Finding broken wires underground

[Q_Goest]

This was pretty amazing. We have an ‘invisible fence’ for the dogs. I installed it myself years ago. It’s just a single strand of wire that wraps around our entire yard that must transmit some kind of signal to the dog’s collar so when they get close to it, the collar starts beeping. If the dog decides they want to ignore the warning and they go closer to the wire, it sends out a jolt of electricity that would fry a cow.

I know it will fry a cow because I tried it on my finger. It took a good 5 minutes to get any feeling back.

Last fall the little box that sits inside the house and to which the wire is connected, started beeping. That means the wire was cut. Second time that happened… First time was at this spot where the wire had been sticking out of the ground for years. I was too lazy to fix it until it broke. Needless to say, that one was easy to find.

So I ask you… how do you find where a broken wire that’s underground, running for hundreds of yards? I don’t know of any way to do it, so I called my brother in law who’s a brilliant electrician. Apparently, there’s some kind of device used to do this but he didn’t have one and didn’t know much about it.

Back to square one.

Not wanting to put a whole new wire in (an all day affair if I had the right tools which I don’t) I searched the internet for a solution and found this:
http://www.instructables.com/id/How-to-find-a-break-in-an-invisible-dog-fence/?ALLSTEPS

They tell you to buy an "RF Choke" from Radio Shack and install it as they show in the picture. It only cost $1.50 so I had nothing to lose. Then I duct taped a small battery operated AM radio to the end of a broom handle, tuned it to the right frequency, and walked around the yard sweeping this thing across the ground like one of those metal detectors.

It worked great! I couldn't believe it was so easy. I’m telling you all this so you know how to fix an invisible fence in the future, so please take notes.

The radio made a distinct, static noise as I waved it just inches off the ground, but only when it was directly above the wire. The radio went virtually silent if more than a few inches to either side of the wire.

Once I got close to the break, the signal gradually disappeared. I thought it might only be because of all the tree branches and things in the area so I walked past the dead area, about 40 feet away, and came up from the opposite direction. Sure enough the signal stopped about 20 feet away from where I lost the signal coming from the opposite direction. I marked the two spots and started clearing away the brush and found a ground hog hole. The ends of the wire were sticking up out of it. Apparently the ground hog had dug his home, busted the wire and neglected to repair it. Thoughtless bugger.

What amazed me though was how this tiny AM radio was able to pick up a signal from the wire. It was VERY accurate. If the radio wasn't within a few inches to either side, it went silent. I’m not sure why the signal fades 10 feet away from the break in the wire. I think it’s either magic, or it has something to do with the wavelength the wire transmits at. Not sure which but it’s probably something the CIA, NSA and some of the folks here know all about. As you can see, I’m easily amazed…

So tell me something – why did the signal fade away about 10 feet from the break and how could you calculate that?

And how does that RF choke work?

Any other insights to this device?

 

[AlephZero]

An RF choke is just an inductor, often wound on a ferrite core. They are usually used to suppress radio frequency signals while transmitting DC current. These days you are more likely to find a ferrite bead on the wire used for the same function, since it's much smaller and less fragile

I suppose the idea works because the choke has the about right impedance at radio frequencies to fool the transmitter into thinking it is connected to a big loop of wire. The two broken wires then act like random-length radio antennas.

Possibly the reason the signal faded out was that the wires either side of the break are acting like two antennas pointing directly at each other, and transmitting signals 180 degrees out of phase, so near the break the signals cancel out. Away from the break, the signal from the wire you are near is much stronger than the other one. The inverse square law probably applies.

Using an AM radio as a cheap RF signal detector is an old (i.e. 50 years or more old) fault finding trick.

 

[Simon Bridge]

If you send a wave down the wire, it will reflect off the broken end.
The resulting interference and some deduction, can tell you how far to the break.
It's been a while since I've done this - it's standard for studying power transmission.

The radio trick is good - it appears to be using the wire as an antenna to transmit a radio signal.

Very loosely:
When you get close to the end of an antenna, the radiation has another direction to go in - straight out the tip.
You are only detecting the part that goes radially from the wire so the strength falls off as you approach the tip.
How it falls off depends on the exact shape of the wire, and the radio receiver - which may screen out signals below a certain strength.

OR

Possibly the reason the signal faded out was that the wires either side of the break are acting like two antennas pointing directly at each other, and transmitting signals 180 degrees out of phase, so near the break the signals cancel out. Away from the break, the signal from the wire you are near is much stronger than the other one.

:D

You want more details - then you know to look up stuff about antennas.

 

[meBigGuy]

According to the article the RF choke fools the unit into thinking the wire is not broken so it will send out the RF signal. The choke provides a DC path, but doesn't short out the RF.

 

[OCR]

So I ask you… how do you find where a broken wire that’s underground, running for hundreds of yards? I don’t know of any way to do it, so I called my brother in law who’s a brilliant electrician. Apparently, there’s some kind of device used to do this but he didn’t have one and didn’t know much about it.

Here's one device...

http://en.wikipedia.org/wiki/Time-domain_reflectometer

 

[AlephZero]

Actually a TDR doesn't find where the break is. It only tells you how long the broken piece of wire is.

I once had a telephone engineer messing around with a very expensive TDR for a whole day, trying to figure out why my landline phone wasn't working. The reason it took so long was, his wiring diagram said the phone was connected to junction box 200 yards east of my house. The wiring actually started out going west, then did two 90 degree turns, and was connected to a box about 500 yards away on another street.

They eventually figured this out by the low-tech method of digging holes and tracing the wires. Luckily, since all the problem was outside of my property, it didn't cost me anything, except the amount of disruption!

 

[OCR]

Actually a TDR doesn't find where the break is.

That's good to know, thank you.

 

[Baluncore]

It only tells you how long the broken piece of wire is.

But that length is measured in time and so you will need an educated guess at the velocity factor.

There are proportional ways of finding faults in lines and twisted pairs. Breaks can be found by measuring the capacitance from both ends. The break is located at a proportional position C1/(C1+C2) along the line. Likewise, finding a short in a cable can be facilitated by making line resistance measurements from both ends.

By having several identical lines available, the Murray and Varley loop techniques can use a resistance ratio measured from one end. They also benefit from accurate balanced bridge techniques. http://en.wikipedia.org/wiki/Murray_loop_bridge

 

[sophiecentaur]

Another possible way to locate a break would be to measure the capacitance between the line and ground - from each end. You could expect something like 60pF per metre but you wouldn't need to know that if you could measure from each end. The ratio of capacitances would give the ratio of distances between the ends and the break. Actually, that method would definitely be 'do-able' even if all you had was a DMM with a Capacitance measuring setting. You just need to have access to each end of the wire and connect directly to the ends and to Earth in a consistent way.

There is a more expensive way but not as flashy as TDR. That involves using a variable RF (UHF /VHF) source and looking at the VSWR over a range of frequencies. The frequency spacing between Maxes and Mins is an indication of the length of the line to the mis-termination. Rhode and Schwartz used to make a beast that did this, before the step recovery diode and sampling had been invented. I cannot remember the name of it. I think this method is what Simon Bridges is referring to.

If you don't want to do it with RF, you can start with a hole, half way along, find which side the break is (testing for continuity) and further divide that half into halves. I reckon that you could locate the break with very few holes.

Next time you bury a wire, you'll have to lay it in plastic drain pipe and that should discourage little beasts.

 

[dlgoff]

So I ask you… how do you find where a broken wire that’s underground, running for hundreds of yards? I don’t know of any way to do it, so I called my brother in law who’s a brilliant electrician. Apparently, there’s some kind of device used to do this but he didn’t have one and didn’t know much about it.

He may have been thinking of something like this Fault Finder.

 

[OCR]

A device called a "Thumper" might also be used... I've actually watched the REA (RUS) use one, in our area...

http://www.hvinc.com/cds.html

About a third of the way down...

http://electrical-engineering-portal.com/how-to-locate-underground-faults-on-cable

 

[Q_Goest]

Wait, wait, wait… We’re running off in the wrong direction here. I realize there’s something out there that pro’s use to find cables and the discussion is interesting and all, but I’m curious about how this whole mess works. Yes, the RF choke does something to the circuit but I don’t understand what. Yes, it is made out of a stick of iron with wire wrapped around it. It’s an inductor, but why? Why use an inductor? What is it doing? Why wouldn’t a resistor work? Or a capacitor? Or a piece of wire? Why wouldn’t it work if a moron like me stuck his two fingers in the electrical sockets? What is that thing doing that I couldn’t do if I licked my fingers?

The other thing I don’t get is why the signal fades about 10 feet away from the break. Simon Bridge, you mention that the radiation goes out the tip. I’m sorry, but I still don’t get it… Why would this occur 10 feet away from the break? The radio is tuned to roughly 600 khz. I wonder if that particular frequency is a function of this particular RF choke and perhaps also related to the 10 feet?

So I’m envisioning an electrical wave which propagates away from the wire like a wave on the ocean in all directions like the diagram below. Except perhaps it has a preferred direction, 90 degrees to the axis of the wire, which might have something to do with the 10 feet that doesn't transmit at the end?

I am also under the impression that the wire has some electrical potential along its length which changes from + to – and back to + again at 600 khz which creates this electrical field that’s propagating outwards with the magnetic wave. Is the electrical potential along the length of the wire linearly proportional to the length of wire? Or did I drink too much beer in college?

 

[meBigGuy]

The unit must have a circuit that does not effectively drive the loop unless there is a DC path and/or a non-infinite RF impedance. The choke provides a temporary path so the unit will work. The two antennas (now that they are not part of a loop) then have RF on them. I have no idea how they designed the driver, but it probably resonates the loop. The inductor looks like a working loop.

As for why the signal drops off as you get close to the break, it could be because you are nearing the point where two antennas driven with opposite polarity come close to each other and the fields cancel. You can test this by disconnecting one of them and see what happens as you near the break.

 

[Baluncore]

The other thing I don’t get is why the signal fades about 10 feet away from the break.

Underground cables at RF are very lossy transmission lines. In wet soils, (dielectric constant near 80), the electric fields are much more highly attenuated than magnetic fields. You are detecting the magnetic field due to the current flowing in the broken conductor. That current is flowing along the conductor to charge the high line capacitance. As you approach the break, the capacitance falls towards zero and so the current falls and the local magnetic field falls. At some point the magnetic field falls below the level detectable by the radio receiver with it's (ferrite rod?) antenna. In your case that is about 10 feet from the end.

 

[sophiecentaur]

Wait, wait, wait… We’re running off in the wrong direction here.

This is PF, remember. :big grin:

There are two issues here. The device was designed with a monitor facility to let you know when the line is broken. A very sensible feature as you wouldn't wan't pooch to get out with our your knowledge. The choke fools it into thinking everything is fine, whether the line is broken or not. Imo, that is not a good 'repair' strategy.

The line is a transmission line which is terminated at the far end, to ensure that there is current along the whole length. The impedance of the line is fairly low so the magnetic field will dominate (hence the ferrite rod works well as a detector). The terminating load will ensure that there is no excessive reflection at the end of the line - avoiding a 'node' at the end, and ensuring that there is current along the whole length.
There is an ancient form of directional MF receiving antenna, called the Beverage Antenna, which consists of a long wire along the ground. This link describes the basics, which also apply to the buried line of your system. Although the design was initially for a long wire suspended on poles, the Beverage antenna works to an extent when just draped along the ground or even under it. It's a very inefficient transmitting antenna but works very well on receive and it much more directive than any other design of similar cost. (Whimsical note: the original paper on the design was presented by Beverage, Rice and Kellog in the twenties, affair. A breakfast antenna, perhaps.)

 

[meBigGuy]

Wow sophie, you sure miss the point --- often. You should carefully read the OP again and try to figure out why the choke was added. Also, you are *way* off about the antenna. It's just a resonated loop. When it breaks, it isn't a loop anymore.

Baluncore: You know more about these things than me, but I'm real interested in what happens when he disconnects one of the broken leads and measures the signal near the break again. I think with with your model there would be no change.

 

[sophiecentaur]

There would be a change as the loop becomes a form of folded dipole, fed at some arbitrary point. (It didn't seem explicit that the wire is connected as a loop - balance fed.) The impedance at the 'feed point' could be anything with a broken wire. If you actually disconnect one of the wires, you will have an end fed monopole, which has effectively an infinite impedance so no power can get into it. The only mode for power to get into one leg would be unbalanced and would involve a deliberate or chance (capacitative ?)connection of the other amp output to earth.
Unless there is a user adjustment, the loop cannot be "resonant" unless its route is explicitly a circle (known inductance).

 

[Baluncore]

Without some idea of the waveform and pulse repetition frequency it is hard to predict behaviour.
My guess is that the line is driven as a current loop at a frequency well below resonance.

The AM radio is probably picking up the broad band “key click” transients at the pulse edges.

 

[Q_Goest]

This is PF, remember.

Yes, I have my cat hearding license from the ASPCA. They teach you that constant course corrections are needed to keep any thread on track!

Perhaps if I shoot out observational blurbs, something will ring a bell, so here goes..

Ok, I don't know what attenuated, folded dipole, pulse repetition blah bla bla means. I'll have to look those up... But here's another observation...

I know if I don't have the radio tuned to the right frequency, around 600 khz, it won't work. Just a tiny bit above or below that and the radio doesn't pick up a signal. Why?

It must have something to do with the RF choke and I suspect something other than this particular RF choke won't work. Radio Shack has lots of different RF chokes, and I haven't tried any others, but I suspect there's a reason the instructions say to use this particular one and it has something to do with the frequency the wire will transmit at. In fact, this particular RF choke and this particular AM frequency is intended for use on ANY underground wire, so I don't think claims about soil capacitence or wire length have anything to do with how this works. Further, I suspect this particular RF choke has something to do with the ability to detect the wire near the break. Perhaps this particular choke produces a 10' gap while other chokes produce a larger gap.

 

[AlephZero]

I know if I don't have the radio tuned to the right frequency, around 600 khz, it won't work. Just a tiny bit above or below that and the radio doesn't pick up a signal. Why?

From your first OP, the system is effectively a radio transmitter with the wire loop as the aerial, and a receiver in the dog collar. The collar is presumably activated by the strength of the signal received, i.e the distance from the wire.

So you have to tune the radio to the transmitter frequency to pick up the signal. If the sound is just "static", it is probably transmitting an unmodulated signal (constant amplitude and frequency). Radios are usually designed to mute the audio output when there is no signal received. When the radio is tuned to the signal, that noise suppression is switched off. The static noise you hear is the random noise generated by the radio's circuits, unrelated to what is being transmitted.

I don't expect the exact type of RF choke is critical, but giving a Radio Shack part number is convenient for people with no background knowledge. Anything with an inductance the same order of magnitude as the unbroken wire loop would probably work fine.

 

[Q_Goest]

From your first OP, the system is effectively a radio transmitter with the wire loop as the aerial, and a receiver in the dog collar. The collar is presumably activated by the strength of the signal received, i.e the distance from the wire.

So you have to tune the radio to the transmitter frequency to pick up the signal.

Are you saying you believe the 600 khz AM signal is what the system normally puts out and is picked up by the dog collars? If that's true, then I should be able to pick it up on my AM radio now that it's repaired, right?

 

[sophiecentaur]

1. What do you actually hear on your mf receiver? I imagine the box produces pulses (bursts) of 600kHz rF which activate pooch's pain collar (did you ever see Twin Peaks?) and which may come through as clicks (or something) on the speaker when the receiver is tuned to the right frequency (as with any received RF signal). This collar needs to be a bit selective in what signals will activate it or a local radio broadcast could send pooch wild!
2. The choke need not be very accurately specified and I guess it is largely there to fool the box into believing there's a complete conductor. Your distance limit near the end may depend on actual routing / lengths / depth buried etc. etc. as well as the choke value. Two wires, fed at the centre, behave like a dipole and they will also operate if you bend them round into a circle - you just get a different impedance. If the added choke acts as a bit of improvement in the matching the broken system then it cannot be highly critical as the break could be anywhere and the routing is unknown.

Also, your loop / wire system is not an 'aerial' as such because it is not required to 'radiate' its power (better not to, in fact). It just couples with a nearby receiver and its near field operation is what counts.

 

[Q_Goest]

What's an MF receiver? I had an AM radio tuned to 600 khz. I haven't tested it yet with the wire intact. I'll check that tonight and report back.

 

[sophiecentaur]

What's an MF receiver? I had an AM radio tuned to 600 khz. I haven't tested it yet with the wire intact. I'll check that tonight and report back.

It's what you have but I am using a better defined term for it. AM refers to the (Amplitude) Modulation system. MF refers to the frequency band, which is more relevant here. AM radios cover a vast range of frequencies - from 60kHz to many GHz.
Your "wireless set" should detect the signal - possibly better, when the unit is working properly. And so should your dog's collar - no more nights out on the town with the local ladies.

 

[meBigGuy]

WOw, how many times do I have to explain the choke. The choke is simple. All it is is a loop simulator. It let's the unit think it is driving a loop so you can track the signals down the broken wire.

Yes, you will receive 600KHz (or a close frequency) when the loop is repaired. That is how the system works.

As for driving the loop at resonance, I have built loop antenna systems that automatically resonate the variable inductance loop. I'm not saying this unit does that, just that it isn't all that hard. Having a decent match is probably not that critical in this application. I would expect a fixed tuning capacitance in the unit such that a typical loop would be near resonance.

 

[Q_Goest]

1. What do you actually hear on your mf receiver? I imagine the box produces pulses (bursts) of 600kHz rF which activate pooch's pain collar (did you ever see Twin Peaks?) and which may come through as clicks (or something) on the speaker when the receiver is tuned to the right frequency (as with any received RF signal).

I haven't seen Twin Peaks, no. I must be missing something again <sigh>.

Anyway, when tuned to 600 kHz, there is a strong, static noise with a bit of a high frequency popping to it.

I tried searching for the wire tonight with the radio and found the same noise. In fact, it seems every radio station up and down the AM dial comes in significantly stronger when the radio is near the wire, but I guess that has to do with the wire acting as a receiving antenna somehow as well. My house has aluminum siding and similarly, signals come in stronger when the radio is held near it as well so that's no oddity.

As a side note, I also tried searching for a signal on FM but didn't come up with anything.

So yea, I guess the choke simply 'fakes out the circuit' into transmitting as it normally would. The fact it's transmitting on an AM wavelength is a bit strange though, local radio stations and all as you say...

Thanks for everyone's inputs. I realize I'm not the easiest student but I'm still learning!

 

[meBigGuy]

I'm still learning too. Learned how to use a choke to fake out a loop transmitter.

FM radio is in the ~100Mhz band.

I suppose either the signal from the loop can overpower or be detected through the local stations if one happens to fall on the same frequency, or the signal is broadband or sophisticated enough to not be affected by an AM station.

I sure would like to know how the radio behaves as you approach the break with one lead disconnected (That last 10ft drop-off). (I guess I'm asking you to undo your fix, though).

 

[AlephZero]

FM radio frequencies would not be a good idea here. The wavelength is only a few meters, so a "random" shape of the loop could lead to problems with the signal strength example if the loop has sharp 90 degree bends at the corners, or doubles back on itself.

On the other hand at 600 kHz the wavelength is about 500 meters, so any reasonable sized loop will act as an untuned wire.

The specified RF choke (100 uH) has an impedance of about 380 ohms at 600 kHz. Presumably that is low enough for the transmitter to think the circuit is unbroken, but high enough so the broken loop in parallel with it still gets most of the power from the transmitter.

As somebody said earlier, this system works using the "near field" radiation from the wire. That is high close to the wire but typically falls off as $1/d^3$ compared with $1/d^2$ for the far field or $1/d$ for a directional antenna. The far field signal will be too low to interfere with a broadcast AM radio on the same frequency.

 

[sophiecentaur]

Perhaps it's because I was in Broadcast Engineering for a long time but I find it strange that many people refer to RF frequencies in terms of the Modulation Systems that are often used at those frequencies. In the 'VHF' bands - which extend from 30HMz to 300MHz (ITU designation) various forms of modulation are used and that use is subject to change. Also, AM is used in some VHF channels. I think it would be a good idea on PF, at least, if we used the more precise terms, in the interest, as usual, of avoiding confusion.

I realize that what's written on peoples' radio receivers may be not what I am recommending but the same goes for a lot of what we find in 'popular' sources of knowledge.

@AlephZero
I am not sure what you are referring to when you talk of a simple inverse law for a "directional antenna". Afaik, there is nothing to interfere with the Inverse Square Law for any EM source at great distance. Could you explain the circumstances for 1/d drop off, please?