Re: [NTLK] eMate battery problems

At 11:40 AM 2/10/2007, you wrote:
>A long time ago Christian wrote:

No, that was me.

> > I've noticed that if you are charging and you set the
> > clock the Newt will complain about the battery not being
> > able to take a charge. I suspect that it uses the
> > current clock time to figure out the charging rate and
> > if it is too slow (since you set the clock ahead a few
> > years) the battery is NFG.
>
>This can't be confirmed at least for a Newton 2100. I just
>soft-resetted one. It started charging with a battery level
>of 12%. Then I changed the year from 2006 to 2007, and the
>Newton did not complain of a dead battery.
>
>Amazingly, though, it went from "Fast Charge" and 12% to
>"Trickle Charge Continuous" and 100% as soon as I changed
>the year. Obviously the charge circuitry DOES monitor the
>actual date and time.

Are you using some sort of battery monitoring software? I was just
going on what the regular, unmodified 2100 gives on the battery level meter.

>This, as I may humbly say, is quite
>stupid. It should monitor the time that has actually passed
>since the charge cycle was started, but NOT by querying the
>actual system time. And, of course, since the actual
>battery voltage can't possible have changed within the
>millisecond it takes to change the date, the battery's
>charge state can't possibly have changed from 12 to 100
>percent.

Well, that is a design decision. They had a real time clock and date
calculation subroutine set available. If someone was dumb enough to
change the clock under them they will get odd results.

I remember that the 100s would fire all of the date reminders between
the first date and the newly set date. The pop ups would go on and on
(I also remember a UNIX box having cron do the same thing back in '83).

>As an indication of truly brilliant and thorough operating
>system design, the battery gauge is still at 12% although
>the text below it says "Charged".

I suspect that it determines "Charged" based on the rapid temperature
rise that NiMH batteries give when they are full, but 12% based on an
ADC voltage with a nominal full charge voltage as a reference. It
sounds like that pack needs a good whack.

>Wow...
>
>This explains a fair amount of weird charging behavior I've
>come across that was obviously due to the fact that I quite
>often set the clock of a fixed Newton or eMate while it is
>charging its battery. Another mystery solved.

I would run into this a lot because my battery doesn't hold a charge,
the supercap drains, the clock resets back to 1996, and I plug it in,
it starts to charge, and I reset the clock. Immediately the battery
is declared bad. I pop it out, wait a few seconds, pop it back in and
it charges.

>I could imagine that the Newton will report the battery to
>be dead if the time is changed while it is in "Waiting"
>mode, i. e. in the mode in which it tries to establish a
>battery condition that it deems safe for charging.

It could be waiting for a very low voltage before it kicks the
charger into fast charge. In which case it would be waiting for a
voltage instead of a dV/dT > X.

>Frank

Andrei
(it's -18C and snowing again, I'm getting sick of this)

-- 
This is the NewtonTalk list - http://www.newtontalk.net/ for all inquiries
Official Newton FAQ: http://www.chuma.org/newton/faq/
WikiWikiNewt for all kinds of articles: http://tools.unna.org/wikiwikinewt/