power consumption after shutdown

[Richard A. Smith]

david at lang.hm wrote:

>> The embedded controller was.  Something needs to be watching for a
>> power button press in order to know when to turn on.
> 
> by default the wireless card remains alive to participate in a
> potential mesh network, disabling wireless should give you a lot more
> time.

In power off the the WLAN is not powered up.  Only the EC remains powered.

> from the XO.  Self-discharge rate has a dependency on storage
> temperature as well.

The self discharge of LiFePO4 is quite low.  The numbers thrown about in 
the marketing docs are between 3-5%/month.  Based on the values in the 
datasheet I calculate ours to be 3.8%/month at 25 degC.  However, its 
highly temperature dependent.  The datasheet suggests that at temps >= 
45 degC   its 33%/month and at 60 degC 100%/month.

This has be somewhat experimentally verified.  There were a few boxes of 
laptops stored in a temp controlled environment for at least a year and 
when I examined those batteries they still had plenty of life but, 
batteries sent to Jordan that were stored in a hot warehouse arrived 
completely discharged.

> I did this just now, on a unit running Q2E27, and another running
> Q2E30, the current is 24mA at 12.9V powered from a large sealed lead
> acid battery with nothing else attached.  Two days of this would be
> 1.15 amp hours (Ah).  Three days would be 1.73 Ah.

While running on eternal power the EC does not attempt to enter a low 
power mode.  This is because its monitoring the battery.  This is 
sub-optimal while on things like solar power so at some point this may 
roll up on the "Fixme:" list but theres a lot of conditions that have to 
be dealt with.  For now Ext power == EC on full.

When powered off and on battery power only the EC attempts to to go into 
the STOP state where power consumption is much less.  Not as low as it 
should be (see below) but much less.

> days later I insert the AC adapter into the (still closed) XO, after 
> the 'power' light comes on, it goes yellow.  [I've seen this both 
> when I was running 'joyride' on the XO, and when I was running 
> 'staging'.]  I take this to mean that *something* was draining some 
> power for the two days the XO was sitting in its "shut down" state.

As mentioned above the EC attempts to go into STOP mode when the power 
is off and not plugged up to external power.  The data sheet indicates 
that STOP mode is supposed to be in the 10s of uA range.  Measurements 
taken (this morning against my EC dev tree) via instrumented XO indicate 
that its 4mA.  Studying the datasheet that suggests that rather then 
going into STOP mode the EC is only going into IDEL mode.  In IDEL the 
instruction fetch is turned off but the clocks and timers still run. 
4mA is listed as the nominal draw in that mode.

Looking at the EC code the go-into-STOP-mode rather than the 
go-into-IDEL-mode bit is getting set.  So the code is _trying_ to go 
into STOP mode but its power draw suggests otherwise.  I'll have to 
study it deeper.

In anycase @ a 4mA draw the EC will see the net ACR diff at about 
3.1%/24hours so across 2 days you would have probably discharged enough 
for the EC to enable the charger to top up the battery.

-- 
Richard Smith  <richard at laptop.org>
One Laptop Per Child