Strange battery behaviour - or perhaps very, very clever?

[jrmwalsh]

The unusual characteristic is that the battery discharge voltage suddenly INCREASES part-way through its discharge cycle.

I have a Z61t Thinkpad whose battery was not giving me enough life. (I actually have three x Z61t laptops and four x batteries for them). The batteries are all Lenovo 42T4512, 5.2 AH.

The battery in question, Battery 2, seemed to be not giving me enough life so I decided to measure its capacity by charging it, then discharging it through a known load (so I could calculate current) while measuring its voltage every minute, until it turned off. This is where I noticed very unusual behaviour.

To discharge the battery-pack I connected a 24V 18W light globe to it's outermost terminals, as a load.
I temporarily connected an ammeter in series with the globe to measure the current, which was about 850 mA. This is similar to the normal load on the battery when the laptop is running.

I connected a data-logging voltmeter across the leads of the globe, and began logging the voltage while the battery discharged by lighting the globe. The discharge voltage began as 11.94 volts and slowly dropped to 10.45 volts over 103 minutes (1 hrs; 43 minutes), then the voltage suddenly INCREASED to 15.9 volts and the battery continued discharging as per the experiment for another 1 hr:20 minutes.

I have repeated this trial with the other batteries and they sometimes behave the same way too.

I have noticed two different, and distinct, discharge patterns;

Pattern 1:
The discharge voltage begins at about 15.5 V - 16 V, then slowly drops off to around 13.9-14 volts, then the voltage suddenly drops by about 2 volts to about 12.0 volts. Presumably this sudden drop shows that one of the cells has fully discharged and has stopped delivering current. The voltage then continues to drop slowly to about 9.6 volts then drops immediately to 0. Presumably this is when the entire battery pack is now discharged.

Pattern 2
As described a few paragraphs above. The voltage starts at about 12 volts, and drops slowly to about 10.3-10.5 volts, over about 1 hr: 20 mins, then the voltage suddenly increases by about 3.5-3.8 volts to about 16 volts, then continues to drop off for about another 1 hr: 20 mins, until the pack is discharged.

I cycled the battery three times, and checked its discharge pattern each time. The first time it followed discharge Pattern 1, then it followed Pattern 2, then Pattern 1 again.

I've checked the other three battery packs and they all show similar characteristics of sometimes following discharge Pattern 1 and sometimes following Pattern 2. I can't predict which Pattern they will follow from one charge to the next.

Can anyone explain what might be happening?
Does the pack have a very sophisticated control system that can switch cells in and out as their charge drops off during a discharge cycle?
I can guess, but does anyone REALLY know?
I can provide all the data in an excel spreadsheet, with graphs, if anyone is interested.

thanks

Mike Walsh