Dramatically reduced battery life after installing Rockbox (h300)

[strandz]

Hi,
hope somebody can help. I just discovered Rockbox and is extremely excited about all the features. However, after installing the software on my iRiverH320 the battery life has more than halfed. The problem does not occur when in playback mode...the problem seems worst when the player is turned off. The battery is gradually (but fast) drained while the player is shut down. I hope somebody can help, as the reduced battery life may force me back to using the old iRiver firmware.

[petur]

Weird. Rockbox should increase battery life.

[Llorean]

I've heard ONE report ever of an increase in the rate at which battery drains while turned off. It wasn't a Rockbox'ed player though. It was someone with a US type player who'd simply installed the Korean firmware at all.

[belly917]

I've heard ONE report ever of an increase in the rate at which battery drains while turned off. It wasn't a Rockbox'ed player though. It was someone with a US type player who'd simply installed the Korean firmware at all.

I noticed that my non-US h320 drew more power when I enabled the iriver firmware's clock... try disabling the clock in the iriver firmware.

[fizze]

one cannot disable the clock. You could just flash on a version without RTC support.... but I doubt that would effect rockbox..... ?

[cheeto81]

i have an ipod color photo, and my battery drains all the time very qucikly when i shut down the rockbox

[XavierGr]

Rockbox increases battery life on all supported iriver targets.
This is dtermined by the various reports posted on the wiki. (check iriverruntime)

There once was a serious bug with H300 that halved the battery runtime but thanks to amiconn these are forgotten days...

[cumulusnimbus]

Yes, there definetly is an issue.

I cannot exactly tell when it started with my player, but it seems to correlate with installing rockbox.

I just replaced my battery with a better one (ipod), because I thought it was an issue with an old battery.
Then I connected it to some measurement instrument in my company just to see what the current consumption is in some usecases.
There I could see that in power off 50mA (!) are drawn from the battery (see attached picture).

I have made some tests with most of the available bootloaders, patched and original (1.20 ... 1.30). The behaviour does not change, still 50mA in power off :-(((
Rockbox update doesn't help too, neither does "set to defaults" (std. firmware).

I'm lost. Does anyone have an idea how to get rid of this problem?

Rüdiger



[attachment deleted by admin, too old]

[XavierGr]

That makes me wonder!

I am sure that my H300 gives more battery time, though all the tests are done to units fully charged, so maybe this can't be seen by a simple battery benchmark.

Have you checked if there is power consumption while the unit is turned off with the default firmware? (unflash the bootloader)

[Febs]

His post indicates that he has already tried that and that there is no difference.

[cumulusnimbus]

yes, no difference between rockbox and std. bootloader

is it possible that there is a HW failure that causes such a symptom?

[pabouk]

I think that someone reported the same problem in this forum. It was more than two months ago. Unfortunately I cannot find the thread anymore.

Could not the problem be caused by some improperly configured HW component i.e. RT clock?

[cumulusnimbus]

maybe it was this one?
http://forums.rockbox.org/index.php?topic=5204.0

unfortunately doesn't help here

[pabouk]

No, it was about H300 series and battery discharging when the player is turned off.

Maybe the same method used by Amiconn can be used here. You can check all the ICs by a thermal sensor or an IR camera to find the place where the power is dissipated. 3.7 V * 50 mA = 185 mW - this should be visible.

[cumulusnimbus]

a first intermediate summary:

the problem (50mA current when powered off, battery completely flat after 24..36h)
- is independent of FW, bootloader and rockbox-SW
- persists after completely disconnecting the battery for half a day
- persists after a HW-reset

questions:
- Are there any configuration data that are buffered in a kind of NV-RAM, EEprom or buffered CMOS RAM?
- Is there another internal buffer battery/capacitor?
- Are there known HW defects that would fit to this kind of problem? I'm thinking e.g. about the charging circuit. In addition to the iriver-charger I used a proprietary charger for some time that - at least unloaded - supplies up to 6.5V.