sending data over one wire (physically one wire, not the "One wire protocol")

[ikalogic]

Hi again!

I was thinking about that concept:

RF is sent using zero wire.
All other communication means i know of send data using at least 2 wires (the ground wire being often not counted).

Is there a way to send just a minimal pulse from one circuit, to be picked up by another circuit, and the only link between them being a single wire?

I feel that is a solution exist, it would be related to some capacitors being charged and discharged..

I think this can have many applications, although i still don't have an exact application right now..

any hints on where to start looking...?


ps: I am not looking for a specific IC that does the job for me, i am trying to understand how such a system would work, on the simplest possible level.

Thanks a lot !

 

[ikalogic]

Just to clarify, i am not talking about single wire earth return.

I am talking about implementing such a system on low voltage (typically 5V) systems.

 

[Ian Rogers]

On my system there is a feature that requires two wire hook up.... Its called an "ATB System" Anti Two Blocking system... We need to use a very strong cable... We use a basic co-axial cable but with a reinforced galvanized inner core with a copper braid... I send several signals down the cable using the RS232 protocol ( one way ) using a 12v rail and a micro at the top regulated by a 5v regulator... Then I use 12 - 7 v signalling and using a current detector at the bottom to re-create the RS232 signal. If you do it this way you can maintain the 5V rail at the top..

Works well for me..

 

[crutschow]

Any wire sending a electrical signal requires a return. You could use one connection with RF but that still uses the earth as a return.

If you use a dielectric signal conductor, such as a fiber optic cable, then you only need one.

 

[MikeMl]

How about heating one end of a metallic conductor and waiting for the heat to transfer to the other end

How about stretching the wire between two posts, and flipping it, and watching for the wave to arrive

How about tying a string between two cans, pulling it tight, holding one up to the mouth and the other to the ear...

 

[Ian Rogers]

Thinking about this again..... If you used some sort of preamble at the beginning of the communication, you could work out the differential of the signal...

If you sampled the voltage first then read the signal.... The only trouble, as you well know, the receiver could be at a higher potential than the transmitter....

 

[ikalogic]

Thank all for your replies.

I have to resign to the fact that there may be no solution to send a simple pulse using only 1 wire.

I just find it frustrating that it will work in RF (using AIR as a medium) and not with a wire... i wish i payed more attention in my physics classes...

 

[Mr RB]

Any wire sending a electrical signal requires a return. You could use one connection with RF but that still uses the earth as a return.
...

I'll argue that. One of the first cool things I made with a electronics kit as a boy was a 1 wire intercom.

It started off as the 2 wire intercom example from the book, but I cranked the gain of the audio receiver amp right up, and it still worked with only 1 wire connecting the 2 kits (no earth, both powered from a 9v battery).

So it can definitely be done, all you need is to send the signal down the wire at a few volts and have a high gain audio amp at the receiver end.

A tone will be much easier to detect than a single pulse.

 

[dr pepper]

I've looked into something similar.
If you want to only use 5v then your probably not going to get far.
If you use something that has high voltages then theres a chance it'll work, you can send high volatge rf along 1 wire and use a metal plate or wire as a ground plane for the invisible 'other' wire.
Taken to the extreme theres a guy on youtube who does this with an auto ignition coil, theres enough energy at the receiving end to run an rc plane motor.

 

[Ramussons]

I've looked into something similar.
If you want to only use 5v then your probably not going to get far.
If you use something that has high voltages then theres a chance it'll work, you can send high volatge rf along 1 wire and use a metal plate or wire as a ground plane for the invisible 'other' wire.
Taken to the extreme theres a guy on youtube who does this with an auto ignition coil, theres enough energy at the receiving end to run an rc plane motor.

ikalogic says he does'nt want an Earth Return. Else we could go back to the Telegraph of 1900's!

Ramesh

 

[kubeek]

I think that using a single wire should be possible, but not very easy.
If you have an object electrostatically charged then if you connect it by a single wire to another object that is at neutral potential the charges will balance and there will be current flowing through the wire for a brief time.
Now what the current will be will depend on the capacitance of the the sources, wire length and charge voltage, and I think this scheme would be quite susceptible to noise (you could use some shielding, but then we are back to the much simpler two wires).

 

[alec_t]

But for that to work the object terminals not connected by the wire would have to be at a certain potential difference (which could be zero). So if the two objects are spaced apart then another conductor (e.g. the ground) would be needed between those terminals to establish that potential difference. Any flow of charge which can be sensed externally of an object requires a closed circuit.

 

[kubeek]

I beg to differ, I imagine this object as a simple metal sphere with only one terminal prefecty insulated from everything else, like for example the ball on top of a Van der Graaf generator. This ball gets stripped of its electrons and becomes positively electrostatically charged (ionized). Now imagine a second generator like this, but one which adds electrons to the ball, so it becomes negatively charged (ionized).
If you leave the two balls just hanging, they should maintain their charge indefinitely. Now if you connect them together with a conductor the electrons will flow to the place with lower concentration, and you will be able to meausre current in the conductor even though there is no closed circuit loop.

You could also imagine it as a cloud to cloud lightning - again no closed circuit needed.

 

[Mr RB]

But for that to work the object terminals not connected by the wire would have to be at a certain potential difference (which could be zero). So if the two objects are spaced apart then another conductor (e.g. the ground) would be needed between those terminals to establish that potential difference. Any flow of charge which can be sensed externally of an object requires a closed circuit.

I'll argue that too. You only need a closed circuit if you want to get any significant amount of current.

You can get a signal across one wire perfectly fine, it's just a matter of receiver gain.

As an example you can wear rubber shoes and make sure you are not grounded in any way, then touch a fingertip to the AC mains live wire. You will feel a "signal" no problems.

The only way you can get no current over a single wire with an AC signal is if the receiving end is infinitely small and can change its average potential instantly with no current required. If the receiving end (device) has any significant mass or capacitance to the environment then tiny AC currents will be present over the single wire.

In my case I used a transmitted signal of a few volts AC, and an audio amp at the other end with a gain of a few hundred. Audio comms over the single wire worked fine. Broken wire sniffers as used in telecoms work also just use a high gain amp to detect an audio tone in a broken wire.

For this to work all the OP needs is to inject a tone into the wire, and have an opamp and a LM567 tone detector at the other end.

 

[crutschow]

I'll argue that. One of the first cool things I made with a electronics kit as a boy was a 1 wire intercom.

It started off as the 2 wire intercom example from the book, but I cranked the gain of the audio receiver amp right up, and it still worked with only 1 wire connecting the 2 kits (no earth, both powered from a 9v battery).

So it can definitely be done, all you need is to send the signal down the wire at a few volts and have a high gain audio amp at the receiver end.

A tone will be much easier to detect than a single pulse.

And I'll argue that the return is through the capacitance to ground of the two circuits, small though it may be. You can't detect a voltage difference between two remote devices without some sort of signal return.

 

[dr pepper]

I agree with both, you need a send and a return, however the return can be through air or ground, and if the rx doesnt need much power or the voltage is high it can be a very high resistance path.

This isnt my kind of thing as its surrounded by silly theories and free energy, however this vid shows someone transmitting a surprising amount of energy with a car ignition coil over one wire.
As I mentioned before the 'return' is via a metal plate.
The geyser doesnt say what the actual freq is, no matter there will probably be an amount radiated as rf from the harmonics which might be an issue, but it does show how effective such a simple circuit can be, it can run an rc motor and prop, and power 2 21w lamps accross one wire over a few meters.
A much lower voltage tone burst could probably be used to transmit data to/from a logic gate.

https://www.youtube.com/watch?v=Qz1-RIcj1HY

 

[RCinFLA]

Single wire waveguided transmission line. "Goubau line "

 

[dr pepper]

The maths in G lines make my head hurt, they use 'launchers' and single core copper conductors which behave not too diffrently from co-ax.
I think we've gone a little silly now.