Hacker News new | past | comments | ask | show | jobs | submit login

"The download of six megabytes of data collected during the six-hour transit, from above Jupiter’s north pole to below its south pole, took one-and-a-half days."

...surely that's a typo and they mean gigabytes?




The solar panels that Juno has produce only 500 watts of power. You would need a lot more then that to be sending gigabytes of data in any meaningful timeframe. So yes it really is just six megabytes of data taking 30h+ to download. Btw the panels would produce 14kW at earth orbit but suns power goes down quite fast as you move away from it. Also the farthest from the Sun we've ever used solar panels. Traditionally missions this far from the Sun have used some kind of a nuclear power source (RTGs). Though I don't think they could produce 500 watts without being very big.

http://www.jpl.nasa.gov/news/news.php?feature=4818


Transmission is one thing, but collection is another: did they really only collect 6 megabytes of data over the course of a 6 hour flyby?

Off topic, ish: does anyone know how large a solar panel array would have to be to supply all of our current energy use, if the array orbited the Sun at the same distance Mercury does?


According to Wikipedia, the world's average energy consumption is 12.3Tw. Given that the Earth is gets 1366 W/m² of solar power at 149,597,870km, this means Mercury gets 9190 W/m² at 57,910,000km. Plugging these together, this tells us that a 12.3Tw solar array of 100% efficiency in Mercury's orbit would have a surface area of 1338 km².

Realistically, you'd at least double that, to produce the same power, as solar panels are <50% efficient.


Nitpick, the farthest distance of Mercury from the Sun on its orbit is closer to 0.46 AU and it spends most of the time there, so for safe estimate it is better to take this greater value for planet-Sun distance. I get around 100x100 km collecting square near the Earth and almost 50x50 km near the Mercury. Taking into account efficiency and design constraints, it's safer to estimate more like 200x200 km near the Earth and 100x100 km near the Mercury. Clearly, building this near the Mercury brings little benefit as opposed to building it near the Earth. If we built this on the surface of the Earth (seems cheapest, even if the area would have to be quadrupled due to night and atmosphere) it does look practically achievable in the future. Solar energy could sustain current consumption of the human civilization. Most alternative sources of energy are not rich enough to supplant nuclear energy. Solar energy seems to be the best contender.


Brilliant, thanks. So how long before we start putting solar panels in close, direct orbit around the Sun?


Then we could beam the energy back to earth using lasers so bright that they would burn plasma holes through our atmosphere and precisely target receiving stations where they would vaporize water which would power a steam turbine. Totally doable, I'd say we could get it up and running in a couple years.


This sounds like sarcasm, but yes: solar collectors in close orbit around the sun, huge flying parabolic mirrors focusing solar energy that's already 7-10x more dense than it is at 1 AU, beaming that energy to receivers sitting in orbit around the Earth (perhaps at Lagrange 1 and, via L4/L5, L2 to reach the night side of the planet), relayed in to near earth orbit satellites, and being beamed down to many surface locations at safe intensities that might well be used to boil ocean water, driving steam turbines and producing fresh water (and salt).

Such a system could provide energy dense power to the entire solar system (skipping the steam turbine step for most locations, and going pure solar electric), up to large installations for entire planets, and down to individual spaceships.

I'm not suggesting we'll do it tomorrow. But within a hundred years, if we get and keep our shit together..?


Short answer: you can't. The Earth's rotation would make collecting that energy very unwieldy. Individual collectors could act like capacitors, but getting that power down to the planet continuously isn't really doable with any current technology.


You wouldn't beam energy directly from near-solar orbit to the surface of the Earth; you would relay it to orbiting receivers around the planet, which would in turn beam it to the surface, and reaching the night-side as well as day-side.


As soon as somebody figures out how to get the energy from the panels back to earth? Assuming we figure out how to economically put them there in the first place...


About the same time we reach Type II Civilization status and can build a Dyson Sphere!

https://en.wikipedia.org/wiki/Kardashev_scale#Type.C2.A0II_c...


The cost would be astronomical, I do not think it would be worth it. Much better to build this here on Earth, only 4x more panels needed, still easier to build and maintain.


You'd have to add some redundency for the rotation of the planet. 58 earth days per mercurial day.


You don't need to be concerned about Mercury's day unless they're on the surface, essentially. You do need to be concerned with the earth's, and their positions relative to the sun.


In the era of big hard drives and fat pipes, people forget that 6 megabytes is a lot of raw data if it's not packed with unnecessary GUI elements, high-res video, and inefficient code.


On the other hand, 6MB is the typical file size for the jpg-compressed images taken by my 20MP smartphone camera.

It is all a matter of perspective.



That's pretty much packed with a lot of useless information, considering the average usage of uploading the picture to Facebook. :P


Almost all of the instruments on Juno are non-visual. The camera was included mostly for PR purposes. 6MB is pretty good for one-dimensional data streams.


In stellar terms, tiny. In human terms, huge.

Googling suggests that global energy usage is estimated to be around 15 terawatts.

Using the most efficient multi-junction solar cells available - which can convert 30% of the light hitting them into electricity - we could generate that with about 6.5 million km² of solar panels: that's about 80% of Australia.

Out at Earth's orbit, it'd need to be 68 million km^2: about the area of the Indian Ocean.


I only get (15E12/1360/0.3) = 37 billion m², which is 190x190 km square or 36000 km². Little more than Taiwan, which looks as quite a little spot compared to the globe: https://en.wikipedia.org/wiki/Geography_of_Taiwan


When I was looking into Forward Error Correction a while back all of the impressive stuff was for downlinks from remote space probes.

I don't know any specifics around the Juno probe however most of the techniques usually sent 2-3x the amount of data actually downloaded because a missed frame was 1.5hrs(43mins @ Speed of light from Earth->Jupiter). Being able to reconstruct a frame with redundant data was critical.

Combined with a low power transmitter 6mb is pretty reasonable.


Nope, it just takes this long for the data to travel 588 million kilometres [1]. Also the space craft will have a low powered transmitter.

[1] http://www.space.com/18383-how-far-away-is-jupiter.html


Why "surely"? One of the main things about the Juno probe is how underpowered its tech is.


Google Fiber really needs to step up the interplanetary deployment!


Imagine you need to ping google.com from mars. The round trip would be at least 40 minutes. Put a whole new meaning into the word lag.

Seriously though, the current TCP/IP protocol is not designed to deal with that kind of distance.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: