Echos from the past, 150 ms ago

Foundations of Amateur Radio

Recently Glynn VK6PAW and I had the opportunity to play radio. This isn't something that happens often so we try to make the most of it. For our efforts we had plenty of frustrations, to the point where we were joking that I should rename this to "Frustrations of Amateur Radio".

That was until we heard something weird on-air. All setup shenanigans forgotten, we marvelled at the experience.

I was playing around on the 10m band, trying to hear people making noise and potentially our first contact for the field day we were participating in, when I heard something odd. Two stations talking to each other, but the audio was strange. It was like they were doubling up, the same audio played a fraction of a second later, until that moment, something I've only ever heard in a radio studio whilst editing using a reel-to-reel tape machine with separate recording and playback heads.

Having just started using a digital only radio, at first I thought this was an artefact of the radio. I took note of the frequency, 28.460 MHz and told Glynn about it. After we moved the telescopic vertical antenna to the analogue radio, we discovered that this was in fact real, not caused by the radios, no doubt a relief to the proud owner of both radios, Glynn, who was thinking more clearly than I. He took note of the callsigns, Dom VK2HJ and Yukiharu JE1CSW. Looking back now, an audio recording would have been helpful.

At the time I suggested that this might be a case of long path and short path signals arriving at our station and being able to hear both. If you're not sure what that means, when you transmit, an antenna essentially radiates in all directions and signals travel all over the globe. Some head directly towards your destination, the short path, others head in exactly the opposite direction, taking the long way around Earth, the long path.

You might think that the majority of contacts are made using the short path, but it regularly happens the other way around, where the long path is heard and the short path is not. As you might know, radio waves essentially bounce up and down between the ionosphere and Earth and it might happen that the signal arrives at the destination antenna, or it might happen that it bounces right over the top, making either short path or long path heard, or not.

In this case, both arrived clearly audible.

It wasn't until I sat down on the couch afterwards with a calculator that I was able to at least prove to my own satisfaction that this is what we heard.

So, what were those calculations and what was the delay?

The circumference of Earth is roughly 40,000 km. RF propagation travels at the speed of light, or about 300,000 km/s. It takes about 0.13 seconds or 130 milliseconds for a radio signal to travel around Earth.

At this point you might realise that 40,000 km is measured at the surface, but ionospheric propagation happens in the ionosphere, making the circumference at the very top of the ionosphere about 45,000 km, which would take 150 ms.

There are several things that need to line up for this all to work.

Propagation aside, the distance between all three stations needs to be such that the number of hops between each combination is a whole number so we can all hear each other.

As it happens, the distance between Perth in Western Australia and Maebashi City in Japan is pretty close to the distance between Goulburn in New South Wales and Japan, and the distance between Goulburn and Perth is roughly half that.

Using back of napkin trigonometry, it appears that 27 hops around the planet are required to make this happen. That's five hops between Perth and Japan, and between Goulburn and Japan, and two hops between Goulburn and Perth, and 27 hops between Perth and Japan the long way around.

Given that the F2 layer where the 10m signal is refracted exists between about 220 km and 800 km, we can estimate that the total delay for the long path is at least 144 ms.

That doesn't really translate into anything you might relate to, but at 8 wpm a Morse code dit takes 150 milliseconds, which gives you a sense of how long the echo delay is. In other words, it's something that you can absolutely hear without needing to measure it.

There are other implications.

WSPR signals are used to test weak signal propagation. Stations around the globe report on what they can hear and when. For this to work, the signal need to be synchronised, something which is commonly implemented using something called NTP, or Network Time Protocol. It can achieve a time accuracy of 10 ms. GPS locked WSPR beacons can achieve an accuracy of 40 nanoseconds.

In other words, if we know that the beacon and the receiver are time synchronised, we can probably detect if the signal arrived using a short path or a long path. The WSPR decoder tracks the time between when the signal arrived and 2 seconds past an even minute as perceived by the receiver. Gwyn G3ZIL wrote an interesting document called "Timescale wsprdaemon database queries V2" on the subject of the data format used by wsprdaemon, a tool used to analyse WSPR beacon transmissions. If this is something you want to play with, check out wsprdaemon.org

From our adventures there was plenty to take away. Stay curious, go portable, take notes, practice putting up an antenna, keep a log, laugh and have fun, and last but not least, get on air and make noise.

Before I forget, make sure your mate brings a pen for logging when your own trusty scribble stick suddenly gives up the ghost for no apparent reason. I knew there was a reason I prefer pencils.

I'm Onno VK6FLAB