-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Where did you even find pricing for Iridium ? The best I could find was voice plans for $70 at 60 minutes a month... that's nothing in terms of data (something like 13kbit/s for 60 minutes). Even with 100% profit margin you'd probably get at least 2Mbps 24/7 at $70 from Starlink, and that's using the old guesses at costs from the above linked NSF post.
Or was the $70 in reference to Comcast and friends? Yeah, they're not the target market. Rural customers often can't even get 1Mbps though at any price, though, so there'll be plenty of those.
Or was the $70 in reference to Comcast and friends? Yeah, they're not the target market. Rural customers often can't even get 1Mbps though at any price, though, so there'll be plenty of those.
Yes the $70/month was referring to a typical wired Internet connection. I agree that the target market isn’t going to be urban people with lots of connectivity choices. It will indeed be isolated or rural customers with the alternative being dial up or dsl or nothing.
I was trying to pop the bubble of expectations that this would compete on price with urban wired internet. I don’t think it will.
The "good enough for gaming" is interesting. I read today that Google is involved in Starlink and at the same time is working on the Yeti gaming platform. It's said to do the rendering on servers allowing the client to use low-cost hardware and just a browser.
Is it not a dangerous assumption?
As the price of launching payloads into space continues to drop, I would expect the average satellite price to drop significantly.
First there is economies of scale. More demand will open up for more competition and more efficient production methods of satellites.
But more importantly, as launching payloads becomes cheaper and faster, you can accept much higher risk of having a satellite malfunction. Lowering the QA / material requirements. Or it would make sense to designed more short lived satellites and retire early in order to reap the benefits of technological advances.
Imagine designing a smartphone that last for 20 years compared to buying a new every 2 years... Tech advances exponentially which makes old tech obsolete very fast.
SpaceX's Starlink will have economies of scale. Current satellite manufacturers do have standard platforms they build off of, but few companies buy multiple of the same satellite (Iridium being one). Lots of design, construction, and testing on each one. Along with what you mentioned that no one wants to launch a new satellite with old tech. So satellite prices will stay high. If someone only needs one satellite, it is cheaper to build one reliable sat than two less reliable with the second as a back-up. Or else have a loss of the service while waiting for a replacement to be built. No that there is much cost to trade off for reliability. Satellites are mostly solid state with only a few moving parts. Lifetime is limited based on maneuvering fuel (if no electronics failures) which is a minor part of the overall cost (esp will high mass capable launch vehicles).
Given that there's huge rural areas with no broadband, and they really want broadband... (AND they're often voting for right-wing crazies because they don't have good access to the Internet and their TV and radio and newspapers have all been bought up by right-wing propagandists...) AND the fixed-wire companies really don't want to spend the money on running the wires to them...
I think Musk is going to ignore the urban market to start with and saturate the rural market. If he can get the price of the "pizza box sized" Starlink connection down low enough, *every* farmhouse will have one, the way they all have satellite TV dishes now (and often did even when the dishes were 6 feet in diameter). This is a monumental market in the US alone, and much larger worldwide. No reason to move into the dense urban market, which is better served with fiber anyway.
I just came back from a trip to the Galapagos. There’s a perfect market for Starlink. We stayed at a luxury safari samp on the most populated island, and their connection was 2 Mbps and it cost them $1100/month. Supposedly the best Internet connection on the island was a government dept. with an 8 Mbps link. There are lots of isolated communities like this, as well as all ocean transport.
SpaceX seeks patent for custom-built Starlink internet satellite antenna design
Phased array patents. Scaling up in process.
Phased array patents. Scaling up in process.
There's more to it than finding a place to put the phased array antenna; moving targets are harder, there's a phenomenon called Raleigh Fading to deal with, trees, bridges... the real answer is to use Starlink as the backhaul for self-sufficient 5G cells pretty much everywhere.If you really wanted to you could built them into the hood (being about the only place on top of the vehicle that isn't mostly glass) but you'd have to switch to using materials RF transparent at the relevant frequencies and as such it would be a complex solution that isn't much needed as most places have cellular coverage. But it is technically feasible, and as such it might some day happen. Hence I would not be surprised, versus I am expecting
If you really wanted to you could built them into the hood (being about the only place on top of the vehicle that isn't mostly glass) but you'd have to switch to using materials RF transparent at the relevant frequencies and as such it would be a complex solution that isn't much needed as most places have cellular coverage. But it is technically feasible, and as such it might some day happen. Hence I would not be surprised, versus I am expecting
Thinking about it - shouldn´t they be able to integrate the antenna *into the roof glass*? I´ve seen antennas for FM radio that were attached to bottom of wind shield. Sorry for off topic.
Fact Checking
Well-Known Member
Those are passive FM antennas - an active beam forming phased array antenna is more like a matrix of chips - not trivial to integrate into glass, and not transparent either:
The active beam forming is achieved by each of those nodes on the grid being independent emitters/receivers, and by adding a time delay (and other transformations) to the base signal. This means that each node is a small chip/shifter that stores the signal digitally and emits it after a (programmable) delay and with a transformation.
By changing the delays and transformation on a per node basis the outgoing signal can be made highly directional. Same applies to reception, which allows the reuse of the same frequency channels by multiple satellites and ground stations, with very low interference.
Such hardware is highly non-trivial, it has to track fast moving LEO satellites, has to be able to hop from one satellite to the next one without significant delay, and has to be able to compensate for a significant dynamic Doppler shift caused by satellites moving at 7-8 km/s as well.
GPS receivers are easy in comparison: unidirectional broadcast signal on well known frequencies that are stable due to the satellites having no relative motion.
Last edited:
Those are passive FM antennas - an active beam forming phased array antenna is more like a matrix of chips - not trivial to integrate into glass, and not transparent either:
The active beam forming is achieved by each of those nodes on the grid being independent emitters/receivers, and by adding a time delay (and other transformations) to the base signal. This means that each node is a small chip/shifter that stores the signal digitally and emits it after a (programmable) delay and with a transformation.
By changing the delays and transformation on a per node basis the outgoing signal can be made highly directional. Same applies to reception, which allows the reuse of the same frequency channels by multiple satellites and ground stations, with very low interference.
Such hardware is highly non-trivial, it has to track fast moving LEO satellites, has to be able to hop from one satellite to the next one without significant delay, and has to be able to compensate for a significant dynamic Doppler shift caused by satellites moving at 7-8 km/s as well.
GPS receivers are easy in comparison: unidirectional broadcast signal on well known frequencies that are stable due to the satellites having no relative motion.
Actually makes me wonder if you could do it with ITO (Indium Tin Oxide, a transparent conductor... used for example in solar cells and LEDs)
But threads about wireless and satellite technology do folks? ?? Mods get the crap threads outta here....
The moving target problem pretty much has to be solved to work anyways. The satellites will be moving pretty quick, the relative motion of the vehicle should be within the error margins needed for tracking the satellites, and even at only a few degrees beam width you end up with pretty big spots at both ends so you only have to get close enough.
But yes, throwing down self sufficient tower w/ solar, battery, starlink, and regular 4/5/whatever-G radios on it is the more practical solution. Cellular coverage in the US is already pretty good, you usually have to go out of your way to find a road without coverage. (though I know of some areas in Texas with no coverage for a few miles, the scenic route to my dad's house, which we won't be taking any more because it's hilly and no Supercharger on the scenic route)
While you might in theory build a phased array antenna on glass, you would end up with an ugly mess of wiring and chips on the glass. Not worth it.
This used to be part of my day job. Constant motion isn't too bad, but think about the car turning! And the other problems remain.
You sound like you know a lot more about this than I do, but it seems like a speed change of +- 90mph wouldn’t be much relative to the rotation of the earth, particularly for a satellite orbiting only 750 miles up. Without looking it up or calculating it, that’s got to be at least a couple thousand mph.
You have a car moving at, say, 100km/h in a straight line, and a satellite moving at 7,000 km/h in a different but approximately straight line. It's easy to do the vector sum and know exactly how to compensate for those motions. But the satellite is about 400km up, and not straight above, so let's say it's 600km away from the car. Now, the car turns 1º. There's an approximation that pilots use, called the 1 in 60 rule, that for small angles is a good way to estimate stuff. The beam that the array was pointing to/from the satellite moves about 10km away from the estimated position. That's a lot for a tight beam. Anyway, too much.
Similar threads
