The AOS battery issue on our device seems to be, as far as I can tell, a reporting bug only the majority of the time.
My best battery life was when AOS was at 60%+.
My new experimental kernels seem to eliminate this reporting bug - AOS now is <5% at all times for me, but drain has never changed - it was never that high to begin with though. Less than 1% an hour in moderate signal, with typical deep sleep percentages at 93-95%. I don't recall of gtg included 10/9 experimental or non-experimental. If CPUSpy says 2.6.35.7, it's non-experimental and AOS will likely get falsely reported high for you. If it says 2.6.35.14, AOS will not get falsely reported high. Actual battery usage will be the same - at least in the cases I've seen on my device, the AOS bug is reporting-only and doesn't cause drain. (Yes, sometimes AOS can be associated with drain - but not always.)
This should have small battery performance improvements compared to the stock kernel, except in some corner cases that might have caused extreme drain before. (On the Infuse, the classic case for codeworkx's cpuidle patch was "weak-signal Gmail sync battery rape - Gmail sync would sometimes suck down 15-20%/hour or more in weak signal conditions prior to the cpuidle fix, that dropped to 5% when it happened after the patch.)
Note that some people have had oddball performance/drain issues that went away after a wipe data/factory reset.
An active radio and associated background data transfers dominate battery usage in our devices, just like they did with the Infuse. I get similar battery results with this device to my Infuse - drain in moderate signal less than 1% per hour, with deep-sleep at 93-95%. An Infuse in airplane mode improves to less than 5% per 24 hours with 99% deep sleep - it is amazing how much our radio dominates battery usage. I haven't tried an airplane mode test on this device.
At my desk at work I'm a bit over 1% an hour, 93% deep sleep. The Infuse approached 1.5-2%/hour in these conditions. (Extremely weak signal, bouncing between 1 bar and no signal.) Weak signal always increases battery usage as the radio cranks up its transmit power to compensate.
Last edited:
