That is what i tried to say, but i agree with ndi that i messed a little with units (''appleoranges" )

fwiw, here is the result of a test run:

# chrge volts %
0 1070 4052 85
1 974 3976 --
2 920 3976 72
3 963 3976 76
4 954 3941 75
5 964 3935 77
6 910 3952 72
7 921 3923 73
8 876 3947 69
9 910 3947 72
10 838 3923 66
11 887 3917 70
12 838 3905 66
13 810 3917 64
14 824 3911 65
15 796 3899 62
16 810 3894 64
17 782 3888 61
18 768 3882 60
19 729 3852 58
20 711 3823 55
21 743 3858 58
22 729 3870 57
23 687 3847 53
24 673 3829 53
25 673 3799 53
26 660 3847 52
27 648 3829 51
28 648 3823 51
29 639 3817 50
30 638 3799 50

I stopped the test run after the 30th boots. My guess is that it would take 110 to 120 boots to kill a 100% charged battery. I have rechecked for any possible typo and there is none.

This is what I did:
=============
Between boot 1 to 9:
1. power on
2. waited for all icons at desktop(#4 view) to show up.
3. wait 10 sec
4. launch battery graph to record the values
5. close battery graph
6. switch off
7. wait about 30sec and Goto 1 again.

Between boot 10 to boot 30th.
1. power on
2. waited for all icons at desktop(#4 view) to show up
3. waited until the CPU load widget showing 1 bar.
4. wait 10 sec
5. launch battery graph to record the values
6. close battery graph
7. switch off
8. wait about 1.5 to 2 min and Goto 1 again.

A few notes about my setup:
=====================
My unit has power 47 kernel loaded(but only the default N900 voltage profile is used).

At boot 0th, the % battery was from the advanced power plugin.

At boot 1st, I forgot to record the % battery.

Wifi is enabled, but there is NO internet connection.

Bluetooth is switched off as shown by the advanced interface switcher.

N900 General Profile is used with System sounds set at level 1; speaker volume is set to OFF.

Brightness is set at max level.

Cheers,

good work 'cheve'

Facts instead of speculation, that the way we like it.

Can see that at boot #30 the voltage was 3.8V which mens that there where about 50% left at that point, so the battery can take up to 60 boots on a charge with your setting and method.

So just doing boots without the waiting etc, it should do up to 100 boots, meaning that the actual drain on the battery is about 1% e.g. 14mA used. but for the sake of the 60 boots we get 1.66% or 22mA used (mA was rounded up)

And for the sake of W we get
3.8V * 14mA = 0.05W (180Ws)
3.8V * 22mA = 0.08W (288Ws)

the Ws was for 'ndi'

Dr frost:

For your first post, it makes me feel you didn't read my last post . I have explicitly stated that I was referring to an impossible scenario. I don't know what more to say.

For your second post, I use Ws because of SI units. Also, for clarity.

@cheve
I read your initial post but couldn't find it again. You posted this experiment before, right?

no, it is the 1st time.

i hope it is somewhat helpful

Ah, it was Matan. I knew I read this before.

very useful, thanks for taking the time to put this "myth" to rest.

I always switched off the mobiles that I had.
Maybe that is the reason why I get up to 3 days with my
N900.
;-)

Since I have my N900 as only phone I even wonder about the strange habit that we all had with landline phones keeping them "online" all night long.
;-)

Stock N900 does maintenance overnight - checks for updates, cleans up calendar, etc.

Unless You change that behavior to do it in noon/twilight/whatever Anyway, shall we mark this topic as "concluded" and maybe do some wiki article? what do you guys think?