I believe a spectrum analyzer test should be made to figure out the short and nominal working frequencies.
Need some figures to work with.

Main problem its big and heavy, I need a hand to move it outside, (when it isnt raining of course!), then I can set up near the fence, or does anyone know how to pick up the frequency from say 15 meters?
I could get a wire strand from the fence to within 15 meters of the lab, would a wire in the LF socket of analyzer pick up the frequency from that distance? I could probably pick out the clicks, and I am assuming that the interference while a different frequency, will peak with the clicks?
Your right I need to experiment. The frequency analyzer is one bit of test kit I havnt used much. I should try and get the manual for it if I can

 

I am about to do a google for manuals, but the equipment I have to hand is the following,

HP8568B frequency analyzer 100Hz- 1.5GHz I have to be careful with this, it took me nearly 2 years to save up for!

KIKUSUI KSG 4100 FM/AM frequency generator 0.1-110 MGHz, this is ace and pretty spot on frequency wise, my general day to day RF generator

HP E4400B 250KHz - 1.0GHz frequency generator, this machine is ace, but its dad so extreme care needed lol,

the other 2 signal generators are my every day bash ones, they are not upto this kind of thing, I keep them for abuse type jobs

 

Hello,

The intermittent shorts on the fence wire will act as a spark gap transmitter.
The wiki has this article on the spark gap transmitter:
http://en.wikipedia.org/wiki/Spark-gap_transmitter

The radiated RF signal will probably scattered over a large frequency range, as there is no real tuned circuit.

Bertus

 

very interesting! So I have a 10KV spark transmitter! thats a start, so in theory there should be a frequency that shows a peak every second (roughly), being 10KV then it should at close range dominate for a brief period each cycle, so find the frequency that has the least amount of background noise, is pretty much clear of other signals, build a bandpass filter to isolate that frequency, build a amp to up the signal. Then add on a meter to show strength, and Robert is my mums brother!
Sounds very easy! LOL.
First job on first rain free day is get a strand of fence near to outside of lab, make sure there is a good ground connection near by for fence, use spectrum analyzer to look for 1 second peaks, isolate the biggest peak at the clearest frequency (nothing much else using that frequency), Then I can start designing amps, detectors and filters blah blah blah, sound good??? Or am I imitating a dog like noise at wood?
I will probably go with a active filter just to be difficult.
Once I have isolated my frequency I can do everything else in the lab and just use a signal gen to replicate frequency I am after???

I have another barking mad project on the go

 

Since the pulse is so short you may have difficulty viewing the frequencies with a spectrum analyzer since they normally sweep the band and thus require a steady signal to view all the frequencies. But perhaps, looking at the signal for an extended time, you will be able to see all the frequencies, each sweep showing a small portion of the total. A FFT analyzer would likely work better since they look at a snapshot of the signal.

Any frequency peak in the signal (if seen) would likely be due to the resonant frequency of the fence line due to its stray distributed inductance and capacitance.

 

I will start at a fast sweep over a narrow band, single sweep it, move band and do again etc, I have FFT on one of my scopes, so I might give that a try.
There is alot of energy in the fence, I am pretty sure it would give a signal, the spectrum analyzer is pretty sensitive, maybe I could rig an external trigger and sync the analyzer and energizer somehow.
On a radio you can pick up the clicks of the fence from a fair distance, even if I dont get it working, its a great project to try!