[Starlink] one dish per household is silly.

Ulrich Speidel u.speidel at auckland.ac.nz
Mon Nov 13 15:34:19 EST 2023


Caution - 250 kW peak sounds more like a horror movie (around 500 W 
average powers your average household) but there's an easy explanation. 
It's also a good example for why Reddit isn't a good source of 
information unless you know and can interpret what it is that you're 
looking at. The figures you've seen are almost certainly EIRP ones 
(effective isotropically radiated power). And EIRP ain't the same as 
power consumption.

Simple example: Take a 100 W light bulb that radiates more or less 
isotropically (same amount of light power in all directions) without any 
reflectors etc. When you look at that light bulb, you see 100 W EIRP. 
Put a mirror behind the lightbulb, so you now see the bulb and its 
reflection, which looks like two bulbs. Makes 200 W EIRP, but it doesn't 
increase your power bill. A simple mirror like that has a gain of 3 dBi.

At a Ku band frequency of 18 GHz, a 1.5 m diameter parabolic dish has a 
gain of over 48 dBi. Each 3 dBi step in there effectively doubles the 
number of transmitters you see reflected to you if you're the receiver. 
48 dBi means 16 such steps. So you have to divide the EIRP of something 
like this by the "2^16" reflections that this gives you. If you're 
having 250 kW EIRP, then the actual transmit power is just a few watts, 
and similarly your power consumption is a lot more modest.

Another way of looking at this is that what goes up (to the satellite in 
terms of bits) has to come down (to another ground station / dishy). 
That happens via a similar path. So if you'd transmit to a satellite 
with anything like even 25 kW actual transmitter power, you'd also need 
a similar amount of transmit power at the satellite to downlink. And 
you'd have to generate that power up there in orbit. Now I have solar 
PV, and I know that generating a measly 5 kW peak takes 20 panels of 
around 2 sqm size each. Generating the sort of power you'd need to 
transmit at 25 kW would require solar arrays on the satellite that are 
more like the size of a football field. Ballpark.

Thankfully EIRP for a dish of fixed size goes up as a linear function of 
dish diameter and transmit frequency as the antenna becomes more 
"pointed" as you move from conventional C band to the Ku and Ka bands 
used by Starlink.

On 14/11/2023 5:22 am, Inemesit Affia via Starlink wrote:
> You have to size power equipment for max power plus a percentage 
> regardless of current traffic needs.
>
> Look for the ANATEL docs on Reddit you'll see it there(per dish at 
> least). Or search the NASASpaceflight forum.
>
> I think it's 250kw peak. Could be 25kw instead can't remember right. 
> But each dish is more than 5kw and there's other equipment than dishes
>
> Nov 13, 2023 5:10:27 PM J Pan <Pan at uvic.ca>:
>
>     is the power consumption related to traffic volume? currently the
>     traffic is very light
>     https://www.reddit.com/r/StarlinkEngineering/comments/17k3jas/intergs_ground_station_satellite_links_much/
>     <https://www.reddit.com/r/StarlinkEngineering/comments/17k3jas/intergs_ground_station_satellite_links_much/>
>
>     -- 
>     J Pan, UVic CSc, ECS566, 250-472-5796 (NO VM), Pan at UVic.CA,
>     Web.UVic.CA/~pan
>     <http://Web.UVic.CA/~pan>
>
>
>     On Sun, Nov 12, 2023 at 10:40 PM Inemesit Affia via Starlink
>     <starlink at lists.bufferbloat.net> wrote:
>
>
>         It's kinda expensive to run. Likely 75KW peak.
>
>         Check the power requirements for a regular gateway and divide
>         by two.
>
>         Might be useful for Taiwanese Islands though
>         _______________________________________________
>         Starlink mailing list
>         Starlink at lists.bufferbloat.net
>         https://lists.bufferbloat.net/listinfo/starlink
>         <https://lists.bufferbloat.net/listinfo/starlink>
>
>
>
> _______________________________________________
> Starlink mailing list
> Starlink at lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/starlink
>
-- 
****************************************************************
Dr. Ulrich Speidel

School of Computer Science

Room 303S.594 (City Campus)

The University of Auckland
u.speidel at auckland.ac.nz  
http://www.cs.auckland.ac.nz/~ulrich/
****************************************************************


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.bufferbloat.net/pipermail/starlink/attachments/20231114/e93f91e0/attachment-0001.html>


More information about the Starlink mailing list