iPod Mini won't update battery status

[eicca]

The battery level on my 2G iPod Mini only refreshes upon bootup or plug-in. Otherwise is stays static at whatever percentage level it started. Kind of makes it impossible to know how much battery is left, which really cripples the usability.

Looks like there are a few other threads from iPod users of various models with similar battery level issues.

How can we fix this?

[eicca]

UPDATE: I have traced the issue to database updating. If auto-update is enabled, or the update/initialize functions are invoked through the settings menu, then both the battery meter and disk indicator get stuck.

Reading battery status through debug menu still works as expected.

[roastpotatoes]

I think I'm having the same issue. The battery indicator doesn't seem to have any meaning

[eicca]

UPDATE: On 3.14 the battery indicator gets stuck once the database update process happens. So if auto-update is enabled, say goodbye to battery updates. On 3.15 it's just stuck all the time.

[eicca]

Any update/input on this? It's kind of a game-breaking bug here. Can't use Rockbox if I can't see what my battery is!

Trying a dev build doesn't help anything.

[roastpotatoes]

Any update/input on this? It's kind of a game-breaking bug here. Can't use Rockbox if I can't see what my battery is!

Trying a dev build doesn't help anything.

Try the latest dev build I believe this issue has been solved.

[eicca]

Any update/input on this? It's kind of a game-breaking bug here. Can't use Rockbox if I can't see what my battery is!

Trying a dev build doesn't help anything.

Try the latest dev build I believe this issue has been solved.

It does indeed look like it has been solved, for the most part. A database update seems to still freeze the indicator, but I don't auto-update so I can manage it.

[speachy]

It does indeed look like it has been solved, for the most part. A database update seems to still freeze the indicator, but I don't auto-update so I can manage it.

This happens on all ATA-based players; we deliberately do not update the reported battery stats while the disk is spinning as it typically causes a considerable dip due to the increased load.  Once the disk has finished doing its thing and is shut down, a more accurate battery level (and thus, expected runtime) can be ascertained.

It can be argued that with some solid-state storage this isn't necessary (notably excluding the mSATA converters, which draw _more_ power than the stock hard drives..) but to date I've not been able to come up with a better-than-50%-reliable heuristic that can be used to gate this behaviour.