View Single Post
igor's Avatar
Posts: 198 | Thanked: 273 times | Joined on Jan 2006 @ Helsinki, Finland
#138
This is something that maybe is not so clear, so please forgive me if i'm repeating the obvious.
When the tablet is charging, it _is_ on.
It stays on as long as the charger cable is plugged in (and the cable connected to the power lines, of course).
This fact has probably already been stated elsewhere and should be obvious by the different time it takes to boot from real off (device off and no charger connected) and from "acting dead".
Another hint is the fact that the charging animation doesn't start immediately, when plugging in the charger in a _really_ off unit. It's because it is actually performing the first phases of the booting.

Similarly, when the charger is disconnected and the device was not showing the desktop (meaning that it was in acting dead mode), it will power off for good.

Pressing the power button right after unplugging the charger is likely to not obtain the desired effect (turning on the device, which is actually performing a shutdown sequence).

Said this, it is possible that there is either some sw or (somewhat unlikely on large scale) hw bug.

The simplest (and quickest) way to turn on a device that is in charging/acting dead mode, is to press the power button and let the unit complete the booting sequence _before_ pulling the cable.

Since there are 3 different watchdogs (1 sw plus 2 hw), it is possible that they kick in and generate a reset, in case of lockup due to some race condition. This can explain why sometimes the time it takes to reboot after becoming unresponsive seems to change. It's because the lockup can happen at any moment within the wd period and therefore it will persist for all the remaining part of such period, then the wd will generate a reset.