4 ms·

> If they place them at the 1200km altitude, it's not clear at all why this would be bad, except that some self-declared experts pointed out that the current GP

> If they place them at the 1200km altitude, it's not clear at all why this would be bad, except that some self-declared experts pointed out that the current GPS satellites are at a higher altitude. The main downside of this is that you typically need more satellites. If I recall, you need a minimum of 4 satellites to get a position, and more is beneficial. With modern atomic clocks and other trappings, that could get expensive.

credit_guy 6y ago
Actually, modern atomic clocks should be less expensive, not more expensive than the ones produced 30 years ago that power the original GPS satellites. For example, chip-scale atomic clocks [1] can be had for about $1000 [2]. Here's a space rated such atomic clock [3]. I couldn't find the cost, but let's say it's one hundred times more expensive than a "civilian" version. It's still negligible even compared to the cost of putting the satellite in orbit only. Speaking of which, this calculator [4] shows me that with the same rocket, going from an altitude of 1000km to 20000km cuts the cargo mass roughly in half. And another difference between 20000km and 1000km is the fact that radio signals decay with the square of the distance. At 20000 km you need an antenna that's 400 times more powerful than at 1000km. That means you need correspondingly more solar panels and more radiators. Overall, the satellite will be many, many times heavier. For example the GPS satellite USA-132 currently in service has a mass of 2 tons [5]. If the lower orbit affords you a mass reduction not by a factor of 400, but only by a factor of 10, you end up with 200 kg. The current OneWeb satellites weigh about 150kg. [1] https://en.wikipedia.org/wiki/Chip-scale_atomic_clock https://en.wikipedia.org/wiki/Chip-scale_atomic_clock [2] https://www.electronicdesign.com/technologies/analog/article/21806867/cant-afford-an-atomic-clock-get-a-molecular-one https://www.electronicdesign.com/technologies/analog/article... [3] https://www.microsemi.com/product-directory/embedded-clocks-frequency-references/5207-space-csac#overview https://www.microsemi.com/product-directory/embedded-clocks-... [4] https://www.satsig.net/orbit-research/delta-v-geo-injection-calculator.htm https://www.satsig.net/orbit-research/delta-v-geo-injection-... [5] https://en.wikipedia.org/wiki/USA-132 https://en.wikipedia.org/wiki/USA-132
- nordsieck 6y ago
  OK. But you're still not addressing the main problem. LEO means you meed more satellites because they're closer to the earth. If look at the proposed OneWeb satellites, they need 650 of them for their constellation. I'm sure there's some overlap there, but probably not a ton. With GPS, you need much more overlap - specifically, every part of the earth need to be able to get a signal from a minimum of 4 satellites. So you might be looking at somewhere between 650 and 2600 satellites in LEO to run a GPS type system instead of 24 at the traditional orbit. I just don't see how that makes sense.
  - credit_guy 6y ago
    The number doesn’t seem to be a problem. Specifically, per wikipedia [1]: “Despite the bankruptcy process in the end of May 2020, OneWeb filed an application to FCC for increasing the number of satellites to 48,000” [1] https://en.wikipedia.org/wiki/OneWeb_satellite_constellation https://en.wikipedia.org/wiki/OneWeb_satellite_constellation
    - nordsieck 6y ago
      > “Despite the bankruptcy process in the end of May 2020, OneWeb filed an application to FCC for increasing the number of satellites to 48,000” It's pretty easy to see that this isn't a real plan; that is to say, if OneWeb weren't bankrupt, they could never accomplish this. The math is pretty simple. If we assume that their satellites have a lifespan 2x as long as SpaceX (10 years) and they continue to launch on Soyuz (34 / launch), then they'd need 48000 / 10 / 34 = 141 launches per year for the life of the company. I hope it's obvious to you that that simply would never happen. Even if OneWeb could afford to buy that many launches, it's doubtful that Russia could manufacture and launch them at that pace. Even SpaceX (who is planning a similar sized constellation) can't launch their lower cost/kg F9 twice per month, and they have the benefit of reusing the 1st stage and fairing halves. > The number doesn’t seem to be a problem. The number matters if the UK cares about money.
      - credit_guy 6y ago

credit_guy 6y ago

Actually, modern atomic clocks should be less expensive, not more expensive than the ones produced 30 years ago that power the original GPS satellites. For example, chip-scale atomic clocks [1] can be had for about $1000 [2]. Here's a space rated such atomic clock [3]. I couldn't find the cost, but let's say it's one hundred times more expensive than a "civilian" version. It's still negligible even compared to the cost of putting the satellite in orbit only. Speaking of which, this calculator [4] shows me that with the same rocket, going from an altitude of 1000km to 20000km cuts the cargo mass roughly in half. And another difference between 20000km and 1000km is the fact that radio signals decay with the square of the distance. At 20000 km you need an antenna that's 400 times more powerful than at 1000km. That means you need correspondingly more solar panels and more radiators. Overall, the satellite will be many, many times heavier. For example the GPS satellite USA-132 currently in service has a mass of 2 tons [5]. If the lower orbit affords you a mass reduction not by a factor of 400, but only by a factor of 10, you end up with 200 kg. The current OneWeb satellites weigh about 150kg. [1] https://en.wikipedia.org/wiki/Chip-scale_atomic_clock https://en.wikipedia.org/wiki/Chip-scale_atomic_clock [2] https://www.electronicdesign.com/technologies/analog/article/21806867/cant-afford-an-atomic-clock-get-a-molecular-one https://www.electronicdesign.com/technologies/analog/article... [3] https://www.microsemi.com/product-directory/embedded-clocks-frequency-references/5207-space-csac#overview https://www.microsemi.com/product-directory/embedded-clocks-... [4] https://www.satsig.net/orbit-research/delta-v-geo-injection-calculator.htm https://www.satsig.net/orbit-research/delta-v-geo-injection-... [5] https://en.wikipedia.org/wiki/USA-132 https://en.wikipedia.org/wiki/USA-132

nordsieck 6y ago
OK. But you're still not addressing the main problem. LEO means you meed more satellites because they're closer to the earth. If look at the proposed OneWeb satellites, they need 650 of them for their constellation. I'm sure there's some overlap there, but probably not a ton. With GPS, you need much more overlap - specifically, every part of the earth need to be able to get a signal from a minimum of 4 satellites. So you might be looking at somewhere between 650 and 2600 satellites in LEO to run a GPS type system instead of 24 at the traditional orbit. I just don't see how that makes sense.
- credit_guy 6y ago
  The number doesn’t seem to be a problem. Specifically, per wikipedia [1]: “Despite the bankruptcy process in the end of May 2020, OneWeb filed an application to FCC for increasing the number of satellites to 48,000” [1] https://en.wikipedia.org/wiki/OneWeb_satellite_constellation https://en.wikipedia.org/wiki/OneWeb_satellite_constellation
  - nordsieck 6y ago
    > “Despite the bankruptcy process in the end of May 2020, OneWeb filed an application to FCC for increasing the number of satellites to 48,000” It's pretty easy to see that this isn't a real plan; that is to say, if OneWeb weren't bankrupt, they could never accomplish this. The math is pretty simple. If we assume that their satellites have a lifespan 2x as long as SpaceX (10 years) and they continue to launch on Soyuz (34 / launch), then they'd need 48000 / 10 / 34 = 141 launches per year for the life of the company. I hope it's obvious to you that that simply would never happen. Even if OneWeb could afford to buy that many launches, it's doubtful that Russia could manufacture and launch them at that pace. Even SpaceX (who is planning a similar sized constellation) can't launch their lower cost/kg F9 twice per month, and they have the benefit of reusing the 1st stage and fairing halves. > The number doesn’t seem to be a problem. The number matters if the UK cares about money.
    - credit_guy 6y ago