Dual Gravity Power Generator

[simonmagus]

This idea was inspired by the gravity lamp:
http://www.vtnews.vt.edu/story.php?relyear=2008&itemno=111

Its principle is basically a weight on a ball-screw falls while turning & pushing the rails. The rails are connected to the rotors which power the 2 generators. Rather than removing the weight to restart the process, a mount can be used to flip the device over like an hour glass. It is a simple design. Any suggestions?   

[hartiberlin]

Interesting idea.
The hour glass flip over reminds me of the idea Mr. Gr?ter Blasius had with his
gravity water device which did not work and was based on the hydrostatic paradoxon.

To flip it over you would need to raise the weight in the gravity field
to the old height, which would need again more energy than could
be won when the weight falls.
So how should this be overunity ?

[Koen1]

Well I also don't really see how it can be OU,
but it is an interesting variation of the gravity lamp.

When I read the article about the gravity lamp, at first I thought
this was its priciple already: that one could just turn the lamp upside down
to make the weight start its slow journey downward over the what was it like
4-hour period?
I thought "hey that's kind of a neat invention, especially for people in
developing countries or way in the outback or on camping trips etc"...
But as it turns out the gravity lamp does not yet have this feature,
and one needs to manually take the weight and lift it back to the top,
then let go and it will slowly slide down again.

Funny eh, that we see that lamp and immediately think of making it
into a two-way version so we can turn it upside down to keep it going?

[zerotensor]

The gravity lamp you linked-to was a design concept, and was never actually built.  It turns out that the inventor did a gross miscalculation, and the performance of the device was off by a couple of orders of magnitude.  The claimed 40W equivalent for 20 minutes was WAY off.  The inventor conceded the prize.

In the case of the device you propose, let's assume a weight of 1kg and a height of 1m.

The gravitational potential energy of 1kg at 1m is E = m g h = 1*9.8*1 = about 10 Joules (say).
10 Joules is enough to:
light a 1W LED for 10 seconds.
A 100 mW will go for about 100sec.
A 10 mW, (which does not put out much light), will go for 1000 seconds, which is about 20 minutes.

For a practical design with a 1kg mass, I'd go with the 100mW range.  It will give you about a minute and a half of light sufficient for reading per flip.  Keeping the beam focused would be advisable.  Of course the output power, or the on-time, if you prefer, will scale linearly with the weight, so a 10kg weight will give you 10x the light of a 1kg model.

So maybe even more practical would be a 10kg unit at 500 mW, which should last about a minute.

These calculations are for perfect efficiency, which of course will not happen.  Consider these values obtained as upper limits on performance.  A good dynamo will net you 50% COP on this scale, I think (guessing here).  BTW, there are some nice low-friction dynamos that are being built into bike hubs nowadays.  If you intend to make a prototype, you might want to consider picking one up.