chatgpt...
Short answer: it’s not a single fixed number—but for the user-link beams
Starlink forms from each satellite, the main-lobe (−3 dB) beamwidth is
roughly on the order of *~2–2.5° at broadside (nadir)*. At Starlink’s
~550 km altitude that corresponds to a ground spot about *~20–25 km in
diameter* directly under the satellite. As the beam steers off-nadir
toward the horizon, the footprint widens and becomes elliptical, so the
long dimension can grow to *tens of kilometers more* depending on look
angle. (ResearchGate
<https://www.researchgate.net/figure/Spot-beam-footprints-at-3dB-and-10dB-for-two-positions-of-a-Starlink-satellite-with_fig17_373190536?utm_source=chatgpt.com>,
Potaroo
<https://www.potaroo.net/presentations/2023-11-28-starlink-ripe87.pdf?utm_source=chatgpt.com>,
Delhi Fence Review
<https://delhidefencereview.com/2020/08/12/spacexs-starlink-project-examining-its-military-potential/?utm_source=chatgpt.com>)
Why those numbers:
*
Academic/engineering analyses that back-out Starlink’s beam patterns
from received signals model the *−3 dB beamwidth at ~2.25°* at
nadir, which maps to ~21–22 km diameter on the ground at 550 km
altitude (simple small-angle geometry). (ResearchGate
<https://www.researchgate.net/figure/Spot-beam-footprints-at-3dB-and-10dB-for-two-positions-of-a-Starlink-satellite-with_fig17_373190536?utm_source=chatgpt.com>)
*
Independent presentations summarizing these studies report multiple
simultaneous spot beams per satellite with user-downlink channels,
consistent with ~2–3° beamwidths. (Potaroo
<https://www.potaroo.net/presentations/2023-11-28-starlink-ripe87.pdf?utm_source=chatgpt.com>)
*
Non-primary but consistent sources describe typical *cell/spot sizes
~15 miles (~24 km)* near nadir. (SatMagazine
<https://www.satmagazine.com/story.php?number=1026762698&utm_source=chatgpt.com>)
Rule of thumb you can use:
*
Ground-spot diameter ≈ 2⋅h⋅tan(θ2)2 \cdot h \cdot
\tan(\tfrac{\theta}{2}), where hh is satellite altitude (~550 km)
and θ\theta is the −3 dB beamwidth in radians.
For θ≈2.25∘\theta\approx2.25^\circ → diameter ≈ 21–22 km (nadir).
Off-nadir, expect larger, elongated footprints.
If you’re trying to model capacity/coverage at a location, tell me your
latitude and typical elevation angles you see—I'll plug in the geometry
and give you a location-specific footprint range.
On 8/18/25 3:05 PM, Mike Hammett wrote:
https://www.starlink.com/public-files/starlinkProgressReport_2024.pdf
Page 26
"96 Gbps per satellite"
--
Mike Hammett
----- Original Message -----
From: "Adam Moffett"<[email protected]>
To:[email protected]
Sent: Monday, August 18, 2025 4:52:40 PM
Subject: Re: [AFMUG] T-Mobile 5G and Fiber Home Internet
Interesting. I said way back that people were over-estimating the capacity.
Articles were saying 10Gig per satellite. I was saying 10Gig for a geographic
area, with the maximum size of the area determined by the size of the beam
coming from the satellite, and it might be a pretty big area. The bottom line
is they have a fixed amount of Hz to work with, and just like any other
wireless, they can't use the same channel in the same location no matter how
smart they are.
I don't know their actual antenna specs, but I see some graphics showing their
2 nd generation satellites are a 4.1m x 2.7m rectangular panel which faces down
towards Earth.
SpaceX says upgraded Starlink satellites have better bandwidth, beams, and more
I'm sure that's artwork rather than a photo, but it looks like it has several
panel antennas at roughly 1sq meter each. It has to have antennas for a
connecting a ground station to the satellite up and down, as well as satellite
to the user up and down.
If I pretend there's 2sqm for downlink to the user, then maybe it's 30ish dBi
and 5-ish degree beamwidth. Tan(2.5) x 350 miles makes an area on the ground 26
miles in diameter. So 10gig for a whole city, not 10gig per satellite. If they
can use the entire surface area for satellite to user, then the beam is roughly
13 miles In diameter. I don't have real numbers on their antennas, but they
only have so much geometry to work with.
Their satellites pre-2023 were a lot smaller. So figure the beamwidth was wider
in the past, and maybe capacity improves as the satellites age out and get
replaced. Maybe we're seeing overage charges because they can only replace them
at a certain rate.
Since 10Gig was a marketing number to begin with, I assume that's before FEC
and with a perfect SNR. Maybe in real life they get 1/2 to 2/3's of that.
A lot of assumptions, yes, but I don't think they're unfair assumptions.
-Adam
From: AF<[email protected]> on behalf of Robert<[email protected]>
Sent: Friday, August 15, 2025 11:32 AM
To:[email protected] <[email protected]>
Subject: Re: [AFMUG] T-Mobile 5G and Fiber Home Internet
SpaceX has reached a limit somewhere.. They are resorting to pricing games and
now putting in caps that they never had before... Either that or the cell phone
greed gene is kicking in. I wonder if they recently hired any cell phone execs
that brought the greed with them. We are seeing ex customers come crawling back
after getting $600 overage bills...
On 8/15/25 7:07 AM, Josh Luthman wrote:
But they got $3M to do nothing in VA.
$76M in TN
https://broadbandbreakfast.com/amazon-and-spacex-undercut-competition-in-tennessee-bead-bidding/
On Fri, Aug 15, 2025 at 9:57 AM Ken Hohhof <[email protected] > wrote:
https://arstechnica.com/tech-policy/2025/08/starlink-tries-to-block-virginias-plan-to-bring-fiber-internet-to-residents/
From: AF <[email protected] > On Behalf Of Josh Luthman
Sent: Friday, August 15, 2025 8:53 AM
To: AnimalFarm Microwave Users Group <[email protected] >
Subject: Re: [AFMUG] T-Mobile 5G and Fiber Home Internet
The real winner is SpaceX.
0 additional work for big fat checks.
On Thu, Aug 14, 2025 at 5:05 PM Mike Hammett <[email protected] > wrote:
T-Mobile and\or their affiliates won big in Louisiana.
Well, still subject to official award.
--
AF mailing list
[email protected]
http://af.afmug.com/mailman/listinfo/af_af.afmug.com
