dhanson865
Well-Known Member
his sats seem to be in 5 or 6 way symmetry. The Starllink stack isn't symmetrical. It's like two vertical stacks of differing heights. I don't think it'll look anything like that KSP video when starlink spin deploys.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
his sats seem to be in 5 or 6 way symmetry. The Starllink stack isn't symmetrical. It's like two vertical stacks of differing heights. I don't think it'll look anything like that KSP video when starlink spin deploys.
Was that in spacex content, or just from the simulation you linked? In that sim it looks like the sats are mounted radially around a central dispenser, not stacked. Either way, spinning still could be part of the dispenser sequence, but--at least near as I can tell from the one image of the flight stack--it would only be releasing two sats at a time.
Possibly, but not necessarily. They would end up so close in altitude that it really doesn't matter too much which direction they're released as far as 'getting back' to the right altitude/orbit. They're all phasing themselves around the orbital plane which is going to require unique thruster firings anyway, so layering in circularization at a specific altitude just changes the timing of the firings a bit.
If it means anything, gut feel says having the upper stage+stack perpendicular to the velocity vector (as opposed to aligned, as shown in the sim) will minimize collision risk in the Monte Carlo simulation...
adiggs
Well-Known Member
I want it to be launch time! I think this is the first SpaceX launch I've been following, where I was more interested in the deploy, than the launch. Since the real point of the launch is the deploy, good!
Thats new. I have never heard of that kind of misplaced priorities
adiggs
Well-Known Member
Yeah - it's a personal thing. I tend to be more excited about the launch and first stage landing, than the satellite deployment
dhanson865
Well-Known Member
Boo, hiss, boo...
Scrubbed launch, holding for a week.
I'd down vote that but it wouldn't be fair to you.
altitude: 550 km
earth's radius: 6,356 km
at 66 sats per plane, circular distance is (550 + 6356)*2*pi/66 = 657 km between satellites.
I saw a article that stated the FCC will not allow normal operation for a satellite before it is in its final orbit.
dhanson865
Well-Known Member
SpaceX filed for a temporary exception for Starlink to allow them to use them during orbit raising and positioning. I don't know if they'll have to keep asking for that exception every so often or if the rule will change or if SpaceX will just do without at some point in the future when the exception expires.
The thing is old school sats at GEO could take weeks to get into position and people needed them to sit still to be able to aim a fixed angle ground sat dish at them. These will get into a usable position faster and all the ground antennas are phased array and can track them as they adjust their orbit. It doesn't make sense to go by the old rule for this new type of sat (in my opinion).
But then I don't work for the FCC so I can't say with any certainty how it'll go long term.
dhanson865
Well-Known Member
as far as I know the only details other than on/off are what frequency they use and minimum beam angle.
Normal rule is a minimum beam angle of 40 degrees to avoid interference/overlapping signals. The exception allows them to use a minimum beam angle of 25 degrees until the constellation is populated enough to not need that.
I didn't pay too much attention to the frequencies. It's plenty complicated.
I think the most recent filing is this one Attachment Technical Informatio SAT-MOD-20181108-00083 and it shows the frequencies and beam angles
By beam angle, are you meaning maximum elevation angle from vertical (more horizontal to compensate for low satellite density) as opposed to beam width?
That doc also calls out initial and final orbital altitude for the constellation, but I didn't see sat usage before reaching either orbit (but it is a long doc).
dhanson865
Well-Known Member
I think beam angle is measured from horizon not from vertical but so long as you start from the same reference angle either concept can get you to the same place.
As to the exception requested that is in a different doc. I don't know which one, I'd have to go digging through all the FCC letters and applications. You'll see mentions of the exception to the 40 degree rule in this doc and that comes from the same exception request.
dhanson865
Well-Known Member
OK I think this is where I got the exception stuff from
An update on SpaceX/Starlink FCC Applications : spacex
"SpaceX is applying to communicate and test the Starlink satellites immediately after deployment, before they reach their intended orbit. This allows them to perform tracking and comms testing ASAP. SpaceX Services seeks special temporary authority for communications between its Ku-band gateway earth station and the first tranche of SpaceX NGSO satellites for 60 days after orbital injection."
"The requested STA [Special Temporary Authority] would allow SpaceX to confirm the operational status of its satellites immediately upon insertion, rather than wait weeks while the satellites are obit raising to ensure proper functioning."
as far as I remember this was approved against OneWeb's objections.
my contention is that this may become the new norm, but then again maybe I'm wrong and SpaceX only wants this exception for the first launch or two.
An update on SpaceX/Starlink FCC Applications : spacex
"SpaceX is applying to communicate and test the Starlink satellites immediately after deployment, before they reach their intended orbit. This allows them to perform tracking and comms testing ASAP. SpaceX Services seeks special temporary authority for communications between its Ku-band gateway earth station and the first tranche of SpaceX NGSO satellites for 60 days after orbital injection."
"The requested STA [Special Temporary Authority] would allow SpaceX to confirm the operational status of its satellites immediately upon insertion, rather than wait weeks while the satellites are obit raising to ensure proper functioning."
as far as I remember this was approved against OneWeb's objections.
my contention is that this may become the new norm, but then again maybe I'm wrong and SpaceX only wants this exception for the first launch or two.
Yeah, that lines up with what I recalled. Communication for verification, not for end user app.
Lots to cover here, gonna try to do it in a single post....
I mean, it was never going to fling them out, but yeah a time lapse animation would be cool. Looks like the 'spin' we've been hearing about was a slow rate in yaw (think: slow-mo version of Mav and Goose in the flat spin) and, most curiously, the stack had a roll angle somewhere around 45 degrees.
Weeks for sure, I'd bound between like 2 and 10. Gut feel is something in the 4-6 week range. It all kinda depends on the size of the thruster. Less thrust = longer time required for the various orbital manners.
Definitely looks like there will be some gravity gradient aspect to the attitude control, though its tough to say how much. The Solar arrays are pretty thin and surely don't weigh that much.
In that image you can see three copper color hoops which are the features that secure the sats to the gold launch structures we talked about that run up the length of the stack. During the convenient "planned loss of signal" those launch structures were ejected somehow, then when the feed goes back to the first seconds of deployment you can see one row of those hoops.
Another cool thing in the shot just before the convenient feed cut out, there's a row of sort of gold brackets hanging off the sats with a while not-quite-circular feature in the middle of those brackets. That's a GPS antenna.
Laser links are, unfortunately, expensive and difficult to create. And, when you don't have a lot of sats in the network the laser links have to look farther away and at different angles. Its plausible that the ISLs SpaceX is building for the network simply can't close the loop with the small number of sats, so they opted to save the money and not launch them...? Having no ISLs makes the sats pretty low value on their own, but certainly spaceX is going to learn a lot about all the functionality that's on these spacecraft and how to improve that functionality in the future. There's no testing environment that can match the one you get on orbit.
The nuance here is that they would not be able to concurrently perform orbital maneuvers (firing the thruster) and simultaneously use the payload. The major reason is that both the payload and the electric propulsion system require significant power; the power system would not be sized to have both on at the same time (not to mention there's probably thermal problems with that too...). There's also a geometrical requirement to align of thrust vector, [satellite] velocity vector, and center of mass during thruster firing...though...if you look at the wire-frame sat in the solar array image linked upthread, you can see the hall thruster offset right of center along the top edge of the satellite. That offset location is certainly to align with the center of mass, I wouldn't be surprised if SpaceX sharpened their pencils to also have it align with the velocity vector.
Anyway, if I were to speculate, I'd say the exemption is really in place because either a) there's some planned time delay in starting orbit raising of some sats vs the others and so you might as well fire up the payloads on the sats that are going to be sitting around for a little while or b) in the event the propulsion system fails or in some other way doesn't perform as expected they can just fire up the payload in whatever orbit the sat ended up in.
There's a bit to unpack in that statement. GEO sats get to GEO altitude like 10-11 hours after launch or something. They don't have one big delta-v event though (like you'd get in a classic Hohman transfer) but rather fire the main thruster a number of times over a few days to de-incline, circularize, and phase. So they really take days to 'get into position'. Note that most actually don't start out in their revenue-generating orbital slot but rather spend the first month in a different slot positioned above whatever ground station they're using for in-orbit test.
The exception to the above is if a satellite is using electric orbit raising--which is still quite rare these days--which turns days into months. Typically 2-6 months, depending on what orbit you start in.
I mean, it was never going to fling them out, but yeah a time lapse animation would be cool.
Looks like the 'spin' we've been hearing about was a slow rate in yaw (think: slow-mo version of Mav and Goose in the flat spin) and, most curiously, the stack had a roll angle somewhere around 45 degrees.Weeks for sure, I'd bound between like 2 and 10. Gut feel is something in the 4-6 week range. It all kinda depends on the size of the thruster. Less thrust = longer time required for the various orbital manners.
https://www.starlink.com/images/satellite/satellite__SOLAR_DISPLAY.jpg
Definitely looks like there will be some gravity gradient aspect to the attitude control, though its tough to say how much. The Solar arrays are pretty thin and surely don't weigh that much.
In that image you can see three copper color hoops which are the features that secure the sats to the gold launch structures we talked about that run up the length of the stack. During the convenient "planned loss of signal" those launch structures were ejected somehow, then when the feed goes back to the first seconds of deployment you can see one row of those hoops.
Another cool thing in the shot just before the convenient feed cut out, there's a row of sort of gold brackets hanging off the sats with a while not-quite-circular feature in the middle of those brackets. That's a GPS antenna.
Laser links are, unfortunately, expensive and difficult to create. And, when you don't have a lot of sats in the network the laser links have to look farther away and at different angles. Its plausible that the ISLs SpaceX is building for the network simply can't close the loop with the small number of sats, so they opted to save the money and not launch them...? Having no ISLs makes the sats pretty low value on their own, but certainly spaceX is going to learn a lot about all the functionality that's on these spacecraft and how to improve that functionality in the future. There's no testing environment that can match the one you get on orbit.
The nuance here is that they would not be able to concurrently perform orbital maneuvers (firing the thruster) and simultaneously use the payload. The major reason is that both the payload and the electric propulsion system require significant power; the power system would not be sized to have both on at the same time (not to mention there's probably thermal problems with that too...). There's also a geometrical requirement to align of thrust vector, [satellite] velocity vector, and center of mass during thruster firing...though...if you look at the wire-frame sat in the solar array image linked upthread, you can see the hall thruster offset right of center along the top edge of the satellite. That offset location is certainly to align with the center of mass, I wouldn't be surprised if SpaceX sharpened their pencils to also have it align with the velocity vector.
Anyway, if I were to speculate, I'd say the exemption is really in place because either a) there's some planned time delay in starting orbit raising of some sats vs the others and so you might as well fire up the payloads on the sats that are going to be sitting around for a little while or b) in the event the propulsion system fails or in some other way doesn't perform as expected they can just fire up the payload in whatever orbit the sat ended up in.
There's a bit to unpack in that statement. GEO sats get to GEO altitude like 10-11 hours after launch or something. They don't have one big delta-v event though (like you'd get in a classic Hohman transfer) but rather fire the main thruster a number of times over a few days to de-incline, circularize, and phase. So they really take days to 'get into position'. Note that most actually don't start out in their revenue-generating orbital slot but rather spend the first month in a different slot positioned above whatever ground station they're using for in-orbit test.
The exception to the above is if a satellite is using electric orbit raising--which is still quite rare these days--which turns days into months. Typically 2-6 months, depending on what orbit you start in.
Here's a great shot of the stack.
--That center vertical bar is one of the structures/beams that supports the sats during launch.
--You're looking at two side by side stacks here, you can really appreciate the half-sat vertical offset between the two sides
--Good view of GPS antennas on both stacks
--Star trackers are way on the corners (one still has a non-flight cover on it!)
--If you look close you can see the solar panels stacked up. They're super thin.
--There's a rectangular feature outboard of the GPS antenna on either stack that has some smaller white and grey rectangular features within it. Best guess is that's T&C antennas.
--That center vertical bar is one of the structures/beams that supports the sats during launch.
--You're looking at two side by side stacks here, you can really appreciate the half-sat vertical offset between the two sides
--Good view of GPS antennas on both stacks
--Star trackers are way on the corners (one still has a non-flight cover on it!)
--If you look close you can see the solar panels stacked up. They're super thin.
--There's a rectangular feature outboard of the GPS antenna on either stack that has some smaller white and grey rectangular features within it. Best guess is that's T&C antennas.
At the risk of narcissistically quoting myself...but at the same time minimizing potential misinformation...
It actually looks like all the star trackers have non-flight covers on them. The one that's solid red (as opposed to most of them that are a red ring with a translucent center) is probably just a different revision.
(self relocating this reply post as it was off topic in the Investors Roundtable thread)
First, Starlink would need to have a feature where a subscriber (i.e., the DoD) can choose to have what amounts to a (possibly many-to-many, perhaps many-to-one or one-to-one, or some combination) VPN with zero connection to the internet (DoD would provide their own firewalling and internet routing at any terminal they choose, on their side of it the connection, if they wanted it).
Second, a reasonable level of encryption for downlinked and uplinked data so that an advisery cannot discover the locations of specific assets by simply intercepting the RF traffic (which would mean having some kind of rolling / encrypted identifier for the base stations themselves, not just the traffic, so that individual base stations cannot be identified once and then located every time they move), then I think they can assume a reasonable level of security. After all, the encryption problem would be common to any DoD specific hardware, so this is not a special case. Besides, by using "commercial" hardware (and assuming all regular consumer terminals have the same level of encryption) you can hide in plain sight rather than being obviously military encryption on military frequencies.
As long as the traffic is never decrypted except at the destination terminal, then it's just as secure as any purely military system would be, and possibly more so (since the data would be hidden among all the other flows, and intercepting and decrypting all the flows around the world in hopes of piecing together any useful data would be impractical even for nation states)
Technically, the encryption portion could even just be handled by using regular VPN systems "in the clear" with no encryption from Starlink itself, then you at most leak your arbitrarily chosen IP addresses - though this enough would be a security concern and likely dealbreaker, since if you could associate an IP with a particular unit you could then track it by watching where traffic destined to it appears in the air. So I suspect that there would need to be end to end encryption, and it would likely exist for consumer use too.
The first feature, if allowing for many to many VPN, would make China happy too, and let them sell service there. The only gateways to the internet for Chinese subscribers would be through China's firewall via their designated terminals. Of course, "we" might prefer they were able to sell unrestricted service in China, but let's be realistic. Better they can sell restricted service than none at all.
Now, I'm not saying that securely providing the above services in a fashion that both keeps traffic flowing only to designated terminals and also encrypted in a way to prevent tracking the movements of terminals is easy... just saying that those should be enough of a feature set to make the military happy.
I wouldn't rule out military use of Starlink, assuming some technical features. SpaceX is already a defense contractor so clearly is trusted to do certain things when they say they will (or to have the oversight by DoD to prove it), so if they implement these features and get them audited, it could happen.
First, Starlink would need to have a feature where a subscriber (i.e., the DoD) can choose to have what amounts to a (possibly many-to-many, perhaps many-to-one or one-to-one, or some combination) VPN with zero connection to the internet (DoD would provide their own firewalling and internet routing at any terminal they choose, on their side of it the connection, if they wanted it).
Second, a reasonable level of encryption for downlinked and uplinked data so that an advisery cannot discover the locations of specific assets by simply intercepting the RF traffic (which would mean having some kind of rolling / encrypted identifier for the base stations themselves, not just the traffic, so that individual base stations cannot be identified once and then located every time they move), then I think they can assume a reasonable level of security. After all, the encryption problem would be common to any DoD specific hardware, so this is not a special case. Besides, by using "commercial" hardware (and assuming all regular consumer terminals have the same level of encryption) you can hide in plain sight rather than being obviously military encryption on military frequencies.
As long as the traffic is never decrypted except at the destination terminal, then it's just as secure as any purely military system would be, and possibly more so (since the data would be hidden among all the other flows, and intercepting and decrypting all the flows around the world in hopes of piecing together any useful data would be impractical even for nation states)
Technically, the encryption portion could even just be handled by using regular VPN systems "in the clear" with no encryption from Starlink itself, then you at most leak your arbitrarily chosen IP addresses - though this enough would be a security concern and likely dealbreaker, since if you could associate an IP with a particular unit you could then track it by watching where traffic destined to it appears in the air. So I suspect that there would need to be end to end encryption, and it would likely exist for consumer use too.
The first feature, if allowing for many to many VPN, would make China happy too, and let them sell service there. The only gateways to the internet for Chinese subscribers would be through China's firewall via their designated terminals. Of course, "we" might prefer they were able to sell unrestricted service in China, but let's be realistic. Better they can sell restricted service than none at all.
Now, I'm not saying that securely providing the above services in a fashion that both keeps traffic flowing only to designated terminals and also encrypted in a way to prevent tracking the movements of terminals is easy... just saying that those should be enough of a feature set to make the military happy.
Similar threads
