Radios use electromagnetic fields.

The EM field can induce a current in an antenna. This current is what is received and amplified. Recall that an antenna, at its simplest, is just a length of wire.

Keep in mind that when you are working with antennas, you have to forget about Kirchoff's laws, because the simplifying assumptions that make Kirchoff's laws work aren't true for antennas. By Kirchoff's laws, you can't have current flowing through a wire that isn't connected to anything. However, the charge density of the wire is changing under the influence of the radio waves--the radio waves are pushing the charges in the wire from one side to the other, and then back again.

There are some great answers here. I'd like to add a little.

You used the term "earthed", which means you are searching for a reference to something. An antenna has its own reference: one side is at a different potential than the other due to induction from the receiving RF wave. It's easier to think about with a dipole antenna, where each half of the antenna receives a different phase of the wave: the relative difference between the poles of the antenna is the voltage induced by the changing wave. It is its own reference.

Now in any circuit that is not literally grounded, there is always a reference voltage that is called ground. Everything is referenced and built to this. Whether or not it is "zero volts" is immaterial, since it is the reference: we can make it zero volts by analysis. The potential induced on the antenna is a quantifiable charge can be compared to this reference, and amplified.

Hmmm. That sounded clearer in my mind, but I just wanted to point out the relative nature of a reference ground in a portable circuit.

Thanks for explaining that. I always had a bit of trouble understanding how wifi and portable RF worked in general, I only accepted that it worked out of magic!

In essence they pick up on waves. An analogous question would be, "How can a ship detect waves on the sea if it isn't anchored?" - The ship is moved up and down, and that acceleration is sensed inside the ship.

In an analogous sort of way, electromagnetic waves induce movement in the charged particles (electrons) in the antenna. They are swept up and down the antenna by the waves.

In my minds eye, It seems a ship can detect motion relative to the earth... it's the earths gravity thats holding both the sea and the ship down... argh. Sorry, I'm still struggling.

A radiated EM wave at far-field distances has both a magnetic and an electric field component.

You can receive the magnetic field component with a - surprise - magnetic antenna, that has an inductor/coil, through which part of the magnetic field passes and thus induces a current (like in a classical transformer):

You can receive the electric field component with an electric antenna, like a dipole. The electric field gradient causes a voltage to be present at the antenna terminals, somewhat like an open capacitor.

Magnetic and electric antennas can be build completely independent of any ground reference. Of course, when they're used above ground it will have an influence on the radiation pattern (due to interference between the incoming direct wave and ground reflections), but they don't need "earth" to work.

If we're having a basic-physics-for-adults, I have a follow-up question...

Suppose I have a permanent magnet, and I blast it with an EM wave like you describe.

Assuming that the magnet and EM waves were very strong, could we (in principle) observe that magnet being pushed towards / away from the radio signal source?

And do radio waves have some kind of stronger effect on molecules that are dipolar?

> Assuming that the magnet and EM waves were very strong, could we (in principle) observe that magnet being pushed towards / away from the radio signal source?

You would need radio waves with impractically low frequencies (and therefore impractically large antennas).

> And do radio waves have some kind of stronger effect on molecules that are dipolar?

Radio waves interact very well with metal, which is not dipolar. If you want to poke around at dipole moments with EM radiation, you're more likely to do it in the IR region, from what I remember from my instrumentation class in chemistry.

EM fields interact with substances because the EM field can be absorbed and make electrons or other charges in the substance move around. Different types of motion require different amounts of energy, and this puts them at different places in the EM spectrum.

In the middle of the EM spectrum you can excite a lot of various vibrational modes. With longer wavelengths, you can interact with things like unpaired electrons. With short wavelengths, you can interact with electrons deep inside a substance. Each different part of the EM spectrum lets you probe a different aspect of the substance you're testing... in one part of the spectrum you might be able to see resonant peaks corresponding to specific structures or shapes, in another part of the spectrum you can see peaks corresponding to specific elements.

Take what I'm saying with a grain of salt because this is all half-remembered from chemistry classes a long, long time ago.

> ...effect on molecules that are dipolar?

That's exactly what happens in a microwave oven. Water molecules are dipolar and vibrates with the microwaves.

I thought microwave heating had more to do with inducing electric currents in the water and the resistance losses heating the food?

That being why they’re bad at defrosting ice unless there’s already some liquid water present, and why they can heat molten glass but not (directly) melt solid glass?

(I’m neither a chemist nor a physicist nor a radio frequency electronics engineer, so I may be way off here).

"Dipole rotation is the mechanism normally referred to as dielectric heating, and is most widely observable in the microwave oven where it operates most effectively on liquid water, and also, but much less so, on fats and sugars.":

https://en.m.wikipedia.org/wiki/Dielectric_heating

In ice the molecules are not free to rotate as they are in liquid water.

> earthed

The entire concept of "earth" and "ground" is inherently sort of DC and not RF, because RF is when wavelengths get short enough that you expect voltage and current to vary over the dimensions of your system, even if you have a perfect conductor.

Misusing terms in a concrete manner to make this conceptually obvious:

Imagine a receiver that is, with respect to some reference, at 100V, receiving an RF signal with a peak to peak voltage of 1V. The receiver is detecting the signal alternating between 100.5V and 99.5V with respect to ground. The receiver only 'cares' about the -0.5 +0.5 part. From the perspective of the receiver there is a signal oscillating around 0V; the local reference (Earth or otherwise) doesn't matter.

In the real world RF systems may need, or at least benefit from, earth reference. Particularly at lower frequencies where antennas are large, transmission lines are long and power supplies are imperfect common mode currents emerge as a problem and it becomes necessary to tie down these elements to a reference. That doesn't change the nature of the RF signal however; it's an alternating current that appears 'on top of' whatever the prevailing reference happens to be.

The answer to your actual question is rather simple - Maxwell’s law. But there’s a more interesting point in your question “how do you reference a signal?”. If you want to really understand this, Id read about diff mode and common mode signals.

Layman answer:

the signal is not DC so you can reference it to itself.

More specific:

The EMF must cause a current in the antenna (Maxwell’s law). This current causes the antenna to have a potential gradient along the length of the antenna. You can define any point along the antenna to be “ground” and take opposite ends of the antenna. (or not; ground is a conversion)

The receiver itself is just a fancy amplifier, it doesnt need a reference (you can keep everything as a differential value) but if you want to keep things simple, you can transform a differential signal to one referenced to an arbitrary DC point. Choose a battery terminal and call that “ground”

For the exact same reason a transformer works when it isn't grounded. Electromagnetic fields do not require a ground. But adding a ground to an AM radio does help to boost the signal (as does a long wire antenna to the top of the hot side of the ferrite coil).

An antenna has a lot of electrons in a conductor. Any passing electrical field can bully these electrons around, because it’s a conductor. If you apply an electrical field you’ll start to push electrons up the wire, where they’ll start collecting, eventually reaching an equilibrium where they are repelling each other hard enough to fight back.

Antennas use passing electromagnetic fields to allow electrons to be sloshed around like that. Then, the buildup of charge is measured and amplified to recreate the signal.

That's easy. Earth is basically very large capacitor (could think of about 1 Farad).

If you will feed AC current via smaller capacitor, it will just have larger resistance, but will also conduct.

And with higher frequency need smaller capacitor to achieve same resistance. Practically, antenna size for 50Hz is about 10th of kilometers, but for Kilo-Hertz, need much smaller antennas and smaller capacitors, and in hundreds Megagertz band, current electronics could easy make radio of match size or even smaller.

Not an EE, but my guess is Capacitors. Or at least one big enough capacitor. I think the capcitor responds to the EM fields out-of-phase with the rest of the circuit creating voltage differentials that can be measured. Not certain though.

It's not entirely wrong to say that an antenna is a capacitor, but it's not a good way of thinking about antennas. The problem is that once you start working with radio waves, everything is a capacitor, everything is an inductor, and everything is an antenna.

With a capacitor, you have two plates and you design the structure of the plates in order to create a strong electrical field between them without leaking current between the plates or leaking the electrical field into free space. Antennas are designed the opposite way---not as a "big enough capacitor", but as an incredibly bad capacitor. Instead of holding a charge, the electrical field created by the antenna leaks into free space and is lost as electromagnetic radiation, or in reverse.

A good antenna is a bad capacitor.

That said, the antenna does have non-trivial capacitance and inductance.

It irritates me when people get downvoted for attempting an explanation from what they understand.

Downvotes aren't a punishment. Marking wrong-ish things in gray is useful for discourse!