Starlink : General Discussion - Thread 1

[Razvan]

There's a team of bright engineers behind Virgin Galactic. I don't like the hate that's spewed towards the company just because people don't like Branson's personality.

I did not mean either hate, or I any disrespect towards Virgin Galactic, by the contrary I was quite enthused about their project with the White Knight and its Spaceship.
It is the fact that, in spite of the good feeling that many shared, the Project did not come to fruition even after prolonged delays. It could be just the result of inadequate budget...

[Robotbeat]

It hasn't disappeared, either. And frankly, something like Virgin Galactic or XCOR is the best chance we have at visiting space.

[Razvan]

http://www.geekwire.com/2015/photos-spacex-founder-elon-musk-unveils-new-10b-space-internet-plan-private-seattle-event/

It seems that Elon has decided to go alone in this new venture covering its triple main components: (1) designing and manufacturing satellites, (2) launching and (3) operating sat network. If the first two were obvious the third one appeared plausible after his friend Greg Wyler went the separate way along with Qualcomm and Virgin Group. It seems that Virgin Group will be in charge of launching the sats which, based on the recent long-effort-no-results of Sir Richard Branson's Space Tourism, I personally doubt very much will be a success. It remains tbs if any truth in "OneWeb backers questioning whether Musk will be able to get rights to the spectrum required to build the SpaceX network."

"Musk has not yet determined where to put the satellite production factory — an unclaimed prize for which Washington and other states will no doubt compete fiercely."
Imo, that place will be in Texas on the same land Elon builds the new Launching Pad. It looks like Texas is gonna be the winner again.

Think Musk will see what each state will be offering. IMO it is more likely to be in either South California, Seattle or Denver. There are existing facilities and human resources concentrated in those areas. There is about zero infrastructure at Boca Chica. Even water have to be trucked in.

Of course Musk will do his own judging. It just made sense to me to put the Plant right by the launching pad. It could also be Brownsville, as @docmordrid stated, and very important the friendly policy Texas state is showing to SpaceX and all other investors of this magnitude.

[MikeAtkinson]

Can Brownsville be used to launch into the satellites' orbits? To cover most of the earth they would need a high inclination.

[Razvan]

It hasn't disappeared, either. And frankly, something like Virgin Galactic or XCOR is the best chance we have at visiting space.

I agree. Unfortunately, it is not enough to have a great idea and brilliant engineers, you should also have good management, too. Or, in our case, to launch a large fleet of satellites in a very short period of time they would need a lot more powerful vehicle, hence more complex problem to deal with.

[Callezetter]

Just posting Elons Seattle video from yesterday if anyone missed it. Mostly about Satellites etc

[TrevorMonty]

http://www.geekwire.com/2015/photos-spacex-founder-elon-musk-unveils-new-10b-space-internet-plan-private-seattle-event/

It seems that Elon has decided to go alone in this new venture covering its triple main components: (1) designing and manufacturing satellites, (2) launching and (3) operating sat network. If the first two were obvious the third one appeared plausible after his friend Greg Wyler went the separate way along with Qualcomm and Virgin Group. It seems that Virgin Group will be in charge of launching the sats which, based on the recent long-effort-no-results of Sir Richard Branson's Space Tourism, I personally doubt very much will be a success. It remains tbs if any truth in "OneWeb backers questioning whether Musk will be able to get rights to the spectrum required to build the SpaceX network."

"Musk has not yet determined where to put the satellite production factory — an unclaimed prize for which Washington and other states will no doubt compete fiercely."
Imo, that place will be in Texas on the same land Elon builds the new Launching Pad. It looks like Texas is gonna be the winner again.

Think Musk will see what each state will be offering. IMO it is more likely to be in either South California, Seattle or Denver. There are existing facilities and human resources concentrated in those areas. There is about zero infrastructure at Boca Chica. Even water have to be trucked in.

Of course Musk will do his own judging. It just made sense to me to put the Plant right by the launching pad. It could also be Brownsville, as @docmordrid stated, and very important the friendly policy Texas state is showing to SpaceX and all other investors of this magnitude.

The problem with Brownsville is attracting talented staff. Satellite factory can be anywhere as transportation of small satellites is not a problem.

[Robotbeat]

http://www.geekwire.com/2015/photos-spacex-founder-elon-musk-unveils-new-10b-space-internet-plan-private-seattle-event/

It seems that Elon has decided to go alone in this new venture covering its triple main components: (1) designing and manufacturing satellites, (2) launching and (3) operating sat network. If the first two were obvious the third one appeared plausible after his friend Greg Wyler went the separate way along with Qualcomm and Virgin Group. It seems that Virgin Group will be in charge of launching the sats which, based on the recent long-effort-no-results of Sir Richard Branson's Space Tourism, I personally doubt very much will be a success. It remains tbs if any truth in "OneWeb backers questioning whether Musk will be able to get rights to the spectrum required to build the SpaceX network."

"Musk has not yet determined where to put the satellite production factory — an unclaimed prize for which Washington and other states will no doubt compete fiercely."
Imo, that place will be in Texas on the same land Elon builds the new Launching Pad. It looks like Texas is gonna be the winner again.

Think Musk will see what each state will be offering. IMO it is more likely to be in either South California, Seattle or Denver. There are existing facilities and human resources concentrated in those areas. There is about zero infrastructure at Boca Chica. Even water have to be trucked in.

Of course Musk will do his own judging. It just made sense to me to put the Plant right by the launching pad. It could also be Brownsville, as @docmordrid stated, and very important the friendly policy Texas state is showing to SpaceX and all other investors of this magnitude.

The problem with Brownsville is attracting talented staff. Satellite factory can be anywhere as transportation of small satellites is not a problem.

Agreed! Was going to say the same thing. Lots of people wouldn't want to move down there, and no reason you need to put the satellite plant there.

Anyway, you're not going to be launching the constellation from Texas, the plane change required would be way too big. Probably would launch from California.

[JH]

I don't see any reason to place the satellite factory near the launch site. First of all, they can only do a dozen launches per year from Boca Chica, meaning that they'd probably also be launching their in house satellites from some of the 3 other pads they currently have access to (even assuming that Boca Chica is at a favorable latitude for the launches). Secondly, satellites aren't so large that the only way to move them around the country is over water.

Then there is the staffing issue. I can't picture people who are used to Seattle happily moving to the far end of Texas.

-typo

[docmordrid]

No need to move down there. U. Texas  Rio Grande Valley has an engineering school there,  and Texas A&M is planning on building another. That would be 3 engineering schools within 60 miles of Brownsville. Not to mention that if the design work is done in Redmond the factory shouldn't need as  many.

[ArbitraryConstant]

My guess is drag it along for 2 or 3 years on SpaceX dime (designing sat offerings and architecture chasing several paths, design the kilofactory and processes, etc.) then spin the subsidiary to investors, with SpaceX maintaining a controlling position and the market/partners coughing up to $15 billion for almost half of the subsidiary's equity.  That distances the big financial gamble from the rest of SpaceX.

This broadly makes sense. Some potential synergies (secondary payload on CRS launches will allow recurring opportunities to validate satellites in orbit).

Also even watching the video it's totally non-obvious what stages this involves and how much investment over what time period. It definitely sounds like they're starting small. This sounds like engineering of the basic spacecraft bus, propulsion, etc for the first few years. Then scale up would get expensive. It's entirely possible they'll go multiple years before they spend the first $100 million, then start getting spendy when they start scaling up. At that point the traditional investment path you describe makes a lot of sense. Then they can go to investors with "here's the satellite bus, we've flown it 5 times and tested it, here's the manufacturing line, here's the launch vehicle".

[ArbitraryConstant]

For deorbit, I think it's possible they will have little tugs that attach to a latch point on a dead sat, and deorbit it.

Just not seeing it. You don't know what condition it's in, whether it's spinning or whatever. Doing this economically seems orders of magnitude harder than just including passive failsafe, harder even than the main payload.

[mfck]

My guess is drag it along for 2 or 3 years on SpaceX dime (designing sat offerings and architecture chasing several paths, design the kilofactory and processes, etc.) then spin the subsidiary to investors, with SpaceX maintaining a controlling position and the market/partners coughing up to $15 billion for almost half of the subsidiary's equity.  That distances the big financial gamble from the rest of SpaceX.

This broadly makes sense. Some potential synergies (secondary payload on CRS launches will allow recurring opportunities to validate satellites in orbit).

Also even watching the video it's totally non-obvious what stages this involves and how much investment over what time period. It definitely sounds like they're starting small. This sounds like engineering of the basic spacecraft bus, propulsion, etc for the first few years. Then scale up would get expensive. It's entirely possible they'll go multiple years before they spend the first $100 million, then start getting spendy when they start scaling up. At that point the traditional investment path you describe makes a lot of sense. Then they can go to investors with "here's the satellite bus, we've flown it 5 times and tested it, here's the manufacturing line, here's the launch vehicle".

There's also a substantial amount of work that needs to be done on network architecture which will probably run in parallel with the bus and payload work.

[Nydoc]

I think a much more interesting question is what technologies they have in mind in order to make a < half ton satellite that can do the job.

Did they find a way to reduce power consumption for example?

These satellites are on an almost 50% power cycle. So you need to produce 2-3 times your average power consumption. Any idea what this level is?

Assuming we're talking about sat-to-sat laser connections and sat-to-ground RF connections at an altitude of 1100-1200km, what is the power requirement per sat?

[RedLineTrain]

Thanks to Callezetter for posting the video.  It sounds as if SpaceX has already filed with the ITU for this system (at 17:20).  If anybody has access to the ITU filing database, it would be great to see SpaceX's filings or a list of its filings.

[AJW]

An element I haven't seen mentioned is that SpX may be able to launch a number of satellite prototypes as secondary payloads on unrelated missions at minimal cost.  They could try different iterations during development while refining capabilities such as de-orbit. 

[meekGee]

A question for those who know - by there being a global Spectrum allocation, I gather that those frequencies are expected to be used, well, globally.

But satellites at 1000 km can only see spots at a radius of about 1000 km, which is sub-continent sized.  From a regulatory point of view, would it be possible for the system to use one set of frequencies, for example, in N. America, a different set in Europe, and a third in Africa?  Would that allow for more options?

From a technical point of view, what would it take to run at multiple frequencies?  I know a dish antenna reflector is good for a very large frequency band, but a multi-element antenna is pretty finely tuned. Phased array receivers work at a certain frequency, but I think that can be soft-controlled, right?


EDIT:  and btw, if you do approximate area math, it looks like they can run a tighter beam and still have very good coverage.  So perhaps a 30-degree high-quality zone, and a 45-degree (half angle) contingency service zone

[guckyfan]

Very satisfactory to me that he mentionend deorbiting satellites when out of service or defective. Important enough to him to actually mentioning it twice.

[go4mars]

There's a SpaceX Genereral section, Reusability Section, Missions section...  I suspect it's time for a SpaceX Satellite section.  This thread seems to have a wide variety of jumbled, but all interesting sub-topics related to the SpaceX satellite business, which might benefit from targeted incisions.

[Robotbeat]

Not nearly time for a whole new section. Come on.