Comment by pdonis | Hacker News Reader

Comment by pdonis | Hacker News Reader

Parent Full thread

There's an interesting comment about super-efficient LED bulbs in that museum of hoaxes article. However, the numbers look marginal for this GravityLight even with super-efficient LED bulbs.

Quick back of the envelope calculation, using what few numbers are given in the article.

Ballast: m = 10 kg (1)

Height: h = 3 m (2)

Energy stored: E = mgh = 300 J

Time: 30 min = t = 1800 s (3)

Wattage available: P = E / t = 0.17 W (4)

Notes:

(1) Article says you can hang anything weighing about 20 lbs.

(2) No numbers given in the article, but the pictures make it look like ceiling height (8 to 10 ft., I took the larger)

(3) Article says light for half an hour.

(4) With current LEDs this is the equivalent of about a 2 W incandescent bulb, i.e., pretty dim. The light in the pictures in the article looks like the equivalent of about a 40 W incandescent bulb, so the numbers come up short by a factor of about 20. That would indeed be "super-efficient" for an LED; I'm not aware of any even on the drawing boards that are that efficient.

[Edit: The numbers actually are not too low compared to kerosene lamps, which are what this light is supposed to replace. See exchange downthread with xd.]

Are you aware of how little light you need to see? Just because you assume the light in the video is similar to a 40w bulb is short sighted (no pun intended). A 2W "pretty dim" incandescent bulb will illuminate a room no problem.

Human eyes can adjust very well to low light levels.

It depends on what you're trying to do. For just minimal seeing, yes. For reading, or many kinds of tasks, not so much. There's not enough in the article to know for sure how they're pitching it to the people that might actually use it.

I once done a stint in the military, and during training we had to use red light at very low levels to read and write at night time .. it is possible, just a world away from what the average human with access to electricity on tap is used to, I guess.

Yes, you're right, I remember similar things from my time in the Navy. This kind of lighting regime is normal in the combat spaces of a ship, even in the daytime.

The GravityLight article does say that it's meant to replace kerosene lamps, which according to Wikipedia range from 20 to 100 lumens:

http://en.wikipedia.org/wiki/Kerosene_lamp

This is about the equivalent of a 1.6 W to 8 W incandescent bulb, so you're right, I was too quick to dismiss the numbers as being too low. They're certainly low compared to what first world citizens are used to, but they are roughly equivalent to what the target users are used to.

It's a shame that few, if anyone else will get to this level of understanding .. instead just branding it a hoax, which sucks.

Considering the goal is 55k, and that the basic principle is sound and surely can be improved in efficiency, I have to say reading this discussion and the jaded way people just dismiss it wholesale as "hoax", and not give a second thought is kinda shameful indeed.

Kickstarter is overflowing with game projects raking in the millions collectively with unverifiable claims altogether, like "awesome multiplayer experience". But this stuff? "It's a hoax, they're trying to rip the public off for 55k!" You simply cannot make this shit up.

The light seen in the video is not the 30 minute-one. There the weight drops much quicker. Look carefuly at about 55 second in the video.

Well, shortening the time by a factor of 20 makes it a minute and a half of light equivalent to a 40 W incandescent bulb. So maybe this is really a combination light and exercise machine? :-)

Line three: you write

   Energy stored E = mgh (300 J)

but by substitution

   Energy stored E = 10 (kg) (9.8 m/sec) 3 (meters)
                     2940 J

2940/1800 is 1.6W

Check your math. I used 10 instead of 9.8 for g since I was just doing a "back of the envelope" calculation. 10 x 10 x 3 = 300.

sigh, you are absolutely correct. The rounding up messed with my brain I saw the 300 and thought "oh that is just 10 * 3 (brain fart) he didn't add in the 9.8"

AND I have to correct mine too since I used 180 for time (30 * 6 rather than 30 * 60)

Your math assumes you have a magical 100% efficient generator to convert potential energy to electrical.

In reality, you will be lucky to get 50% back, and that's with a very good generator.

Electric generators can get up to 95% efficiency, meaning electrical output as a percentage of mechanical energy input. In the current case, the relevant input is mechanical energy produced by a slowly dropping weight, so there would be losses from friction. But overall, I'd expect much better than 50% in the conversion of gravitational potential energy to electrical energy.

However, in commercial power generation, the mechanical energy usually comes from some sort of heat engine, the heat for which is provided by fossil fuel combustion or nuclear fission. There is a big haircut in that step. Heat engines are generally only 35 to 60 percent efficient.

Hydro power does much better, but it doesn't use a heat engine; it's just a much larger scale example of converting gravitational potential energy into electricity.

I think the reason this contraption is valuable is that LEDs don't need much power to produce enough light to please someone used to kerosene lamps.

Link to a generator with 95% efficiency please.

> I think the reason this contraption is valuable is that LEDs don't need much power to produce enough light

Actually, that contraption cannot produce enough light, unless you have a crazy weight, a lot of elevation, and a magical efficient system that converts energy to electricity, and then transforms it to whatever voltage/amperage appropriate for you LED lights.

As posts upthread have shown, the light output will be comparable to a kerosene lamp, which is what these lights are supposed to replace. It doesn't take a "crazy weight" or "a lot of elevation". And the difference between 50% efficiency and 100% efficiency is only a factor of two; that's well within the range of making reasonable adjustments in the weight and/or height.

As far as voltage/amperage is concerned, I would expect the generator to be low voltage DC, matched to some voltage in the range the LED light could support. AFAIK LED lights are fairly tolerant of a range of low DC voltages, so I don't see this as a major issue.

I've not seen a generator (quite likely 'cause I've never gone looking), but I've seen model plane motors claiming almost 95% efficiency (a quick hunt found this one claiming 94% max efficiency: http://www.astroflight.com/index.php?main_page=product_info&... )

I'm guessing that's pretty reversible - if you spun that motor backwards with the right speed and torque and loaded the output up in just the right fashion, I think you'd get 94% of the energy you put in out as electricity.

The problem I see is that a slowly falling weight is unlikely to provide "the right speed and torque" without some sort of lossy gearbox in between - there's the physical equivalent of an impedance mismatch there. If I had to imagine a motor/generator that'd be likely to work on a direct-drive to a falling weight it'd probably be several feet in diameter. (For model plane motors, the diameter of the motor is a significant factor in the kV constant and the rpm at max efficiency. Small diameter motor spin fast, large diameter motors spin slow.)

You're right. I did say it was a back of the envelope calculation. :-)

If you have a 50% efficient generator, just double the size of the ballast; 20kg is still reasonable. Or double the height, though that would be somewhat harder; you would need some kind of pulley arrangement. Or some combination of the two.

Also, I think low voltage DC generators can do considerably better than 50%; the numbers for this kind of application can be quite different than those for the kind of large scale AC generators we're used to.

Reply on news.ycombinator.com