Alt-BEAM Archive
Message #01645
To: "Feser, Jason" jfeser@ea.com
From: Steven Bolt sbolt@xs4all.nl
Date: Sat, 20 Mar 1999 09:52:37 +0100 (CET)
Subject: [alt-beam] Re: Solar charger for NI-CAD walker...
On Fri, 19 Mar 1999, Feser, Jason wrote:
> I was going through my 1000 Circuits book from none other than Forrest
> Mims and found a circuit for a solar charger. I was thinking that I could
> implement a solarize my walker by trickle charging a NICAD with a solar
> cell.
> My question is this - can you be charging a nicad and drawing from it
> at the same time - I think that NICADS have memory, so would this damage the
> battery's ability to store its maximum power level?
Trickle charging nicads is bad because it promotes the growth of
large cadmium crystals, which leads to higher internal resistance
and lower capacity. A moderately fast charge/discharge cycle gives
you the best performance. Electromotors and `bot construction being
what they are, the discharge is likely to be fast enough. Good
chargers use a current of about C (500mA for 500mAh nicad) or a
little less and switch off when the cell voltage stops rising/
starts dropping, or when the temperature starts to rise.
Sealed lead/acid batteries are stil the best if you want to
trickle charge, but they won't stand high current discharges.
That said, it is possible to trickle charge a nicad, and partially
use its capacity as you suggest, if you properly discharge and
charge it once in a while. `Proper' means a discharge to 1V per
cell under a light load (about 0.1C), and a fairly fast charge
using the above voltage drop or temperature rise trigger to stop.
These days there are good chargers on a chip to achieve that, and a
proper discharger isn't difficult to make either.
So what does "once in a while" mean? That depends on the career of
the `bot. Perhaps once every couple of weeks of continuous use.
> Also, the circuit I have doesn't seem to have any protection
> against overcharging - anyone know how to implement this?
For sealed lead acid it is a simple matter of limiting the voltage.
There is no solution for nicads, unless you have a constant current
available so you can implement a voltage drop or temperature rise
trigger to stop the charging. Solar panels won't allow that, So you
have to limit the charge current to something the nicad can stand
almost indefinitely without suffering permanent damage. That limits
you to 0.05C, which means that a solar charge will be very slow,
taking a couple of days to achieve a full charge.
> I was also thinking that when the power level of the NICAD got below a
> certain threshold, say about 3.5v, it might at that point use a IR recieve
> to get a signal from a IR transmit mounted to a wall wart and then head the
> walker in that direction. Then when you saw your bot scrabbling against the
> wall near the trasmitter, you would know that it was time to juice up. Is
> this a do-able thing? Would it require the use of a microprocessor?
One of my old `bots (http://www.xs4all.nl/~sbolt/e-andere_robots.html#willie)
hunted all light sources, in the hope of finding the 15W lamp which
marked its `feeding tray'. Everything about it, including its
procedure to work itself around obstacles, was aimed at maintaining
`lock' on a light source. To avoid fatal hang-ups, it measured
its collision frequency. If its feelers touched more than seven
times within 40 seconds, it closed its eyes for a while, in order to
blindly wander away from the false light. There was no uC in its
brain - the right kind wasn'tavailable in 1986 - but using a uC would
reduce the parts count.
Using IR doesn't mean much, as a lamp is in fact mostly IR. You'd
have to modulate the light to make it special. And then there might
still be reflections to distract the `bot, or glass doors between
it and its target. The collision frequency count and random walk
method proved quite successful. My Spider will use sound to find
its feeder, because that makes it possible work without a direct
line of sight (and because I find it more interesting :)
> I've never played around with an IR setup before so I'm not too
> sure if this is something that is easily done.
Light source tracking and IR are easy, and there is plenty info
about it on the Web. Perhaps the IR `radar' of my Spider will give
you some ideas (http://www.xs4all.nl/~sbolt/e-spider_IRbumper.html).
Best,
Steve
----------------------------------------------------------------------
# sbolt@xs4all.nl # Steven Bolt # popular science monthly KIJK #
----------------------------------------------------------------------
------------------------------------------------------------------------
eGroup home: http://www.eGroups.com/list/alt-beam
Free Web-based e-mail groups by eGroups.com
Home