Main Battery - alternatives ?

[SAE140]

Up until yesterday I had one excellent (2 hrs plus) 4200MaH battery, which is seldom used as I tend to run my 600's straight from a 12 volt accumulator (via an adapter made from an old battery case, with diodes, an over-voltage zener and anti-spike toroids inside).

Yesterday I tried to power-up via the main battery only to discover that it was 'dead'. Not surprising really, considering how little use it gets. Checking with a multimeter returned 10.25 volts. Assuming that this was just below the permitted charging threshold, I redirected some voltage from a 12 volt gel-cell (via a resistor and ammeter) until the battery registered 10.6 volts. When re-inserted, it then began to charge ..... until the 'green led' glowed, but with only 25 % capacity being indicated by the 'smart-stuff' inside.

Well, I've had enough of playing with Li-Ion cells - they may be great if you're using a laptop on a regular basis, but I don't ... They begin to deteriorate from the first day of manufacture, whether they are being used or not - so I'm looking for an alternative solution.

Can't do a useful Forum search for battery-related issues, as the common search-term "battery" is ignored.

Ok. I'm confident of fitting 8 x 2700 mAh NiHM's into an empty case (chargeable via an external charger) - which, although of a lower capacity, would be more appropriate for my style of use. But before trying anything radical like this - does anyone with the right equipment happen to know what information is conveyed via the 2 x central pins of a Li-Ion battery: an analogue voltage (simply 'sense pins' ?), or a stream of digital information ? Could be either.

[SAE140]

I decided to run a couple of 'feasibility tests', using 8x and 9x 2500 mAH Uniross Encore AA-sized NiMH cells ('cause that's what I had at hand).
Although the 9x cell pack measured 12.90 volts (IBM's max battery voltage spec being 12.6 volts), as soon as a load was applied the measured voltage was below 12 volts.

Machine on test was a Thinkpad 600 (P2-300, 64Mb ram), generally run under 'worst-case' conditions: no power saving measures, screen on full brightness etc., except medium HDD read activity, no writing (afaik) during the test.

The 8x cell-pack gave 45 mins run time, the x9 pack gave 58 mins. Using 9x 2700 mAH cells would take the run-time to a little over an hour.

Is this a worthwhile project ? Well, yes - if like me you're an infrequent battery user.