From: Brian <GCAdidas13_at_columbus.rr.com>
Date: Sun, 19 Jan 2003 17:52:16 -0500

Regarding this, is it possible that temperature plays a part in the
battery's percentage? For example, yesterday i took it skiing with me:

Car ride there - 100%-90%
turn off
then after an hour on slopes without it on, i turn it on and it is at 75%
use for next 4 hours or so
turn off
car ride home - right when i get in car, turn on, at 16%
15 mins later - 25%
30 min later - 10%
and then down to 4% then it cut off.

I dont know how this happened, but one other thing to consider is that this
was only the second day i used my new Recorder.



----- Original Message -----
From: "Uwe Freese" <mail_at_uwe-freese.de>
To: "Holger Lembke" <rockbox_at_cool.haxx.se>
Sent: Sunday, January 19, 2003 4:38 PM
Subject: Re: some Ideas about Batteries in Powermanagemant


> Hello Holger,
>
> Holger Lembke wrote on Sunday, January 19, 2003, 8:34:50 PM:
>
> HL> Ok, the story. My AJBR has a distracting capacity display. If I start
using
> HL> it, it has 100%. Then the capacity display drops rather fast to
something
> HL> around 60%. It falls slowly to 40% and 30% sometimes comes back and
than
> HL> quickly falls down to low percentages and my AJBR stops working.
>
> These quick changes aren't usual. Your batteries or your unit could be
> broken!
>
> HL> After looking at
> HL> http://www.stud.uni-karlsruhe.de/~usng/rockbox/ladeverfahren.html and
> HL> looking at 'static int percent_to_volt_decharge[11]' I think it is
easy to
> HL> see why it is this way.
>
> HL> The voltage values simply dont give the resultion to get a better idea
> HL> about the capacity at the flat part of the discharge curve.
>
> The "resolution" is enough. In the middle part of the courve, the
> voltage has a linear form there.
>
> HL> If this is right, could there be a better way of showing the capacity?
>
> HL> Is it possible to calculate it by runtime? Simply have a thread that
> HL> checkes the on/off state energy consuming items (hdd, backlite etc),
knows
> HL> the drain and capacity and does some math?
>
> I also had this idea some months ago. I already started to write a test
> version of this kind. But I rejected this idea after some tests.
> If you charge/uncharge the batteries completely every time, it may work.
> But if you charge/decharge only litte amounts, it will drift off very
> quickly.
>
> HL> Or use the history (that from charging) for a better calculation?
>
> In which way? If you have ideas, let us know.
>
> Bye, Uwe.
>
Received on 2003-01-19