Rockbox.org home
release
dev builds
extras
themes manual
wiki
device status forums
mailing lists
IRC bugs
patches
dev guide



Rockbox mail archive

Subject: Re[2]: charging algorithm stops and restarts
From: Uwe Freese (mail_at_uwe-freese.de)
Date: 2002-08-17


Hello Heikki,

Heikki Hannikainen wrote on Saturday, August 17, 2002, 8:45:29 AM:

>> minute later. That happened here several times with charging really
>> empty batteries (where the charging stopped in the middle of the
>> whole charging time because of a low gradient).

HH> Hm, what was the voltage and did it stop with low or negative
HH> delta?

I think the voltage was about 5.5 v (normally the charging stops here
at about 5,8 v), but I don't know which delta it was. Stopping and
Starting happened here several times within the same charging cycle.
It could be that is first was the 30min low delta, but because of the
charging pause of one minute, a negative delta was detected some
minutes later.

HH> Interesting, I haven't seen this happening with my batteries
HH> (completely discharged or not). My batteries are pretty new,
HH> though.

My recorder is only some weeks old (and the batteries, too). I charged
them many times when they were ~50% full and the complete charging
yesterday was the first one I tried.

>> Disabling charging for one minute disturbs the calculation of the
>> voltages and deltas and is not so nice.

HH> It doesn't disturb that, since batteries are charged for at least
HH> CHARGE_MIN_TIME (10) minutes before deltas are calculated or
HH> checked. This is intended to get over the erroneus voltage drops
HH> when starting to charge empty/old batteries. But wait, you're
HH> right - CHARGE_END_ZEROD is now 30 minutes and the voltage might
HH> well have been higher 30 minutes ago if the unit was not powered
HH> down before it got the charger attached! I wonder if we should
HH> clear the history when starting charging? This might just fix the
HH> problem.

I don't think that is the right way. It's a false detection of the
ending of the charging cycle. So we should not start a new charging
cycle but prevent that the false detection occurs.

>> Add a BATTERY_LEVEL_CHARGE_END.
>> Do not stop charging until voltage > BATTERY_LEVEL_CHARGE_END.

HH> I don't think this is a good idea, as we might end up charging
HH> some bad/old batteries forever.

But it is exactly the same as it is now: The voltage is so low that a
new charging is started. I only want that the same charging cycle goes
on and not that a new one is started.

(This is not the solution here, but) if we want to prevent a "charging
forever", we could allow only one charging cycle while the charger is
attached. For the next charging cycle the user has to detach and
reattach the charger.

HH> We do not really know the full/empty levels of
HH> a given battery set before measuring the individual cells.

The level should be low enough and I think 85% could work.

To summarize the problem: The charging has a phase in the middle where
the voltage goes up only a little bit. And this is detected as the end
of the charging sometimes. Now you know why I added the "phase
detection" feature (read my old mail) to my old charger. :-)
Something has to be done so that the charging doesn't stop in the
"middle" (when the battries are only charged to ~50%). My ideas are
- don't stop until voltage > x
- "phase detection"

Bye, Uwe.



Page was last modified "Jan 10 2012" The Rockbox Crew
aaa