1C or more), and could be dealt with use of parallel connection of large electrolytic capacitors. However, it is not applicable in small portable devices such as pocket cameras. The whole device and software relies on the performance of a battery and I guess, the prototypes are tested with standard types of batteries. The technology used in Infinium/Eneloop Ni-MH leads to extension of storage time without self discharge at the expense of increased internal resistance. In such case, those batteries are best suited for low power devices, and NOT for digicams.
Eventually - solution the problem is to forget about the problem and buy a high performance Ni-MHs. I just wondered, if such an advanced team developing CHDK would have any means to address this issue. Regards
