I know with a yagi antenna that putting it next to metal will cause parasitic coupling and detune the antenna. Starlink using what is essentially aesa, does this negate the issue? If so I don't see how because even traditional phased arrays are susceptible to parasitic coupling. I ask this because traditional transmitting antenna systems you always want to keep them away from metal. However it seems to be the consensus that metal roofs do not effect starlink. If this is truly the case then why?

· · Web · 1 · 0 · 1

@omnipotens you mean a starlink antenna can be 100% covered by a metal roof and still work? wtf?

@piggo @omnipotens
well in that case I think it's not the absolute distance that matters, but distance divided by wavelength, right? So maybe to a 60GHz wave, the roof seems far enough?

Also if the antenna has its own reflector between it and the roof, wouldn't that make the roof completely invisible to the antenna?

@wolf480pl @piggo Well a yagi has a reflector too. The best guess is that ku band being up to 18 Ghz doesn't take much distance from the to escape the metal


@wolf480pl @omnipotens i've seen that done with .. something. there's a way to modify an antenna to change between transverse and longitudinal waves and one of them will transmit to a radio inside a portable faraday cage.

dunno if he's using that.


@icedquinn @omnipotens
does that exploit the holes in the cage, or would it work with solid metal as well?

re: :comfycoffeewoozy: 

@wolf480pl @omnipotens this is far beyond my comprehension but i'll try (and again, i've never seen a schematic for a starlink antenna so i couldn't tell you if they're doing this)

hertzian radio waves are thrown so that they wobble in 3d space. when you hit particular things (iron especially) some of the energy is consumed by hitting the surface. or something.

non-hertzian waves are 'standing' and the math is fucky where they exist in 2d space and one of the dimensions is actually imaginary. so they just .. don't interact with most of the environment.

i've only seen a very small amount of attempts to work with these though and i don't know all the right calculuses to check.

re: :comfycoffeewoozy: 

@wolf480pl @omnipotens there may be other less theoretical means they're bopping through metal plates, dunno.

re: :comfycoffeewoozy: 

@icedquinn @omnipotens
there's no such thing as non-herzian waves

There are reactive / out-of-phase waves, but they can't transfer any energy (by receiving them you're turning them in-phase).

The most likely way of an EM wave getting through a metal plate is the metal plate not being a perfect conductor (duh, metal usually has non-zero resistance)

re: :comfycoffeewoozy: 

@wolf480pl @omnipotens this probably also could have been what was happening to make the radio appear to work, since its skipping over the surfaces rather than traveling through air.

well, disappointing in one way. neat in a different one. :gummythink:
Sign in to participate in the conversation

Linux geeks doing what Linux geeks do...