Feature − Radio at Depth (Part 2)

Mike Bedford, G4AEE, continues his examination of sub-surface communication by looking at alternatives to LF for cave radio, and at methods of communicating between underground locations.

Mike Bedford G4AEE continues his examination of sub-surface communication by looking at alternatives to LF for cave radio, and at methods of communicating between underground locations.

Cave radios, operating in the LF portion of the radio spectrum, can communicate to a depth of several hundred metres through limestone. This provides a valuable service for cave rescue teams in their quest to save those who get into difficulty underground, as we discovered in part one of our investigation into sub-surface radio ( PW September). This month, to conclude our introduction to this unusual and fascinating area of radio communication, we’ll address two further subjects. First, we’ll look at how different frequency bands offer several potential benefits compared to LF. And second, we’ll delve into alternatives to through-the-earth radio for communicating between two underground parties.

VLF Cave Radio and Below

Last time we saw, in passing, how the higher conductivity of coal measures, compared to limestone, means that VLF or ULF is needed, instead of LF, to provide emergency communication in collieries. The negative impact of geology first manifested itself to cave radio enthusiasts following an invitation, by Network Rail, to provide advice on communication in a long-disused railway tunnel for search and rescue operations. While recognising that this area of the Pennines certainly wasn’t limestone country, having used LF cave radios successfully in an archaeological copper mine in Cheshire, an old tin mine in Cornwall and a working gypsum mine in Staffordshire, the team was reasonably confident of success at the Standedge Tunnel, despite the different geology. How wrong they were. Using an 87kHz HeyPhone, which was capable of operating through 1,000m of limestone, a surface station tracking an underground station from the East Portal lost contact just a few hundred metres into the 4.75km tunnel, at which point the ground cover was less than 100m. Perusing a geological survey map revealed alternating layers of gritstone and shale, the latter being several orders of magnitude more conductive than limestone. Clearly a considerably lower frequency would be needed.

Prompted, in part, by amateur experimentation at VLF and below, and with the motivation of providing communication where LF fails, cave radio experimenters have recently turned their attention to the bottom-end of the radio spectrum. An LF cave radio can make do with a comparatively small loop antenna but, down at VLF, ULF or SLF, loops would have to be huge to achieve resonance, and this hampered the very early experiments into baseband cave radio. Using a ground

ed dipole, however, which is now standard procedure at LF, all this changes and a portable rig becomes feasible. Also, very different from the early days of cave radio is the transmitter and receiver circuitry. A software defined radio approach is being pursued and minimal hardware and computing power is needed. In particular, Raspberry Pi hardware is being used and the transmitter and receiver are being implemented using the GNU Radio Companion, which allows the architecture to be defined by wiring up functional blocks on screen. The current prototype is an SSB radio at 9kHz, very different from amateur experimentation at VLF. This is feasible, though, because of the use of magnetic induction instead of true radiation, thereby offering negligible potential for interference to the very narrow bandwidths normally used at VLF.

Going Higher

It has long been believed by cave radio enthusiasts that a frequency in the LF part of the spectrum is needed to achieve adequate signal penetration through the rock. But things aren’t quite so simple. While the attenuation increases with conductivity and frequency, it’s not correct to believe that the relationship is always one of direct proportionality. Attenuation is only directly proportional to frequency if the rock is considered a good conductor. We’re not going to get embroiled in the maths here, but let’s just say that rock ceases to be a good conductor as the frequency increases and this causes a graph of attenuation against frequency to start to flatten off. The bottom line is that higher frequencies aren’t attenuated as much as the naïve theory predicts, and in limestone this suggests that frequencies as high as a few megahertz could be effective. It certainly won’t be true, for example, that a 3.5MHz signal would be subject to 40 times the attenuation of one at 87kHz, as the ratio of those two frequencies might suggest.

With this in mind, radio amateurs have been engaged in HF cave radio experimentation. After all, there are several potential benefits, as well as some unique challenges. A major difference between LF and HF, especially in a constrained cave environment where antenna sizes are severely limited, is that an LF cave radio operates by induction, while at HF true radiation will occur. This, in turn, means that attenuation follows an inverse square relationship, not the inverse cube law that so much limits the range of LF cave radios. But if the purpose of a cave radio is just to provide a link between an underground party and any point on the surface, does an increased range really offer any practical benefits? Actually, we can envisage a couple of advantages. Imagine, if you will, a cave rescue team member given the job of remaining in touch with the underground party, which requires him to spend hours on a windy hillside, at night, in a snowstorm. Now, imag

ine that same surface operator, this time in a warm vehicle on the closest road. The difference between these two scenarios could well be the difference between using an LF and an HF cave radio. The second benefit of a greater range is to provide cavers with warnings of adverse weather that could result in underground flooding. Some expeditions do indeed remain in contact with underground parties but, realistically, it’s just not going to be feasible for cavers to regularly be in contact with a surface party which, at LF, would have to be in the immediate vicinity. At HF, on the other hand, while the regulatory issues would need to be addressed, from a purely technical viewpoint, we can imagine a regional warning broadcast scheme.

So, it appears that there might be benefits, but is through-the-earth radio really feasible at HF? Current experimentation suggests that it most certainly is. Working at 3.5MHz and 7MHz, radio amateurs have been able to replicate the through-rock performance of LF cave radios, but the surface station doesn’t have to be rooted to the spot, pretty much vertically above the underground party. In early tests, a team one hundred metres underground was able to remain in voice contact with a mobile surface station as they drove away from the cave to a distance of over five kilometres. Using slow speed data, however, much more impressive results were achieved, especially on the downlink. However, since the purpose of this work is to offer real-world benefits, slow speed Morse at one word per fortnight was dismissed in favour of WSPR for initial tests. Operating on 7MHz in the Valley Entrance of Kingsdale Master Cave in the Yorkshire Dales, a station 100m below the surface was able to receive 17 stations in nine countries, during a period of about 30 minutes, the most distant being in Sweden at a range of 1,327km. No reception of signals transmitted from this location was reported. However, better uplink performance was achieved from another North Yorkshire location, Short Drop Cave, even though there was less rock above the underground station. In particular, reports were received from stations in Liverpool, London and Southampton, the latter being at a distance of 370km. However, the most impressive result, to date, was from another Yorkshire Dales station, this time established in Yordas Cave. To be honest, in being a huge cave chamber, capable of housing a full-sized 80m inverted-V, this cave is by no means typical. It is also easily accessible from the surface, and it’s only a few tens of metres underground. However, underground it is, and an SSB contact was made with a station in Ipswich, at a range of 340km.

Along-Passage Communication

As an alternative to through-the-earth radio, other methods are available for communicating between a cave and the surface or between two underground locations.

What’s more, some of them have rather more in common with ordinary radio. Hang Son Doong in Vietnam is the world’s largest known cave. The main passage is over 5km long with parts reaching up to 200m tall and 150m wide. Some of the huge chambers have collapsed roofs, thereby allowing tropical jungle to become established underground. In 2015, Swedish photographer Martin Edström led a 50-member expedition to undertake a photographic project on behalf of National Geographic. With members of the party who handled the several immensely powerful lights needed to illuminate these vast spaces being several hundred metres from the core photography team, communication was essential. Handheld VHF or UHF radios provided excellent service, with propagation in large cave chambers providing the same line-of-sight performance as in open air. Propagation along cave passages, on the other hand, is entirely different.

You’ll recall that through-the-earth LF cave radios failed to work satisfactorily at the Standedge railway tunnel. The request by Network Rail to investigate communication at this location was prompted by their observation that VHF radios had a range of just 300m in the tunnel, and while this increased to 1km with UHF PMR 446 radios, this wasn’t nearly sufficient to provide communication throughout the tunnel, even with an operator at both ends. The solution eventually offered by cave radio experimenters was a microwave link, following a demonstration that 23cm (1.3GHz) amateur radio equipment could communicate from end to end through the entire 4.75km of the tunnel. So, with this example of how microwaves can provide communication along a railway tunnel, and with the question of whether the same can be achieved along a cave passage, it’s time to take a look at the theory of tunnel communication.

Some Theory

Waveguides are the hollow metal tubes that are used as feeders instead of coax for use with high power microwave transmitters. A key characteristic of a waveguide is its cutoff frequency, which depends on its crosssectional dimensions, and is a measure of

1

2 Fig. 1: Yordas Cave isn’t typical, in allowing a full-sized 80m inverted-V to be erected, nor is it very far underground, but being able to make a 2 x SSB contact with a station 340km away was remarkable. (Ian Cooper) Fig. 2: Rob Gill G8DS and Robin Gape G8DQX are experimenting with the use of a Raspberry Pi as a software defined SSB radio operating on 9kHz. (Rob Gill) Fig. 3: Attenuation doesn’t continue to rise with frequency but starts to flatten off as the rock becomes a poor conductor at higher frequencies. Fig. 4: HF antennas are much shorter than usual underground, because of ground proximity, but they still pose a challenge in confined cave passages. (Gregory Collins)

4

3

6 Fig. 5: If you’re building equipment that’s going to be used underground, it has to be designed to take some knocks, oh, and a good soaking too. (Mike Bedford) Fig. 6: HF cave radio provides much greater coverage than LF. Red: Stations received from 100m underground. Blue: Reports received for signals transmitted from 20m underground. (Map Data ©2016 Geo-Basis/DE-BKG (©2009), Google Inst. Geogr. Nacional) Fig. 7: Following successful tests at 1.3GHz in Standedge Tunnel, where VHF and UHF had failed, tests with 802.11g WiFi equipment, at 2.4GHz, provided data transfer over the entire 4.75km, compared to 350m above ground. (Mike Bedford) Fig. 8: Juan Corrin with the drone he adapted to explore areas of caves that are otherwise only accessible via a potentially fruitless climb. (Mike Bedford)

8 7