Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Starlink direct to T-Mobile phone

This site may earn commission on affiliate links.

Cosmacelf

Well-Known Member
Supporting Member
Mar 6, 2013
12,686
46,766
San Diego
So the big announcement is that SpaceX has partnered with T-Mobile to roll out a direct to cell phone service. Will work with existing cell phones - no extra customer hardware needed. T-Mobile is going to dedicate a slice of mid-band PCS spectrum to work with V2 Starlink satellites. Expected end of 2023.

This will be a messaging, low bandwidth service. They expect the total bandwidth for each satellite cell will be 2 Mbps to 4 Mbps. The technology will allow for a store and forward service, so a satellite doesn't even need to be in view when you send the text message. The phone will wait until it gets satellite connectivity and then send the message. This also means that it can start to be operational well before full satellite coverage is available. Elon said they could turn it on (certainly for emergency services) when they would get like a satellite zipping overhead once every fifteen minutes or so.

T-Mobile said that aspirationally they would include this service for free in their higher end (they said "most popular") services. T-Mobile and SpaceX also made the call out to international carriers to come on board and allow people to roam worldwide.

The biggest use case for this technology will be safety and security for people outside cell phone range, either hiking or driving a car. Of course, texting a picture standing on top of Mt Whitney will also be popular :)
 
Last edited:
BTW, this will require T-Mobile to work with messaging app providers to ensure compatibility. So, the regular phone SMS text feature should work out of the box, but iMessaging and WhatsApp and even RCS will need to be tweaked somewhat to ensure bandwidth compatibility. You don't want to allow those multi-purpose apps to open up a real time video calling stream and swamp the satellite cell's total bandwidth.
 
This direct to cell phone technology might have come from SpaceX's acquisition of Swarm. They may yet offer an IoT service, but I am guessing the technology to allow direct to cell phone is similar to what Swarm had.
I had thoughts about that pre-presentation, but this is using the normal cell frequency and possibly protocol but instead of a cell tower a few miles away, it's a satellite flying through space. Which is feasible if you have a really big antenna. Apparently like 5 meters by 5 meters big.

Antenna beam width is inversely proportional to frequency and size. T-Mobile has Band 2 (1900 MHz) Band 5 (850 MHz) and Band 4 (1700/2100 MHz). If we compare the 1.7 GHz band to Starlink at 10.7-12.7GHz an antenna for cell needs to have dimensions 6.5 times larger for the same spot size as a user terminal. And that's 6.5x each side so an area 42 times larger. To get enough recieved signal from the phone, it may need to be even larger (and cell size may need to be smaller regardless)
 
Thanks for the synopsis! Imagine what a Tesla phone built from the ground up with this technology will be able to do...

The future is looking brighter!

There is no Tesla phone to make from "this technology" it is all about the satellites. Standard phones send and receive to the sats because the sats have really good software and antennas.

You can build a phone from scratch and it'll just be another phone, nothing to gain from this.
 
I had thoughts about that pre-presentation, but this is using the normal cell frequency and possibly protocol but instead of a cell tower a few miles away, it's a satellite flying through space. Which is feasible if you have a really big antenna. Apparently like 5 meters by 5 meters big.

Antenna beam width is inversely proportional to frequency and size. T-Mobile has Band 2 (1900 MHz) Band 5 (850 MHz) and Band 4 (1700/2100 MHz). If we compare the 1.7 GHz band to Starlink at 10.7-12.7GHz an antenna for cell needs to have dimensions 6.5 times larger for the same spot size as a user terminal. And that's 6.5x each side so an area 42 times larger. To get enough received signal from the phone, it may need to be even larger (and cell size may need to be smaller regardless)

This will be at 2500 mhz (Band 41), what was previously sprint bandwidth.
 
In the big picture:

The breakthrough here isn't the phone, it isn't even the satellite. It's the spaceship that will launch the satellite.

The phone is good enough.

The satellites couldn't be made large enough until we had a spaceship large enough.

So if you want to keep track of this new tech, the thing to watch is Starship launches. That's what decides how quick this "new tech" hits the market.
 
This will be at 2500 mhz (Band 41), what was previously sprint bandwidth.
At least that is my take on it.

If you listen to the video they just say "mid band PCS" without mentioning a frequency.

Why mid-band matters for 5G lists 2.5 GHZ as the nationwide mid band option in Tmobile speak.

A lot of the early articles are saying it is 1.9 GHZ instead, but I think they are wrong, just assuming because 1900 mhz PCS was called "mid band PCS" in the past.

But TMobile considers anything between 1 and 6 GHZ to be mid band per the page I posted above.

mid-band, the range between 1 GHz and 6 GHz.
 
  • Like
Reactions: petit_bateau
This will be at 2500 mhz (Band 41), what was previously sprint bandwidth
Its a less common band globally.

The press conference suggested the SpaceX will look to partner with other carriers, so will be interesting to see what range of bands they can target.

If also note that AFAIK all Tesla vehicles are currently 3G/4G.. so we'd need to see a new 5G modem for Teslas to access this.

Not that this is the point. Given the bandwidth limitations this purely looks like an emergency/ low bandwidth text based service that is about coverage rather than throughput.
 
  • Like
Reactions: petit_bateau
This will be at 2500 mhz (Band 41), what was previously sprint bandwidth.

At least that is my take on it.

If you listen to the video they just say "mid band PCS" without mentioning a frequency.

Why mid-band matters for 5G lists 2.5 GHZ as the nationwide mid band option in Tmobile speak.

A lot of the early articles are saying it is 1.9 GHZ instead, but I think they are wrong, just assuming because 1900 mhz PCS was called "mid band PCS" in the past.

But TMobile considers anything between 1 and 6 GHZ to be mid band per the page I posted above.

Broadband Personal Communications Service (PCS)
FCC calls out PCS as 1850 MHz to 1990 MHz.

Sprint had CDMA in Band 25 at 1900 MHz, Band 41 @ 2.5 GHz was for 5G expansion.

Cheat sheet: which 4G LTE bands do AT&T, Verizon, T-Mobile and Sprint use in the USA? - IPS Inter Press Service Business

2.5 GHz would help with antenna size though.
 
There is no Tesla phone to make from "this technology" it is all about the satellites. Standard phones send and receive to the sats because the sats have really good software and antennas.

You can build a phone from scratch and it'll just be another phone, nothing to gain from this.
If Tesla made a "Tesla Phone", tons of Tesla fans would clamor for it. I mean, if a freaking whistle sells out in mere minutes, you'd think demand for a Tesla Phone would also be off the charts.
 
  • Funny
Reactions: scaesare
Agreed, not the StarLink service but the StarLink satellite “cell towers” using cellular frequencies from way up in orbit. Will be interesting if the orbital transmissions are precise enough to avoid small localized terrestrial locations. Seems much easier to use vehicle mapping to indicate no transmit zones for the vehicle transmitters.

Ideally, both geofences are needed. Otherwise terrestrial cell phones would be in constant high power search mode. Not sure if phones do that though.
Satellite aiming is very accurate and precise, but side lobes mean that the no service zone will extend beyond the black out area.
 
In the big picture:

The breakthrough here isn't the phone, it isn't even the satellite. It's the spaceship that will launch the satellite.

The phone is good enough.

The satellites couldn't be made large enough until we had a spaceship large enough.

So if you want to keep track of this new tech, the thing to watch is Starship launches. That's what decides how quick this "new tech" hits the market.
so far, seems like only 2 starship launches scheduled, upcoming, one in september (next month)

and one in 2024
Cargo Starship-Super Heavy | Superbird-9 (called super heavy)

bunch of falcon heavies in december 2022 (Road Trip to watch in 'real time"!), but land on drone ships

Falcon heavy in july 2023 landing back at the Cape at LZ-1 and LZ-2. definately road trip
(just park anywhere you can, launch to landing is around less than 10 minutes, really big flash crowds)

I would keep an eye on nextspaceflight.com and see when they announce more Starship launches then the 2 so far

 
  • Like
Reactions: petit_bateau