Invisible Dog Fence Repair

Thread Starter

wayneh

Joined Sep 9, 2010
17,495
An invisible pet fence is a loop of wire barely underground, reportedly broadcasting at 600kHz, although other reports say 10.65kHz. There's a popular method for finding a break in the wire loop that involves placing this 100µH RF choke onto the transmitter, in parallel with the broken loop. Some write-ups suggest you additionally wrap the ends of the broken loop wire around the leads of the choke before reconnecting to the transmitter, instead of just putting them in parallel. You then use a portable AM radio tuned to ~600kHz to scan for the lack of a throbbing signal, indicating the location of the break. Many folks have had luck with this, so I'm sure it works at least for some systems.

I'd like to understand why this works, and whether we can improve on it.

I understand that one role of the choke is simply to provide continuity. That allows the transmitter to stop beeping and produce the signal. If that was all it was doing though, why not use a resistor? Because the choke won't allow AC to pass and therefore keeps the RF signal stronger into the boundary wire loop?

Another suggestion is that wrapping the boundary wire leads around the choke leads provides a crude isolation transformer. Pulses through the leads of the choke are picked up by the wrapped wire. Perhaps this increases the signal you would get without wrapping?

Any suggestions or insights?

EDIT Is there a way to identify a similar choke in my parts heap? I've probably got a bunch of them, but I've never used one and don't know what I'm looking at.

EDIT2 Well, I lucked out: An old solar light had a 100µH inductor on it (brown, black, brown, silver, measured resistance at ~3Ω, so it's not a resistor). It's smaller than the "choke" available at RS but has the same impedance rating. Any reason to think it won't work? It's roughly 1/4" long and 3/16" wide and green, fwiw.
 
Last edited:

#12

Joined Nov 30, 2010
18,224
I just keep getting this crazy idea in my head, "Why don't you use and invisible volt meter?"

Sorry. It's such a good straight line.
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,495
Yes, "hidden" or "buried" would be more accurate, but we can't let that get in the way of good marketing. ;)

Getting back to my doggie dilemma, can ANY inductor be pressed into duty as a choke? As I understand it, "choke" defines the application, not the device itself.

I guess I'm wondering if the frequency of the application would make you choose one inductor over another despite identical inductance ratings. I think the core material matters. The RS device is called an "RF choke" and must have the right core for that application. The little inductor I pulled from an old solar light may not?
 

#12

Joined Nov 30, 2010
18,224
The core is important. Iron just doesn't work right at rf. However, your green coil is probably Ok because its very common to use ferrite in that size. Do a few finger tests on it when you start to be sure it isn't going to lose its smoke. If it stays cool, procede with test.

That's my opinion.
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,495
I think it works (no smoke!) but I've learned that this is a PetSafe system, which doesn't use 600 kHz. So my AM radio could detect a pulsing, but only when held directly against the inductor. Holding it against the perimeter wire gave me nothing. I tried the entire AM band, and was hopeful when I got a little something at 600 kHz, but that was all.

The PetSafe solution is a $50 transmitter that you use as a stand-in for the regular dog zapper. It put a different tone on the two wires, so it's easy to tell where one ends and the other begins.
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,495
Still looking for a clever (meaning cheap ;)) solution to detect a wire break in a dog fence. The PetSafe system broadcasts at 10.65kHz thru the perimeter wire as an antenna. So I either need a receiver for that frequency OR another transmitter that would allow me to use a receiver I may already have (AM or FM radio or walkie talkie). I think the commercial device ($50) is an AM transmitter, although it appears to come with its own receiver, so it's hard to be sure exactly what frequency it uses.

Putting my inductor in parallel with the broken wire loop did allow the transmitter to function, and it's probably pulsing the perimeter wire, but the dog collar only fired when near the transmitter. It did not fire when held to the perimeter wire, so it's not a useful tool for finding a break in the wire. Besides that, it's more than a little unpleasant to endure the zapping.

The only transmitters I have that came to mind are my 49MHz walkie talkies, and a kid's "mr. microphone" FM transmitter. Maybe replacing the short wire antenna on that toy with the perimeter wire would give me a signal I could detect with an FM radio?
 

#12

Joined Nov 30, 2010
18,224
There are transmitter/receivers that electricians use to play, "find the breaker". I don't know if they can be reversed to do, "find the break in the wire". It would be so helpful in a house with aluminum wiring!

Anyway, you can check into the Electricians stores and see if they have the quality you need or a better price. Please report if you find something that works.
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,495
I'm leaning towards building my own AM transmitter, as described here and here. (The second link has better pictures but muddies the purpose of the transformer.)
Elec_img039.jpg

Note how the high impedance side of the transformer limits power to the oscillator, allowing the audio signal to amplitude-modulate the carrier. Can I replace the audio transformer with a ~1k resistor in series with the oscillator and a capacitor to couple and isolate the audio signal? I'm sure I might lose fidelity but I just want to broadcast a continuous tone, say 1 kHz. I can generate that with an app on my iPod. What size cap should I use to be sure to not overload the iPod output, which I think is 32Ω impedance?

See if you agree with my strategy, of choosing to build a transmitter instead of a receiver:
1) This is simpler than building a receiver for 10.65 kHz, the native PetSafe dog fence frequency. (Which may vary between models)
2) I can test my transmitter with a standard AM radio at 1000 kHz, and use the same radio to find a break in the wire. If I build a receiver for 10.65 kHz, I cannot test it except with the original transmitter, which I cannot assume is working correctly.
3) This solution will work for every buried-wire system since it uses nothing of the original system, except the perimeter wire.
4) The transmitter strategy is used in the $50 commercial device.
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,495
For anyone that may follow this thread, I've posted a completed project that should show up soon in that forum. The circuit I ended up with is below. In short, it worked fine and my friend's dog can now run around outside with no worries.
Picture 2.png

The build
IMG_2550 small.jpg
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,495
Hah! Yes the adjustment of the Petsafe thing is a bit touchy, despite having a dial going from 1 to 10 (or is it 11?). The manual is also quiet about the fact that the only way to test the range is by manually trying yourself. Not pleasant. At low settings, say 5 or below, there's no response when the collar passes over the fence. At 7 and higher, the poor dog gets zapped almost everywhere in the yard. Tweaked to 6.25, he gets it when he gets within ~4 feet or so of the wire. Not bad.
 

someonesdad

Joined Jul 7, 2009
1,583
If it's for an invisible dog, how does your friend tell if the dog is outside?

Another tool to consider is something like Gardner Bender's LAN Tracker (GET-4220K). I have an earlier version and it's pretty good at tracking wires as long as they're not buried too deeply. It uses a 2 kHz FM'd signal at about 35 Vpp.
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,495
Hmmm.... I see no new Completed Projects for the last month. That's kind of anemic. Mine's been in the queue for a few days, and I guess I can expect it to linger there a while longer. I suppose the mods have better things to do in warm summer months. :cool: Hope so! I'm not pushing, just noticing.

Anyway, here's a question: How does the load of the wire affect the oscillator? The oscillator is said to be able to drive 10 TTL loads. Is that a lot?

A dog fence perimeter wire is up to maybe a 1000 ft long, insulated, and underground a couple inches. I'm guessing this would not risk overloading the oscillator but it's just a guess. Wookie posted a calculator for inductance of a wire recently, and I get about 25µH for my 60 meter wire, guessing it's 1 mm in diameter. The calculator doesn't adjust for the wire being buried. Grover's seminal reference probably addresses this but I haven't found it online. Maybe the inductance is less importance than the capacitance?
 
Last edited:

brozizds

Joined Aug 15, 2010
135
I took 110 ac (fused) and attached it to the wires then use a voltage tic to locate the break and repaired it. My wire is just under my grass not deep. Wire was broke but insulation wasn't
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,495
OK, I Googled it and learned about voltage tick testers. I could see that in some cases they might work for finding the wire. But can they really sense a wire that might be 4 inches underground? There wouldn't be any current on the wire, since it's broken.
 

someonesdad

Joined Jul 7, 2009
1,583
OK, I Googled it and learned about voltage tick testers. I could see that in some cases they might work for finding the wire. But can they really sense a wire that might be 4 inches underground? There wouldn't be any current on the wire, since it's broken.
Must be some foreign term, as I've never heard of that either -- they've always been called non-contact voltage testers when I've seen them. You have to play with one; I used mine last weekend to trace a known-hot wire behind some drywall, but I didn't trust the results because it wasn't sharply telling me where the wire was. Greenlee makes a nice one that I've seen on the web that can adjust from line voltages down to around 5 VAC, so I'd like to try one of those one of these days. They work by detecting the electrostatic field, so current in the wire is irrelevant.
 
I have about 2000 feet of buried fence which occasionally gets a break in it.

Here's my cheap solution. As a broad-spectrum frequency generator which you can pick up with any portable AM receiver as static, simply disconnect your transmitter completely, take a thin wire (awg 18 or so) and wrap it around the spark plug wire of your lawnmower (DO NOT CONNECT IT TO THE SPARK PLUG!). Then connect it to only one end of your fence wire.

Start the lawnmower. Tune the AM radio to around 600khz off any station and hold it near the ground by the start of your fence. You should hear static and a buzz at the spark frequency of the engine. Then just walk along following the radio static until it quits and voila, you found the break. To be sure, stop the mower and connect the thin wire to the other end of the fence wire, restart the engine, go back to where the static quit, and see if now it doesn't continue from where you left off.

In my case with 2000 feet of wire, I put a "manhole" (6" diameter by 8" long plastic pipe) about every 200 feet, so in my case I can move my lawnmower noise generator closer to the break if I need more signal strength (the signal tends to fade with distance and ground dampness).

Hope this helps (and it's free!).
 
Top