Alt-BEAM Archive
Message #02078
To: Bob Shannon bshannon@tiac.net
From: Steven Bolt sbolt@xs4all.nl
Date: Wed, 7 Apr 1999 10:51:08 +0200 (CEST)
Subject: [alt-beam] Re: Microcore memory for 100Farad caps.
On Wed, 7 Apr 1999, Bob Shannon wrote:
---8<---
> If you want to drive a SE, expecially a cmos logic SE, the small VL2320
> or similar cells (with a series resistor) are a good choice. If you
> want current for a walker, its hard to go wrong with small packs of
> Varta Ni-MHD cells and a solar constant current charger. You can build
> a small charger and battery pack is less space than a 1.5 farad gold
> series capacitor.
But capacitors do have some advantages. They don't care about slow
or fast charge, as long as the voltage stays below max. A solar
constant current charger would have to be limited to about 0.05C,
so a full charge would take days where a cap might recharge in
minutes. Moreover, you get an almost unlimited number of charge
cycles from a cap; chemical rechargeables lose significant capacity
after just a few hundred cycles.
The slowly dropping voltage of a cap during discharge doesn't have
to be much of a problem. For 74HC and similar logic, you might
start at 2.5V and drop to less than 1V, using more than 80% of
available energy and most of it at very reasonable Icc.
Parallax has a nice Stamp application note demonstrating outdoor
temperature measurement and an infrared link to relay the data to
an indoor PC every 12 seconds. The Stamp, a TLC555 and the infrared
emitter are powered by a small solar panel with a 1F 5.5V cap as
backup. The panel provides 9mA @ 10V in direct sunlight, the
Stamp's voltage regulator nicely avoids overcharging the cap.
The result:
We ran the project `round the clock for several days, periodically
reviewing the time-tagged data files for breaks or erratic data
that would indicate power failure. None occurred. The lowest
voltage across the super cap, which occurred after about 10 hours
of darkness, was 3.65 volts, just enough to keep the Stamp going.
Less than an hour after sunrise the cap would charge back up to 5V.
One could do better with an AT90S1200 (which happily runs at 1.8V)
or with certain PICs, and this particular application could also be
implemented in 74HC hardwired logic.
I don't think a walker should use a cap for energy storage. Typical
machine walking is too inefficient, making weight a primary
concern, and you wan't a fairly stable voltage for that application.
But a `bot on wheels on a smooth surface is something else, and for
many cmos logic projects, a cap and solar panel do very well indeed.
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