Bobby Powers bobbypowers at gmail.com
Sat Jul 26 21:34:56 EDT 2008

```On Tue, Jul 22, 2008 at 3:49 PM, Richard A. Smith <richard at laptop.org> wrote:
> mianeil2 at yahoo.com wrote:
>
>> I was also wondering if you could give me feedback on this table. The
>> table shows how much kWh is needed a year to power a xo based on
>> different scenarios. If you think I should add or change anything
>
> As I often state in my discussions on laptop power, calculations like
> this are actually pretty complex and the simplistic approach while good
> for ballpark estimations can have a large amount of error.
>
> First as others pointed out you units are wrong.  You need to substitute
>  Watts every where you have kW.
>
> The next issue is that you are assuming a perfect conversion on the
> recharge half of your cycle.  Which is not correct.  The avg power draw
> of 5-7 watts for the XO is measured internally either via the battery
> sensor or by our instrumented XO.  It does not take into account the
> efficiency of the DC/DC converter when recharging the XO from external
> power.
> It also does not take into consideration the charge efficiency of the
> battery.  It takes more power charge a battery than just the usable
> capacity of the battery.
>
> The DC/DC converter's efficiency is affected by the difference between
> the input voltage and the output voltage.  We don't really have any
> numbers on the exact range of efficiency for the XO @ 12V but typically
> your average DC/DC converter is around 85%. The 88% number pops into
> mind from when I was last looking at such things.
>
> Then there is the charge efficiency.  Which is more complex because its
> actually 2 numbers. One for constant current (CC) charge mode and then
> another for constant voltage (CV) charge.  The batteries start off in CC
> mode and then switch to CV mode after certain criteria are reached the
> criteria happens around the same time but is a bit different for each
> battery and much more different between the 2 types of batteries.
>
> I don't have numbers for these efficiencies.  The EC code has comments
> with magical constants that suggest certain numbers for these values but
> I've learned that a lot of those comments may be wrong or apply to
> earlier versions of the batteries.   The range suggested is 80 - 90%.

would (somehow) updating these magical constants for current batteries
(depending on LiFe or NiCad chemistries) improve anything?

bobby

> The only way to know exactly what a good average for charge efficiency
> is would be to measure and compare the power in with power out across
> several batteries of each type (remember we have 2 chemistries).
>
> Thats possible in the case where the XO is powered up and you can read
> the battery sensor, but when the XO is off its not so easy.  Guess what?
>  Your %'s will be different in the 2 cases because the charge rate is
> much faster when the XO is off.
>
> Buts lets just say for a quick ball park that DC/DC is 88% and average
> battery CE is 85%.  Now your recharge numbers are off by about 25%.
>
> Listing things in "cranking hours" may also be problematic.   If you
> really were cranking some human power device your output would be so
> variable that the only way to get meaningful data is to measure it and
> develop some sort of profile for what the average person can really do.
>
> --
> Richard Smith  <richard at laptop.org>
> One Laptop Per Child
> _______________________________________________
> Devel mailing list
> Devel at lists.laptop.org
> http://lists.laptop.org/listinfo/devel
>
```