Wind-up light that does not use a battery to store power?

October 11, 2008 3:16 PM

Would a wind-up light that doesn't rely on charging a battery, but uses a clockwork mechanism to directly power a small generator which lights up an LED, be possible?

However cool wind-up radios and lights are, they all seem to be based on twisting a crank to charge a battery, and then using the battery to operate the device.

Would it be possible to have a light or radio that operated more like a mechanical watch, and did not rely on the intermediate step of charging a battery?

posted by yesno to technology (17 comments total)

It wouldn't violate the laws of physics, but I don't think it would work as well as you think. A spring doesn't really store all that many joules, and a miniature generator wouldn't convert the mechanical energy to electricity very efficiently, especially at low power levels, so it wouldn't run for very long before it ran down -- say, a couple of minutes, or even less.
posted by Class Goat at 3:25 PM on October 11, 2008

Two more efficient non-battery possibilities are a flywheel (as in a dynamo torch) or a capacitor (as in the Faraday Flashlight).
posted by jedicus at 3:40 PM on October 11, 2008

I have one of these flashlights and it seems to work as you want.

Basically, I can spin the crank and give power to a battery. Then, when I need a flashlight, I can just hit the power button. However, if the battery is dead, I can still power the flashlight directly by spinning the crank. (In addition, by cranking the flashlight while it's using battery power, I can get more light from the LEDs than if I just ran it off the batteries.)
posted by 47triple2 at 4:05 PM on October 11, 2008

Gravity lamp.
posted by glibhamdreck at 4:18 PM on October 11, 2008

I think yesno is talking about storing energy in a spring rather than a battery, not about cranking a dynamo when the battery is flat.

According to this article, Trevor Baylis' wind-up radio was powered by a proper clockwork mechanism. So yes, it is possible, but as Class Goat says, not terribly efficient compared with a battery.
posted by le morte de bea arthur at 4:18 PM on October 11, 2008

glibhamdreck: not really.
posted by dmd at 6:39 PM on October 11, 2008

Is this a technical curiosity thing, or are you thinking about building something? If you're building something, why the interested in forgoing batteries?
posted by Kid Charlemagne at 6:48 PM on October 11, 2008

Light takes a lot more energy to make than you think. You'd be cranking a lot, even with LEDs, to get useful levels. To get dim light, you might use a mechanism like that.

There is no such thing as a free lunch. Power is the limiting factor in portability, usually.
posted by FauxScot at 7:20 PM on October 11, 2008

Light takes a lot more energy to make than you think.

Exactly so.

It's difficult to give a mental model of how much energy you exert by winding a spring, so let's use lifting a mass instead. Let's consider how the "gravity lamp" would work if it hadn't been a hoax.

Potential energy = mgy

where m is the mass in kilograms, y is the vertical offset in meters, and g is the acceleration due to gravity (9.8 meters per second per second).

So let's pull some numbers out of an obscene orifice: rather arbitrarily we have a 10 kilogram weight (22 pounds) and lift it 2 meters (about six and a half feet), then use a contraption to let it fall slowly, and harness that mechanical energy to create electricity for a 5 watt flashlight bulb.

10 kg * 2 m * 9.8 m/s² = 196 joules
5 watts is 5 joules per second
So if the system is perfectly efficient (which is impossible) the light lasts 39 seconds before the weight reaches the floor.

And the system just described (10 kilogram mass dropping 2 meters) has a lot more energy than any reasonable spring can hold. I bet a really mondo hand-windable spring probably isn't good for more than 10 joules. For a 5 watt bulb, that's just a couple of seconds.
posted by Class Goat at 9:57 PM on October 11, 2008

Or another way of looking at it: Did you ever have one of those combination light/generators that took power off your bike wheel? That's almost exactly what you're talking about, except it stores the energy in the momentum of the bike. Do you also remember how much harder it was to pedal with it on? you couldn't coast except down hill. And remember what kind of crappy light that gave off?
posted by Ookseer at 2:59 AM on October 12, 2008

I have the first-generation Freeplay Radio that la morte de bea arthur mentioned, and it works just as advertised -- turn a crank, which tensions a constant-force spring, which releases through a little transmission system to turn a generator. I think they traded out the spring for capacitors and batteries a long time ago. Still think mine is pretty cool, though at times the whirring mechanism could be louder than the radio...

One of my students is, right now, trying to build a clockwork wind-up LED lamp. Those above are completely correct in saying how hard it is to store meaningful amounts of energy in these sorts of systems; I think his lamp is going to run for something like 10 seconds after a minute of cranking. It's due in 4 days so I'll let you know then if it works.
posted by range at 6:06 AM on October 12, 2008

Another difficulty of clockwork is that you need to build a fairly beefy geartrain if you want the power to be stored, and able to be switched on at will. The reason the original FreePlay radios had to be cranked as needed (yep, another first gen owner here) was due to Bayliss's discovery that the nylon gears would deform if left under tension for any length of time.

Ookseer, bike dynamos high drag? You obviously haven't used a good one.
posted by scruss at 6:39 AM on October 12, 2008

And although it feels sometimes like I'm trying to set a record for posting the same Wikipedia page over and over again, check out the table over at the Energy Density page -- springs lose to the worst batteries by three orders of magnitude, meaning that (roughly) your .8oz AA battery needs to get replaced by 800oz (50 pounds!)* of spring.

*Or, for you pesky other 95% of the world, 23g to 23kg.
posted by range at 7:25 AM on October 12, 2008

Radios are a different bag than lights. Especially if they play through earphones, they don't use that much energy.

When I was a kid I had an AM radio with an earphone that didn't require batteries at all. It was powered completely by the energy of the RF it was receiving.

The operating power (from the RF) was probably only a few milliwatts. But that was all that was needed to produce the teeny tiny amount of sound in the earphone.

If you use five milliwatts to power an LED, you can see the LED. In fact, it will be reasonably bright. But it can't be used for illumination. You'd need to go up at least 3 orders of magnitude, and at that point you're beginning to talk about significant power consumption, way beyond anything a spring reasonably could support.
posted by Class Goat at 11:52 AM on October 12, 2008

Kid Charlemagne, just curiosity.

Thank you, everyone, for the great answers.
posted by yesno at 4:04 PM on October 12, 2008

Ookseer, bike dynamos high drag? You obviously haven't used a good one.
You're right scruss, the last time I used one was probably 1978, but the experience tainted me for life on electrical pedal power.

Thanks for the excellent Link, range. First time I've seen it.
posted by Ookseer at 7:10 PM on October 12, 2008

(Update: My student's project worked. Clockworks from a toy powered one red LED for 5-10 seconds per windup. Not exactly the reading lamp he was thinking of at the start, but a good demo nonetheless.)
posted by range at 2:44 PM on October 18, 2008