Propagation

A mirror made of electrons

At 60 to 500 km up, the sun knocks electrons off the atoms of the thin air. Where there are enough free electrons, shortwave signals are refracted: a ray coming in at a slant bends further and further back towards the earth until it runs level and comes down again. How strongly a height refracts is given by its plasma frequency. If your signal's frequency is below it, even a vertical ray turns round.

The electrons aren't spread evenly but sit in layers. In the picture on the left each height is shaded as brightly as there are electrons. Move the pointer into it and the plasma frequency of that height shows below.

D, E, F2

D layer (60 to 90 km): daytime only. It hardly refracts but absorbs, because the electrons there keep colliding with air molecules and give off their energy as heat. Low frequencies are hit hardest, roughly with one over f². That's why 80 m is dead over long distances at midday and open in the evening.

E layer (around 110 km): by day a critical frequency of about 3 to 4 MHz, almost gone at night. Low-angle rays on 40 m often turn round there already. In summer there are also dense clouds in the E layer, Sporadic E, which bring 10 m, 6 m and sometimes 2 m over 1000 to 2000 km.

F2 layer (about 250 to 400 km): the most important one for DX. Its critical frequency foF2 is measured by an ionosonde, which transmits straight up and listens for the echo. Depending on time of day and solar activity it lies between 2 and over 12 MHz. The height hmF2 decides how far one hop reaches.

What the model leaves out

The calculation uses a round earth and an ionosphere that is the same everywhere. So every hop is the same length. In reality the ionosphere changes along the path, especially at the boundary between day and night. The earth's magnetic field is missing too: in reality it splits every signal into two waves with slightly different critical frequencies. To get a feel for how angle, frequency and layer hang together, that's enough.

Lower goes further

A steep ray needs a lot of electrons to turn round, a low one only a few. So at the same frequency the low ray comes back while the steep one goes through. The relationship is called the secant law: the highest frequency that still turns round at a given angle of incidence is foF2 times the secant of that angle. Because of the earth's curvature even a ray sent out horizontally doesn't hit the F2 layer quite flat, so one hop can't do more than three to four times foF2.

MUF

The MUF (maximum usable frequency) always applies to a path. Over 3000 km, roughly one low hop via the F2 layer, it is about three times foF2. Ionosondes report exactly this value as MUF(3000). In the model with a height of 300 km, foF2 7.2 MHz gives a MUF(3000) of 22 MHz. For shorter paths it drops down to foF2, longer ones need several hops.

Just below the MUF is often best: absorption in the D layer is lowest there. The lower limit, below which absorption swallows everything, is called the LUF.

Skip zone

If your frequency is above foF2, every ray above a certain angle goes through. The steepest one that still turns round lands closest. Before that nothing arrives, only the ground wave reaches a few dozen kilometres: that's the skip zone. The higher the frequency, the bigger it gets. With the default values it is about 1400 km on 20 m, smaller on 30 m, and on 40 m there is none, because 7 MHz is below foF2.

Interestingly, the steepest ray doesn't necessarily make the shortest hop: just below the critical angle the ray runs along inside the layer for a long way and lands further out again. You can see that nicely in the fan.

Several hops

The ground reflects the ray back up and it goes again. Every hop costs something: twice through the D layer and a bit at the ground (sea reflects better than dry land). So a few long hops beat many short ones, and that's why low take-off angles matter so much for DX.

By day

The sun is high, foF2 rises, the high bands open. At the same time the D layer is there and swallows the low ones: on 80 m at midday only steeply radiated signals arrive over short distances, on 40 m one low hop costs around 40 dB in the model. On 20 m it's only around 13 dB with the same sun.

At night

Without the sun the D layer disappears within minutes, the E layer almost completely. The F2 layer stays, but foF2 slowly drops, in winter often to 2 to 4 MHz, and the layer rises higher. Now it's the turn of 160, 80 and 40 m: hardly any absorption, long range. The high bands close because even the lowest ray goes through.

Just the D layer gone?

The tick box “D layer absorbs” only removes the absorption, not the daylight. The D layer at 60 to 90 km hardly refracts the rays, it only absorbs them: in the picture absorbed rays get paler, the path stays the same. If the sun is still high, the E layer is there too. On it, 80 and 40 m already turn round at about 100 km and make short hops instead of going up to the F2 layer. The long night hops only come when the sun is well below the horizon (below −12°): then D and E are gone. For that, turn the sun elevation down or pick “Night”.

In between

In the morning and evening the ionosphere changes fastest. Along the boundary between day and night (the grey line) the D layer is already gone but the F2 layer is still dense: on 160 to 40 m paths often open then that otherwise don't. The model here uses one ionosphere for the whole path and so doesn't show the grey line.

Sunspots

With lots of spots on the sun foF2 rises clearly, and at the peak of the eleven-year cycle 10 m is open worldwide during the day. The D layer gets a bit stronger then too, though. In the sunspots field you set how strongly it absorbs. How things stand right now is shown in propagation conditions with measured ionosonde values.

Where the antenna radiates

Which rays from the fan really carry power is decided by the antenna. A horizontal dipole radiates up and down, and what goes down is reflected by the ground. Direct and reflected rays add up, and depending on the angle they reinforce or cancel each other. The elevation pattern shows the result.

Height in wavelengths

Over perfect ground a dipole at λ/2 height has its maximum at 30°, at λ height it has two equally strong lobes at 14.5° and 49°, and at λ/4 height it radiates almost straight up. A 40 m dipole at 10 m height is only at about λ/4: good for contacts within Germany (NVIS), poor for DX. The same antenna on 20 m sits at λ/2 and has its maximum around 30°.

Ground

Real ground reflects worse than metal, especially at low angles. Under a horizontal dipole a bit gets lost that way, but the maximum stays roughly where it is. For a vertical the ground decides everything: over salt water its maximum is below 10°, over dry sand only at a good 25°.

The pattern here is simplified: antenna and mirror image, Fresnel reflection, without the ground's effect on the feed. For more detail with gain, feed point and azimuth pattern, use the antenna simulator.