SpaceX Internet Satellite Network: Starlink

[MP3Mike]

"Better performance in hot weather. Download speeds are typically 3x better at > 35°C (95°F)"

Any idea why would that be the case? My computer drastically slows down when it runs hot. But of course there is no CPU in the dish, but still wondering would heat increase the download speed?

The 3x better is compared to the standard dishy. Which maybe at 95* can only do 15 Mbps. Where this one would be able to do 45 Mbps. (Just making numbers up.)

In other words, this dish is less impacted by heat than the standard dish.

And there certainly is a CPU in the dish. (Unless they made it remote in this version, but that seems highly unlikely.)

 

[dhanson865]

"Better performance in hot weather. Download speeds are typically 3x better at > 35°C (95°F)"

Any idea why would that be the case? My computer drastically slows down when it runs hot. But of course there is no CPU in the dish, but still wondering would heat increase the download speed?

larger dish has more heat sink materials (blocks of metal with fins to radiate heat) more surface area lets it disperse the heat away from the chips doing the work.

Basically they spent more materials to address certain issues. Having 4x the retail cost gives them leeway to spend more on making a better dish.

So when they say better at better at > 35°C (95°F)" they don't mean better than 20°C (68°F)", they mean $600 dish at 35°C (95°F) is slower than $2500 dish at 35°C (95°F) so the $2500 dish is better at 35°C (95°F) than the cheaper dish.

 

[Electroman]

So when they say better at better at > 35°C (95°F)" they don't mean better than 20°C (68°F)", they mean $600 dish at 35°C (95°F) is slower than $2500 dish at 35°C (95°F) so the $2500 dish is better at 35°C (95°F) than the cheaper dish.

Thanks. That makes sense. 3x better than the smaller less expensive model at high temps.

 

[mongo]

larger dish has more heat sink materials (blocks of metal with fins to radiate heat) more surface area lets it disperse the heat away from the chips doing the work.

Warning, only partly researched.

I think a larger antenna would have more elements so the effective power per unit area is constant. Unless they shift spacing to a different fraction of a wavelength, but I'm not sure that works.

Another possibility is that the larger array has a higher gain due to aperture size which improves the signal to thermal noise ratio (at least for mechanically steered setups). Performance Dishy overall area is 2x standard, so a 3x performance improvement may result from that.
They may also be using better chips which would help explain the disproportionate (to area) price jump (though that is likely also volume induced).

 

[dhanson865]

Warning, only partly researched.

I think a larger antenna would have more elements so the effective power per unit area is constant. Unless they shift spacing to a different fraction of a wavelength, but I'm not sure that works.

Another possibility is that the larger array has a higher gain due to aperture size which improves the signal to thermal noise ratio (at least for mechanically steered setups). Performance Dishy overall area is 2x standard, so a 3x performance improvement may result from that.
They may also be using better chips which would help explain the disproportionate (to area) price jump (though that is likely also volume induced).

The smaller dish has a large amount of space allocated to the motor that changes tilt.

The larger dish is fixed position and can have all it's heat sinks closer to the back of the dish housing without that large gap the smaller dish has for the motor to move around. I haven't seen a tear down of the high performance dish but I know if I designed it minus this motor cavity I'd take advantage of that to move heat away from the board by extending heat sinks into area previously not available.


I'm saying the smaller dish only gets to use the face and sides to radiate heat effectively. The high performance dish has the option to move heat out the rear more effectively.

 

[petit_bateau]

The smaller dish has a large amount of space allocated to the motor that changes tilt.

The larger dish is fixed position and can have all it's heat sinks closer to the back of the dish housing without that large gap the smaller dish has for the motor to move around. I haven't seen a tear down of the high performance dish but I know if I designed it minus this motor cavity I'd take advantage of that to move heat away from the board by extending heat sinks into area previously not available.


I'm saying the smaller dish only gets to use the face and sides to radiate heat effectively. The high performance dish has the option to move heat out the rear more effectively.

Except that the larger dish is probably often being flat-mounted straight onto something like an RV roof made out of non-thermally conducting fibreglass. So the large rear surface area may not actually be a good place to dissipate heat from. Nothing in engineering is ever as easy as one might hope. Can they get the heat out the front of the dish ?

 

[dhanson865]

Except that the larger dish is probably often being flat-mounted straight onto something like an RV roof made out of non-thermally conducting fibreglass. So the large rear surface area may not actually be a good place to dissipate heat from. Nothing in engineering is ever as easy as one might hope. Can they get the heat out the front of the dish ?

It doesn't matter a great deal what it's mounted on as long as that surface isn't a direct heat source.

Air is a poor conductor of radiant heat. Solid fiberglass conducts heat better than air. Go find an RV or Boat in a hanger/garage in the morning and put your hand on the fiberglass shell and see how cold it feels as it pulls heat out of your hand. It will feel colder than the air around it because it has better conductivity and more thermal mass.

And maybe you still didn't get my point from earlier. They have reclaimed space that a motor and mount and such were at that they can put additional heat sink material. Even if it doesn't radiate out the back it takes heat away from the ICs.

 

[petit_bateau]

It doesn't matter a great deal what it's mounted on as long as that surface isn't a direct heat source.

Air is a poor conductor of radiant heat. Solid fiberglass conducts heat better than air. Go find an RV or Boat in a hanger/garage in the morning and put your hand on the fiberglass shell and see how cold it feels as it pulls heat out of your hand. It will feel colder than the air around it because it has better conductivity and more thermal mass.

And maybe you still didn't get my point from earlier. They have reclaimed space that a motor and mount and such were at that they can put additional heat sink material. Even if it doesn't radiate out the back it takes heat away from the ICs.

I did get your point earlier, and yes I do know about thermal conductivity and thermal mass. Though it does depend on what the mix is in the fibreglass. If it is worse than air as a conductor, then it may have a greater thermal mass but in the limit the heat will still have accumulated and the temperature of the ICs will then rise, even if the motor gap is utilised. Hence my comment.

(Here we are only au fait with UK radars. Do the Ticonderogas/AEGIS radars have a liquid wet-loop going into the back of each tx element ? Or are they able to use passive heatsinking per Starlink dishes?)

 

[bxr140]

Interesting development: Xfinity just bumped up my DL speed from 100 to 200 megs at no cost (I pay $76/month). While I acknowledge this is possibly just confirmation bias, it's hard to imagine Comcast's play is anything but an attempt to motivate me and presumably all of my neighbors--up in the woods above Silicon Valley--to stay with Comcast rather than switching to Starlink. Give my occluded sky view my only practical alternative to Comcast is Starlink, and I suspect most of the folks that live in these here woods have similar foliage constraints. It turns out big trees are big.

 

[Cosmacelf]

SpaceX just filed their final Gen 2 configuration. Copy and pasted from Reddit:

This means the Gen2 constellation will be authorized for Ka, Ku, V, and E Band. This encompass frequencies from 10.7Ghz to 86Ghz!

The Gen1 constellation is authorized for Ka, Ku, and V Band.

Gen2

Sat to UT

• ⁠10.7 to 12.75Ghz
• ⁠17.8 to 18.6Ghz
• ⁠18.8 to 19.3Ghz
• ⁠19.7to 20.2Ghz
• ⁠37.5 to 42Ghz
• ⁠Total of 8.35Ghz (currently 2Ghz)

UT to Sat

• ⁠12.75 to 13.25Ghz
• ⁠14 to 14.5Ghz
• ⁠28.35 to 29.1Ghz
• ⁠29.5 to 30.0Ghz
• ⁠47.2 to 50.2Ghz
• ⁠Total of 5.25Ghz (currently 0.5Ghz)

Sat to GW

• ⁠17.8 to 18.6Ghz
• ⁠18.8 to 19.3Ghz
• ⁠71 to 76Ghz (Quite a bit of attenuation through the atmosphere)
• ⁠Possibly 37.5 to 42Ghz on up to 7518 Sats
• ⁠Total of up to 10.8GHz (currently 1.3Ghz)

GW to Sat

• ⁠27.5 to 29.1Ghz
• ⁠29.5 to 30Ghz
• ⁠81 to 86Ghz (Some, but not as bad attenuation through the atmosphere)
• ⁠Possibly 47.2 to 50.2Ghz on up to 7518 Sats
• ⁠Total of up to 10.1Ghz (currently 2.1Ghz)

They also applied to use 50.4-51.4Ghz for GW links, but the FCC has deferred on approval.

 

[dhanson865]

https://www.reddit.com/r/Starlink/comments/yi8694

 

[bxr140]

Interesting development: Xfinity just bumped up my DL speed from 100 to 200 megs at no cost (I pay $76/month).

And even more interesting, Comcast sent me a "so sorry about that last note about 200megs--we really meant to say 400!".
No surprise, my speed tests didn't budge after the first "go faster" note let alone the second...

 

[Cosmacelf]

And even more interesting, Comcast sent me a "so sorry about that last note about 200megs--we really meant to say 400!".
No surprise, my speed tests didn't budge after the first "go faster" note let alone the second...

Yeah, when a cable company says 400 Mbps, they mean at 3am once a month...

 

[dhanson865]

Australia opened up

became

>and in case the question is asked, as it often is.. that dark circle on the left is this site: Inyarrimanha Ilgari Bundara, our Murchison Radio-astronomy Observatory - CSIRO

Radio quiet zone
The Australian and Western Australian Governments have established a ‘radio quiet’ zone to protect this unique radio astronomy site from noise created by modern life.
The zone is an area 520km in diameter, centered on the observatory, in which licensed communications and electronic devices such as television transmitters, mobile telephones base stations and CB radios are controlled to limit electromagnetic interference to the radio telescopes on site.

 

[dhanson865]

Northern Alaska, Canada, Northern Sweden opened up

The stripe exists because, although the polar shell (Group 3)( isn't complete (only half the planned satellites are on station), they get close together over the poles, so above a certain latitude, coverage is complete; below that point where there are gaps, the 70 degree shell (group 2) was supposed to fill in north of the point where the 53 degree (groups 1 and 4) satellites can reach. They planned to launch the groups in order, but congestion at the lower latitudes got so bad after group 1 was completed that they diverted the launches to adding group 4 before they really got started on 2 and 3, and they are only now getting back to work on those now that 1 and 4 are both essentially complete (missing a few failed sats each)... so the stripe is where they are too far north for complete coverzage from 1 and 4 but too far south for the incomplete 2 and 3 groups to insure continuous coverage.

 

[dhanson865]

South America extended

Since that update this showed up in South America. Is it just me or did they make that area of open cells look like Antarctica?

View attachment 866932

zoomed in

View attachment 866933


and part of Japan is live now

View attachment 866934

All of Japan open now

and much of South America expanded, affecting multiple countries vs the quote above.

 

[dhanson865]

Something new about ground stations?

 

[petit_bateau]

Something new about ground stations?

View attachment 872170

If it is not lasers, then perhaps they are able to use other/existing ground stations. How else to explain the rapid shift in coverage ?

 

[mongo]

If it is not lasers, then perhaps they are able to use other/existing ground stations. How else to explain the rapid shift in coverage ?

Could be bouncing using solar powered battery backed up no backbone ground stations?

 

[petit_bateau]

Could be bouncing using solar powered battery backed up no backbone ground stations?

That would be a lot of ground stations to deploy, and fast, would it not ?