Here are the results of some experiments on 15" 667 and 800 MHz battery runtime estimate experiments. The most interesting result is that I repeated Case 2 with different amounts of RAM installed, 256 MB up to 1 GB, and found that there was **no difference** in the battery runtime estimate. There is quite a bit of "noise" in these estimates - it varies plus or minus 10 minutes with each trial - but there was definitely no correlation with the amount of RAM installed. Runtime Estimate For Each Case ---------------- Purchase Battery SerNum 1 2 3 Date Usage since purchase -------------- ---- ---- ---- -------- ----------------------------- ZZ2322DGL8VB 2:53 2:00 1:36 2002 Sep Used half-time in 667 MHz PB LW1406VXL8W 2:46 1:31 2002 Jan Used half-time in 667 MHz PB ZZ21507AL8VB 2:29 1:15 2002 May Used full-time in 800 MHz PB Experiment was done with three different batteries, all model number M8511 (667 MHz and 800 MHz generation). All batteries have been "calibrated"; that is, allowed to run down until the computer went to sleep. Each battery was fully charged to begin the experiement. In each case, I installed the battery, pulled the power cord and allowed the battery estimate to stabilize for 2 minutes, then read the "Time Remaining" Runtime Estimate. Case 1. 667 MHz 15-inch Powerbook G4 with 512 MB RAM, 30 GB HD, Display 1152 x 768, Brightness Maximum, running only Finder with no windows open. Fan off, Airport installed but off. Case 2. Same as Case 1 except a different Powerbook: 800 MHz 15-inch with 40 GB HD, Display 1280 x 854. Case 3. Same 800 MHz Powerbook as Case 2, except running FInder with 4 windows in the background, LaunchBar, CopyPaste, Entourage, Word with 3 documents open, QuicKeys X, Synchronize Pro! X and Internet Explorer. Entourage was in front, all other applications idle. Fan was off. -- Jerry Krinock San Jose, CA USA The opinions expressed in this message are those of the signatory and do not necessarily reflect the policies of the current un-elected United States government, and vice versa.