More experiments on this subject.
iPAQ h5550 naked (no expansion jackets):
Operational mode, wifi driver (usbvnetr loaded)
backlight full -- 560mA -- about 2 hours battery life
backlight off -- 390mA -- about 3 hours
560-390=170mA consumed by the backlight
wifi driver unloaded, backlight off -- 140mA -- about 9 hours
390-155=135mA consumed by wifi
the same, Bluetooth radio on -- 155mA
the same, bnep on, IP connection established -- 155mA
The above figures are not very accurate, they are the average values
obtained from /proc/asic/battery.
Now testing the sleep mode.
wifi driver unloaded, MMC card removed, Bluetooth off -- 14.5mA (over 1.5
hour period). This gives us about 90 hours battery life (4 days). This is
inconsistent with my previous experiments when the expansion jacket was
installed (an additional battery and a microdrive IBM 1G). Need to check
this thoroughly.
Alex
On Fri, 27 Aug 2004, Alex Bystrov wrote:
> Under the same setup, but with the driver loaded the unit lost further
> 20% in 10 hours of sleep. The orange LED was off as I managed to guess the
> time and to press "suspend" so that it was caught in the "off" state.
>
> So, the driver unloading helps, but something else drains the power. I'll
> try unloading Bluetooth and MMC card drivers also. I am having doubts now
> regarding the CPU being suspended properly.
>
> I am talking about hh36.11 kernel on h5550 and the unstable gpe.
>
> Alex
>
>
> On Thu, 26 Aug 2004, Alex Bystrov wrote:
> > Is it the fault in the driver that it cannot be removed without the above
> > hassle? I am experimenting now checking battery life in the sleep mode.
> > The first impression is that without the driver the consumption is lower.
> > Still the unit looses about 30% in 20 hours and I am using the expansion
> > pack with a second battery.
>
>
>
Received on Fri Aug 27 11:42:13 2004
This archive was generated by hypermail 2.1.8 : Fri Aug 27 2004 - 11:42:34 EDT