SpaceX Internet Satellite Network: Starlink

[Cosmacelf]

Aside:
B4RN: The World's Fastest Rural Broadband

A group of people in the UK got together and did it themselves.
Not only is it fast, it's cheap.

Yes, B4RN is very impressive indeed. Here’s a short case study I wrote on another small Welsh community. This one was only 200 people, very rural. They built their gigabit fiber Internet system for about £1,250 per connection. It’s a well engineered system too, no shortcuts. Here’s the case study:

FTTH.Build

Btw, the FTTH.Build website provides information for small communities on how to build their own fiber networks. It isn’t particularly hard. It isn’t rocket science or anything

 

[Oil4AsphaultOnly]

Antenna size is inversely proportional to frequency for the same beam width/ gain. Antenna receive performance is identical to transmit ability (reciprocity).

I'm having trouble understanding the logic of this one? There are plenty of 2.4GHz antenna that have varying sizes and shapes, whose performance varies as well.

 

[ItsNotAboutTheMoney]

Yes, B4RN is very impressive indeed. Here’s a short case study I wrote on another small Welsh community. This one was only 200 people, very rural. They built their gigabit fiber Internet system for about £1,250 per connection. It’s a well engineered system too, no shortcuts. Here’s the case study:

FTTH.Build

Btw, the FTTH.Build website provides information for small communities on how to build their own fiber networks. It isn’t particularly hard. It isn’t rocket science or anything

Off topic, but I was immediately reminded:

 

[mongo]

I'm having trouble understanding the logic of this one? There are plenty of 2.4GHz antenna that have varying sizes and shapes, whose performance varies as well.

If the receiver circuit is the same, then the radiation pattern of the antenna is the critical difference. Thus can range from a fully omnidirectional antenna to a tight beam width one. The tight beam one will be larger and have a much higher signal strength (stick vs pringles can). The omni antenna (half dipole) can also have different strengths(gain) based on size (quarter wavelength vs 5/8)

If you mess with the front end of the radio section, you can add a matching circuit to compensate fir a bad antenna's impedance.

Normal stick antennas:
Dipole antenna - Wikipedia

Dish:
Parabolic antenna - Wikipedia

Phased array:
Phased array - Wikipedia

 

[Grendal]

Perhaps eventually the Starlink threads can all be merged? The first one was started by Grendal four years ago and it's now 11 pages.
SpaceX Internet Satellite Network: Starlink
The thread in TSLA Investor Discussions is up to 10 pages in less than a year.
Super fast Internet via SpaceX Satellites / Starlink
They look remarkably similar. Someone trying to create a convincing argument for a multiverse?

Done. I can't merge the investor Discussion. Hopefully, the topics diverge enough to leave both discussions up.

 

[Oil4AsphaultOnly]

If the receiver circuit is the same, then the radiation pattern of the antenna is the critical difference. Thus can range from a fully omnidirectional antenna to a tight beam width one. The tight beam one will be larger and have a much higher signal strength (stick vs pringles can). The omni antenna (half dipole) can also have different strengths(gain) based on size (quarter wavelength vs 5/8)

If you mess with the front end of the radio section, you can add a matching circuit to compensate fir a bad antenna's impedance.

Normal stick antennas:
Dipole antenna - Wikipedia

Dish:
Parabolic antenna - Wikipedia

Phased array:
Phased array - Wikipedia

Right, so the above makes this earlier statement irrelevant:
"Antenna size is inversely proportional to frequency for the same beam width/ gain. Antenna receive performance is identical to transmit ability (reciprocity)."

I just want to make sure that antenna size/shape is not an absolute restriction (e.g. must be flat pizza box size and shape or no connection possible).

 

[mongo]

Right, so the above makes this earlier statement irrelevant:
"Antenna size is inversely proportional to frequency for the same beam width/ gain. Antenna receive performance is identical to transmit ability (reciprocity)."

I just want to make sure that antenna size/shape is not an absolute restriction (e.g. must be flat pizza box size and shape or no connection possible).

GPS antennas are like that due to needing full sky view and the satellite transmissions are set up to work when overlapping. With Starlink, the sat's tight beam would give some level of specificity, but you would not be able to uplink at all (due to need transmit power and you would impact all sats in view). Not sure on the receiver/ antenna gain you would need.

 

[BioSehnsucht]

Right, so the above makes this earlier statement irrelevant:
"Antenna size is inversely proportional to frequency for the same beam width/ gain. Antenna receive performance is identical to transmit ability (reciprocity)."

I just want to make sure that antenna size/shape is not an absolute restriction (e.g. must be flat pizza box size and shape or no connection possible).

While it might be possible (I'd say it should be possible but I'm not actually an RF engineer, so ...) to engineer a non-flat phased array, as far as I know nobody has done so yet. And of course there must still be some depth to it, so if the flat one was pizza box depth, the curved one probably is as well, so you haven't gained much.

A phased array is basically a virtual dish that can be (practically) instantaneously re-oriented across the entire operating range, and even potentially be "pointed" in several different directions "at once" (because it can be "re-aimed" in lockstep with the time slots of multiple sats.

So other than a pizza box sized phased array, you could in theory work with a bunch of steerable dishes (each one aimed at another satellite in view), but obviously that's worse in (nearly?) every way.

 

[Ed Hart]

Starlight is coming together now. I think the strategy is as simple at this:

 

[mongo]

Starlight is coming together now. I think the strategy is as simple at this:View attachment 350563

5. Deployment is not early, standard FCC requirements for a constellation. SpaceX asked for delay in requirement to get to 50% but was denied.

6. No SpaceX IPO, Elon is totally against it. Hates that Tesla is a public company

 

[AudubonB]

Done. I can't merge the investor Discussion. Hopefully, the topics diverge enough to leave both discussions up.

I have moved the offending thread out of Investors into this sub-forum; if Grendal now will merge them the Moderators' work will be done....

 

[mongo]

I have moved the offending thread out of Investors into here; if Grendal now will merge them the Moderators' work will be done....

People post from dusk to dawn,
A moderator's work is never done.
(but is much appreciated)

 

[AudubonB]

as per "done...."

 

[ecarfan]

No SpaceX IPO, Elon is totally against it. Hates that Tesla is a public company

Agreed, not going to happen.

 

[dm28997]

An interesting take on how the speed of space travel internet can assist speed of transmission beating out fiber internet. Of course this is speculation since Elon comes up with some solutions outside of the box. The article suggest high frequency trading could be worth billions.

The first detailed look at how Elon Musk's space internet could work | New Scientist

 

[mongo]

An interesting take on how the speed of space travel internet can assist speed of transmission beating out fiber internet. Of course this is speculation since Elon comes up with some solutions outside of the box. The article suggest high frequency trading could be worth billions.

The first detailed look at how Elon Musk's space internet could work | New Scientist

Thanks for the link!

One issue I have is

Paying customers will be crucial, because the costs of launch and maintenance will be eye-watering. Launching a single satellite costs tens of millions of dollars.

This statement ignores launching multiple sats at once, which it a huge cost save. Even Iridium was sub ten million (~6 million) a sat at a quantity of 10.

 

[ecarfan]

This statement ignores launching multiple sats at once, which it a huge cost save. Even Iridium was sub ten million (~6 million) a sat at a quantity of 10.

Yes, that is a surprising mistake by the author of that article. It’s simple math. And when you factor in re-using the booster 10 times or more...

Here’s another quote from that article: “Each Starlink satellite will probably only last a few years, so SpaceX will need to launch new satellites to replace dead ones every few weeks.”

I’m not sure that is a well-founded assumption.

 

[GWord]

Yes, that is a surprising mistake by the author of that article. It’s simple math. And when you factor in re-using the booster 10 times or more...

Here’s another quote from that article: “Each Starlink satellite will probably only last a few years, so SpaceX will need to launch new satellites to replace dead ones every few weeks.”

I’m not sure that is a well-founded assumption.

The author seems proud to have mastered the concept of averaging but in practical terms, the cadence of failures is not likely to be linear. And in any respect, even if it is, on orbit spares give SpaceX a chance to wait until they need to replenish a bunch in one flight. Or alternatively, just ride share a few whenever available. Seems like a solvable problem.

 

[mongo]

The author seems proud to have mastered the concept of averaging but in practical terms, the cadence of failures is not likely to be linear. And in any respect, even if it is, on orbit spares give SpaceX a chance to wait until they need to replenish a bunch in one flight. Or alternatively, just ride share a few whenever available. Seems like a solvable problem.

Definitely solvable (esp once BFR is launching)

Won't the failure rate track the launch rate? If they launch at regular intervals, and they fail in a typical bell curve rate, then they should fail at regular intervals once in the main part of the curve. If it doesn't, then it will be peaky and may need spares launched preemptively (again regular replacement intervals).

As you mention, the sheer number of sats does mean replacement is not a time critical activity.

 

[ItsNotAboutTheMoney]

Definitely solvable (esp once BFR is launching)

Won't the failure rate track the launch rate? If they launch at regular intervals, and they fail in a typical bell curve rate, then they should fail at regular intervals once in the main part of the curve. If it doesn't, then it will be peaky and may need spares launched preemptively (again regular replacement intervals).

As you mention, the sheer number of sats does mean replacement is not a time critical activity.

A regular launch cadence for scheduled replacement, plus any special replacements, maybe hitching a ride, could make sense. Are there any technicalities on having to put replacement satellites into specific orbits that would make it difficult to shove a hitchhiker in there?