Alt-BEAM Archive

Message #01055



To: beam@corp.sgi.com
From: Steven Bolt sbolt@xs4all.nl
Date: Sat, 27 Feb 1999 10:47:43 +0100 (CET)
Subject: [alt-beam] Re: ARG, stupid 'vore! + V vs A (fwd)


On Fri, 26 Feb 1999 TurtleTek@aol.com wrote:

> I'm sorry, I meant 24x22. I don't use the "BP-xxxxxx" to identify
> them because I've been too lazy to memorize them.

That is a 6-cell panel, but the cells are too small for SE
applications in anything but the best light. That's why the 5-cell
24x33 is considered much better.

> It triggers between 3.2v-3.4v, usually. And yes, I'm pretty sure
> I'm using the J.

That sounds like an L. You might test it out of circuit to be sure,
if the markings are unclear.

> > That one [37x33] consists of 8 cells (each the same size as
> > those on the 5-cell 24x33). In the same light, it gives you
> > the same current at a 60% higher voltage.
>
> Oh, I thought it gave more (although a small amount) because the
> Solarbotics site advertises it as such.

It's best to associate the number of cells with voltage, and their
size with current. But note that the 37x33 doesn't merely have more
cells; compared to the 24x33, the active panel area is 60% larger.
That makes a difference, even if the number of cells doesn't match
your trigger voltage and light intensity too well. What I said is
perhaps incomplete:

"Only if the voltage is sufficient in the light conditions you
design for. Once you have the voltage, increasing the number of
cells does not make your SE charge faster, but increasing cell size
does."
Both of course increase the panel area. It's just that if you have
a choice, then at a switch-on level of 2V or a little more, 5 large
cells are preferrable to 8 small cells, given the same panel area.

And we now know that your switch-on level is much higher, actually
about 60% higher than what I've found to be best for SunEater_III.
It seems to me that you are likely to waste *much* energy in the
first fraction of a second of discharge, when the motor accelerates
too fast to take the `bot with it. Giving it just enough for a firm
`step' is best, I think.
But if you really want so high a switch-on, an 8 cell panel is
actually needed. 5 or 6 cells will easily reach 3.4V only in direct
sunlight, even when unloaded.

> argg....I don't understand the concept of voltage, and
> "electrical pressure" means nothing to me. At least I understand
> current. Good thing I'm taking that physics course next year,
> perhaps that'll help me.

It's all about power, really, as that is what motors convert into
motion. Power is watts = voltage x current. Your motors can't eat
power in just any combination; to get the best motion `kick' out of
your storage cap, you have to charge the right cap to the right
voltage, then switch it using the right transistor, able to deliver
the current demanded by the motor.

The reasoning behind the number of cells has to do with performance
in good and bad light. As Justin pointed out, Solar Engines are
about converting light to motion - light as opposed to darkness.
It's all very well to want some performance in bad light, but
preferrably not at the cost of wasting good light!
When the light intensity increases, a solar panel consisting of a
too large number of cells wil try to increase the voltage. But the
switch-on level of the SE has to be fixed, to meet the appetite of
the motors. So extra voltage can't be used.
A panel with the same area and the right number of cells will
increase current, which we *can* use. The storage cap will be
charged faster. At some point the motors may run almost
continuously, even on a 24x33 panel.

In short, having the right number of cells means your `bot will
become much more active when it finds good light. If the cells are
too many and too small, your critter won't care so much if it finds
the sun. It will do marginally better in bad light, but that is not
as interesting.

> BTW, Steve, I must say that I am one of your "biggest fans", by
> that I mean that I look forward to your posts and find them very
> straight forward and informative. Keep it up!

Thanks; I hope to avoid becoming a pedantic old know-it-all...
I certainly can't know everything, so make sure you confirm the
interesting bits by doing good experiments.

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