3
$\begingroup$

I am new to ham radio and have been playing with a RTL-SDR receiver to which I have connected a vertical 1.5m monopole antenna.

However, I can receive transmissions belonging to the 40m band (albeit with quite a lot of noise).

I have read in ham radio manuals that it is recommended to use a monopole of half the wavelength. But in this case, the length is much shorter.

Why an antenna of that length is able to receive bigger waves, is it because of the harmonics of the frequency or is it another physical phenomenon?

Thank you very much in advance.

—Fabian CD6FIQ

73

$\endgroup$

3 Answers 3

4
$\begingroup$

An antenna that's the wrong length doesn't stop being an antenna; it just becomes a less efficient antenna for the given frequency.

As a matter of fact, basically every conductor is an antenna. That's why your antenna cable cannot just be a single wire running from your RTL-SDR to the antenna; it would become part of the antenna and pick up signals, as well.

As cheap and insensitive an RTL-SDR is, compared to say your phone for cellular frequencies it's designed for, you seem to be picking up a lot of the 40 m transmission – that happens, though you could expect quite significant attenuation.

Now, it does also happen that transmissions on some other bands actually get mixed up in frequency, before they reach your antenna. But then you wouldn't "see" them in the 40m band, but somewhere else!

The cause for this frequency mixing would be a defect either with the transmitter, or a large antenna structure close to you, introducing nonlinearities. In the case of active components like transmitter power amplifiers, that is something you always need to take care about when designing the transmitter, and add sufficient filtering to suppress such harmonics (or else you'd be in violation of your transmitting license). In the case of passive components, like cabling, filters, and antennas themselves, there's a thing called Passive Intermodulation Products, and they happen when you either put too strong a field across a material that starts to behave non-linearily, or when corrosion on a metal connector lead to semiconductor properties. Both annoying!

$\endgroup$
6
$\begingroup$

As Marcus Müller says, a small antenna is still an antenna, just a less efficient one. Adding to that: an inefficient transmitting antenna is something we try really hard to avoid if we can. An inefficient transmitting antenna means wasted power and less signal available to the receiver.

But an inefficient receiving antenna, a lot of the time, is not a problem at all. As long as the noise received from the antenna (natural or manmade) is significantly stronger than the thermal noise coming from inside the receiver, making the antenna more or less efficient has basically no effect on the signal-to-noise ratio, because the dominant source of noise increases or decreases proportionally to the signal. On HF and below, where the noise levels are always well above the sensitivity of modern receivers, it's perfectly fine to use an antenna that's <1% efficient for receiving.

$\endgroup$
1
$\begingroup$

A small antenna can be as efficient as a half wave one, but with a much smaller bandwidth. As an example I made a crystal receiver for MW, about 1 MHz, using a whip antenna of length 0,5 meters (0.0017 wavelength).

See it as a capacitive probe into the two dimensional waveguide between ground and the ionosphere. By use of a very good and fairly large ferrite core I could tune it to the right frequency only loading it with the very small extra capacitor of a microwave germanium diode. With an appropriate transformer on the DC/audio side I could get enough power to hear MW transmissions from Germany about 700 km from here using a sensitive ear-phone. It means I received enough RF power to hear without applying any DC power to the receiver.

The bandwidth was small, about 2 kHz, largely because of the loading at the low frequency side. The unloaded Q was above 1000.

People operate mobile on 80 meters from cars with small antennas. It is just a matter of impedance matching. Efficiency may be modest due to losses in the coils needed.

You could improve sensitivity with an impedance transformer. A ferrite core with a primary that has in inductance that matches the capacitive impedance of your 1.5 metre wire and a secondary with a couple of turns that match the input impedance of your RTL-SDR. Bandwidth would be small, but sensitivity similar to what you would get from a quarter wave wire. More turns on the secondary would improve bandwidth and decrease sensitivity. Beware of stray capacitances. That is why the size of the toroid matters.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .