Battery sometimes goes dead while phone is on charger.

[Mike (mwester)]

Russell Sears wrote:
> I wrote the support list and filed a bug about this.  Andy said this is
> a more appropriate forum.
> 
> Sometimes, when I plug in my phone to the wall charger, it doesn't
> charge, and instead slowly discharges the battery.  I'm not the only one
> to see this, and it seems to be a software bug.
> 
> Here are the symptoms, and a summary of the findings so far:
> 
> apm reports:
> 
>  Off-line, battery status low: 49% (1d+3:40:00)
> 
> and the GUI icon doesn't show a lightning bolt.
> 
> charger_type contains:
> 
>  charger 1A mode 1A
> 
> dump_regs contains:
> 
>  13 84 00 00 00 00 00 80 00 00 00 00 00 d3 aa 4a
>  15 44 ff 01 00 07 00 00 02 08 6b 01 00 0a 1b 02
>  00 22 2f 01 00 22 00 00 3f 01 05 20 11 18 02 18
>  02 00 00 00 00 15 00 15 01 16 01 00 00 ff 3f 00
>  00 17 63 e6 28 19 ff ff 00 02 00 33 20 50 19 00
>  00 08 00 00 0e a3 00 f0 00 42 04 23 00 20 06 08
>  7f 7f 3f 07 3f 1f ff 00 00 00 00 00 00 00 00 00
>  00 00 00 00 00 00 00 c9 72 d9 0a 27 60 08 c0 b8
>  13 c0 03 07 00 00 00 03 e2 ff fa 62 11 00 00 00
>  00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>  00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>  00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>  00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>  00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>  00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>  00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> 
> which confirms the charger is off.  The bug is here:
> 
> https://docs.openmoko.org/trac/ticket/1158
> 
> Michael Shiloh suggested I apply this kernel patch, and force the
> charger type to 1A.  (I haven't been able to do this yet; I'm still
> figuring out the build system... Reply to me w/ instructions for
> rebuilding the kernel, and I'll update the wiki if necessary.)
> 
> http://lists.openmoko.org/pipermail/openmoko-kernel/2008-July/003761.html
> 
> -Rusty

Possibly related, maybe not -- when charging a Nokia BL-6C battery,
chgcur is blank (no reason for it to not report the charging current,
even if the 50633 doesn't know how full the device is).  Far stranger
is that one can watch the battery voltage on the BL-6C rise over time
(it's definitely charging) until suddenly the battery voltage as
reported by battvolt goes from about 4000 mv to 65535 mv.  (That's a lot
of voltage... :D :P)

(Oh, yes -- I've also observed the charge ceasing at some point, as
described in this bug report.  Once that happens, the device continues
to run until it drains the battery completely.)

Mike (mwester)