Article: 8462 of comp.sys.laptops Path: uwm.edu!cs.utexas.edu!uunet!noc.near.net!news.bbn.com!news.bbn.com!wbe From: wbe@bbn.com (Winston Edmond) Newsgroups: comp.sys.laptops Subject: Very dead T1000SE battery revived Date: 16 Dec 92 03:01:31 Organization: Bolt Beranek and Newman, Inc., Cambridge, MA Lines: 87 Distribution: world Message-ID: NNTP-Posting-Host: crystal.bbn.com Thanks to suggestions from this newsgroup, and after some experimenting, I seem to have managed to revive a completely dead T1000SE battery pack. The message rambles a bit so that others that have experienced the same problem may compare the solution I found with others that have been discussed in this newsgroup. Completely dead battery = 0V output, 0 Ohms resistance, regardless of how long it was "charged" on a T1000SE. Current state = 7.2V output, and it just powered my T1000SE for 75 minutes from full charge (green light) to first low charge beep, while running the screen at medium brightness and using the 2400 bps modem. Yes, 75 minutes isn't as good as new, but this is the first full charge/discharge cycle after revival, so I don't yet know if it will get better with proper use, or get worse again. BACKGROUND (how I got into this mess and what didn't work): My problem started when the battery pack began running out of power much sooner than it "should have". I tried deep discharging, dropping the battery pack from a moderately high distance (and other forms of physical bashing that some people thought might help), and various other tricks over the course of time. At first, these seemed to help slightly, but the improvements were only temporary. Eventually, while using an auto light bulb to discharge the battery pack, I forgot the advice to not go below 1 Volt and let the pack discharge completely. After trying to recharge it, I could get the green light to come on, but the system would INSTANTLY shut down if the wall-plug power unit was unplugged. A Voltmeter showed that the battery was only putting out 4.8V. (NOTE: the green light doesn't mean the battery pack is fully charged -- it means the battery pack isn't likely to accept any more charge, and that only means "fully charged" if all the cells are working.) Continued attempts to revive the battery pack eventually left me with a battery pack in which all cells were "dead" (in "cell reversal" mode, I think, but I don't understand NiCads well enough to be sure) -- 0V, 0 Ohms resistance, immune to all the simple attempts I tried to charge it. WHAT DID WORK: Someone on this newsgroup suggested using a large electrolytic capacitor charged to 8V and discharged through a completely discharged battery pack. I didn't have a big enough electrolytic capacitor to do the job, so I tried two alternatives: (1) another battery pack, and (2) an industrial power supply capable of supplying 10 Volts at 10 Amps. Partial success was obtained by connecting a working, fully charged battery pack's + to the now-completely-dead battery pack's +, and - to -, with just wires, for a few seconds. This produces a modestly bright spark. Using a second battery pack was good enough to revive 4 of the 6 NiCad cells in the pack, but wasn't enough to bring back the last two. It might have been enough if I'd put some regular batteries in series (and in parallel) to get the voltage and current capacity up a bit. The first industrial power supply I tried could only provide up to 5A at up to 10V. This wasn't enough. The second power supply I tried could supply 10A at up to 20V, and 10A at just 10-12V was enough. The power supply was a high grade unit with adjustable current and voltage limits. I connected + to +, - to -, and let 10A at 12V flow through the battery for a few seconds. That brought the battery pack back up to a full 7.2V! The rest of the charging I did on the T1000SE, because I think feeding 120W into a few NiCad batteries for more than a few seconds at a time is unwise. :-) DISCLAIMER: This message provides history and opinions, not advice. If you elect to try to duplicate this success, you do so at your own risk. I did find that a Voltmeter for measuring the battery pack voltage (which was always N * 1.2V) was indispensible. COMMENTS ON OTHER SOLUTIONS (given what worked for me): * The big electrolytic capacitor method probably would work, though it might take several "zaps". * Using a 12V car battery probably also works, but is overkill and looks to be much more dangerous since the current isn't limited and it probably causes sparks. (The bench supply I used caused no sparks when I touched the probes to the battery pack, even though it instantly went between 0A and 10A output.) * Connecting up a set of 7 1.5V dry cells (or 9 1.2V NiCad cells) in series to get 10+V, probably with N >= 2 sets in parallel to get sufficient amperage, should also be able to do the job. I don't know of anyone that's tried this, though. -WBE