In a cave we are currently exploring, there is a temporary siphon that sometimes blocks our way.

I was thinking about monitoring it (pressure sensor or other level sensor at the siphon, transmitter via mobile phone outside the cave.)

The problem is that there is more than half a kilometer between the entrance and the siphon (including some narrow passages,) so I was wondering if it is possible to transmit information over such a long distance (let's say 1km) with a SINGLE wire (ie no "ground" wire, but both ends can connect to earth.)

The requirements:

  • a single wire between the siphon and the mobile phone emitter (ie no ground wire, but each side can be connected to the soil ("earth"))
  • distance : 1km
  • transmission in single direction (from cave to outside)
  • very low bandwidth needed (I'm fine with one bit per minute)
  • using a thin wire would be best (lower cost and weight).
  • ideally not consuming too much power (both sides will be battery powered, I need to transmit 8bits once per hour)
  • one micro-controller on each side (+ sensor or 4G module) + the necessary electronics for transmission
  • only safe voltages/currents (ie no injury if insulation is damaged and someone touches the wire)

Do you know if something like that is feasible? If so, do you know where to start looking about information on how to build such a transmission (I have no idea how it might be called, so it's difficult to search for it.)

  • \$\begingroup\$ Comments are not for extended discussion; this conversation has been moved to chat. \$\endgroup\$
    – Voltage Spike
    Commented Jan 11, 2022 at 23:27
  • \$\begingroup\$ Single-wire ground return phones are used by cave rescue organisations. Check out the British Cave Research Associations (BCRA) Cave Radio & Electronics Group (CREG) Journal; a quick search for "single-wire" returns a few detailed results. Of particular interest might be this article on data transfer over single-wire ground return systems. \$\endgroup\$ Commented Jan 13, 2022 at 13:59
  • \$\begingroup\$ It seems like the best way to answer this question would be to start with the wire. Buy whatever single (or twisted pair) wire that you can get within your budget, and try sending some signals through it. \$\endgroup\$
    – kgutwin
    Commented Jan 13, 2022 at 16:07

9 Answers 9


Why a single wire? Why not a twisted pair? With a twisted pair (think "phone line") you can definitely get kilometer distance reliably and you can send power down the line as well, so that you only have to deal with batteries at the entrance to the cave. Splicing is easy with Scotchloks, which is an advantage over fiber. 300bps (and probably at least 9600bps) possible easily using 1970s technology (except "microcontroller").

Phone wires have been around for well over a hundred years. As noted in a comment, you can often find military or other surplus telephone wire extremely cheap. Not high enough quality for modern networks, but more than adequate for voice or low-speed data.

If the goal is to monitor a single open or close status then you may be able to avoid a complex system by literally opening /closing a switch on a pair of wires. If you can go do that with a passive device - float switch, bimetal thermostat etc. then you don't even need power at the far end. Changing batteries at the entrance is easy. If you miss a week or two due to weather or other reasons and can't change a remote battery then the system becomes useless.

  • \$\begingroup\$ Comments are not for extended discussion; this conversation has been moved to chat. \$\endgroup\$
    – Voltage Spike
    Commented Jan 11, 2022 at 23:28

Yes, you can do this. This is how early telegraphs and even telephones worked: a single wire per circuit with an Earth return.

Low data rate inside a cave (shielded from most interference) would seem to be an environment favorable to this approach.

  • 2
    \$\begingroup\$ When I read the question my initial reaction was exactly this, "Telegraphy?". The next logical step is to use no wires at all and call it "radiotelegraphy", and the rest is history! \$\endgroup\$
    – crasic
    Commented Jan 11, 2022 at 15:26
  • \$\begingroup\$ They didn't even use copper wire, or insulation, but fence wires \$\endgroup\$
    – Chris H
    Commented Jan 11, 2022 at 15:40
  • 5
    \$\begingroup\$ @crasic The trouble with wireless is that a cave is a lousy waveguide. \$\endgroup\$
    – John Doty
    Commented Jan 11, 2022 at 15:44
  • 1
    \$\begingroup\$ @JohnDoty indeed they did. That's probably more successful than it would be in a wet cave \$\endgroup\$
    – Chris H
    Commented Jan 11, 2022 at 16:34
  • 3
    \$\begingroup\$ en.wikipedia.org/wiki/Through-the-earth_mine_communications or Cave-Link cavelink.com/cl3x_neu/index.php/en \$\endgroup\$
    – D Duck
    Commented Jan 12, 2022 at 23:41

Outside the box solution: ditch the wire completely and build an electric cricket. Crickets chirp at different rates based on the ambient temperature. You can measure how often a cricket chirps and use that to fairly accurately calculate the temperature. You can use a similar principle to transmit your small data packets.

Your equipment inside the cave would have a piezoelectric speaker capable of generating loud, sharp chirps (similar to what you hear when you test a home fire alarm). Every time you take your hourly reading, the result is converted into a chirp pattern that is played out the speaker. Your equipment at the mouth of the cave will listen for these chirps, decode them, and forward you the results.

Audio is not normally a good candidate for long-distance communication because of how rapidly it attenuates with distance. In a cave environment, however, sound will echo off the walls and travel a lot farther than it will in open air. Your system will have to account for these echoes, but lowering your bit rate should compensate for it. For example, the chirps could be 10 seconds apart, with 2 chirps indicating "no siphon" and 4 chirps indicating "siphon present".

The variables here would be how well sound carries in your particular cave's geology, how noisy the ambient environment is at the mouth of the cave (and whether any noise is in the same frequency range as your chirps), and whether or not your chirping would disturb anything nearby (bats and neighbors would hate it). With judicious usage of the low-power sleep modes on your microcontrollers, your electric cricket could run for weeks off of a decent-sized battery pack.


Rather than a single wire, like the other answers I'd recommend twisted pair.

1km single twisted pair is cheap but quite fragile.

Cat5 isn't much more expensive. In fact it's pretty damn cheap due to being everywhere. Here's 305m for 50€. I'd just use that. It's pretty tough and can be reused for many other things, unlike the other more specialized options.

The extra wire pairs can also be used, for example as backup if one wire breaks... or you could even use them to install a telephone or intercom at the end of your cave, just in case.


Use a fiber optic cable instead. Same size as a single wire. More impervious of the elements. Lighter.

  • 9
    \$\begingroup\$ Good idea, but I'd be afraid it might get kinked and fail. You can't just strip the fiber and twist the ends together to make an emergency repair either. \$\endgroup\$
    – Transistor
    Commented Jan 10, 2022 at 23:22
  • 2
    \$\begingroup\$ I suppose whatever I use (fiber-optic cable or electric cable), it will brake "regularly" (excepted if using really thick/sold cables, but that will be too expensive for 1km). So I'd rather use electric cable that is easy to repair by hand. There are in particular 2 narrow passages were I expect thin cable to brake on nearly every caving day, so I was thinking to put thicker cables for those 2 passages (it's "just" 40m of strong cable instead of 1km) \$\endgroup\$
    – Sandro
    Commented Jan 10, 2022 at 23:45
  • 2
    \$\begingroup\$ @hacktastical power over fibre is a thing - basically put a laser down it; 10s of mW is trivial, 100s of mW possible. I've used fibre for laser delivery more than for comms in fact. A little care in the design would be needed as back-reflections of the power delivery would be greater than the signal coming from the sensor, but the sensor could transmit only when the line goes dark, with the power-supply laser being turned off to request a reading \$\endgroup\$
    – Chris H
    Commented Jan 11, 2022 at 13:57
  • 2
    \$\begingroup\$ @ChrisH So yeah... crazy expensive, complicated, and way overkill for a hobby project. \$\endgroup\$
    – J...
    Commented Jan 11, 2022 at 14:29
  • 1
    \$\begingroup\$ An optic float switch is certainly possible, but I would expect them to be expensive. No, wait! It is enough to dip the polished end of the fiber into the water on the high level, that would certainly change the reflection at the end of the fiber. \$\endgroup\$ Commented Jan 13, 2022 at 6:24

Probably not a lot off the shelf. The only single wire protocols I have seen in "common" use is 1-wire or LIN. Neither are quite what you want.

To get a signal from one spot to another I see three major variables:

  • Baud: Very slow signals are less likely to be misinterpreted
  • Power: Hard to miss a 5kV signal
  • Error Correction: If you only need half the bits to be heard, it's more likely

Swinging big voltages is harder than one might think so I'll focus on the other two.

Bitbang a GPIO, baud rate under 2400, something as simple as a Hamming(7,4) is probably good enough and you'll likely need a driver/receiver circuit (12-15v?). Season all of the variables to taste until you get the performance you're looking for.

  • \$\begingroup\$ Indeed, I was thinking about low baud rate. A baud rate of 1 bit/second is already enough for me. Error correction is also not a problem (I can waste quite a lot of bits, as I have so few data to transmit). The ideal would be using a signle Lipo 1S, so I can power the micro without any converter. But a second battery, or a boost converter turnd on only when transmitting, is also fine. Or using a 12V battery and a step down converter with very low power consumption when output current is low \$\endgroup\$
    – Sandro
    Commented Jan 11, 2022 at 22:42
  • \$\begingroup\$ Hamming(7,4) in practice can be a lookup table. Very simple. There is a tradeoff between baud and power usage, gotta drive the line. But surely a baud of 1 would make it, so there perfect answer probably lies between 1 and 4800 baud. You could try throwing 3.3 - 5v down the line but you may be fighting the resistance of the line at those voltages. \$\endgroup\$
    – foreverska
    Commented Jan 12, 2022 at 16:26

A single wire could certainly be used to communicate; back in the land-line telegraphy era, the signal was carried on a single wire. Each station on the line had a connection to ground, which was used to complete the circuit.

I suppose that a mobile phone could be used to send the data, but I'd think that a custom device based around a microcontroller such as an Arduino device, or based around a single-board computer such as a Raspberry Pi, would be more suitable.

Miners often communicate with VHF or UHF handheld radios. (Unfortunately I don't know what bands they use, or what distances are practical.) You might consider using radio rather than a wire. You would likely need more battery power, but you wouldn't depend on a fragile wire. The battery could still be made to last quite a long time if the transmissions were quick and infrequent.


about your specific case: the twisted pair should work well, however since you are in a cave you could also use sonic communication quite easily, like how submarines do it.

however as a answer on the general question

Yes you can communicate with one wire. you don't even need a "ground" note this method relies on a parasitic grounding effect which typically happens on places like earth. you can use the signal and a capacitor, instead of measuring the current state you measure the changes, so you measure it kind of like you do with a capacitor isolated circuit like in medical or audio equipment.

so when it is High first on when the capacitor charges it will send a positive pressure in the circuit, then it might go back to around 0, but that isn't bad, since even if it goes back to 0 there will be a negative pressure spike generated when the voltage of the signal is brought to low. so you get a pulse when the signal turns high and when it turns low

this will work well and you only need one wire and a capacitor to decouple it so you get the relative voltage change rather than the exact voltage.

however it doesn't end here because you can also use diodes to rectify the signal, you might not even need a capacitor that way, however you can use one to get a refference high and low voltage, when doing it like that you can even use most common communication protocols which need a ground.

protocol the data would based on the frequency, so you can either use the pulse length to determine if it is a 1 or 0 or if you monitor something simple like sound or a analog value then you can directly use that as the PWM, this is actually quite simple to how microphone amplifiers work, however now you have much more voltage and current and don't need a ground.

in short the most easy way to use a electric single wire for communication is by using the voltage change instead of the voltage itself by decoupling it with a transistor. if desired a reference voltage can be made using 2 diodes and a capacitor, since the diodes have a small voltage drop anything above the reference voltage high is high, and anything below the voltage drop low is low. for example while using 12V for communication the reference voltage for low would be around 0.7V and for high around 11.3V. this actually makes it more easy to work with. however all of that is rather unnecessary, since you don't need the reference voltage in most cases when you decouple it and measure the change instead of the actual value, in some minimalist circuits it might make it more easy however.

Disclaimer you should use higher voltage for this to work properly, it will kind of automatically use the air and the earth as a ground, so it wouldn't work as well in space, you can see it kind of like gravity lines or magnetic fields while in theory they shouldn't as easily interfere due to many factors they still will do so. with a magnet as a example when you rotate a flying magnet it should go back and the other magnet shouldn't notice this, however due to the magnets mass and air resistance or other resistance, so the arm holding the measuring magnet will measure a magnetic peak force. using a ground wire will make it a lot more stable and efficient however, it will give a much stronger signal. when sending signals through long wires you also need higher voltages typically, and due to internal resistance and capacitance it will become a lot harder for the described method to work properly, when you connect both grounds to the earth however instead of relying on the parasitic ground effect it will be a lot more effective.


I'd recommend high frequency alternating current RESONANT single wire power transmission through a carbon fiber thread, there are minimal wire losses since it's an electric field oscillation, as opposed to conventional conduction current (i.e. its a waveguide, not a conductor). One would need matched Tesla coils at each end, and a fast DC rectifier.

  • Would you WANT to do it? Doubtful
  • Is it worth mentioning? YES
  • Carbon fiber thread is cheap, lightweight, and strong
  • if it breaks, just tie it back together
  • I am able to do it at shorter distances, 1km shouldn't be an issue
  • See me demonstrate it: https://youtu.be/YYXr6vgef04 Schematic of single wire transmission line and RF-to-DC rectifier
  • \$\begingroup\$ Your youtube link appears to be broken. \$\endgroup\$ Commented Mar 22 at 23:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.