Yeah, yeah ..I know what you're thinking....and no, I'm not suicidal!
I am trying to use an Arduino to measure the voltage of an agricultural electric fence. The idea is to have the device at the end of the fence to make sure the fence isn't shorting to ground somewhere along its length and notify the farmer if it is. To be safe, I should allow up to 10,000 volts to pass through the device. Although the voltage is high, the pulse width (by law) is less than half a second so as not to kill humans if they touch it.
The goal is to measure the peak of this very short pulse. I cannot use anything like a shunt, as the device, instead of detecting shorts, will be the short.
What hardware is required to achieve this without zapping my Arduino or taking much 'oomph' out of the fence?
Wow, that is indeed a challenging task to deal with. Finding various coupling components (capacitors, transformers, even resistors) rated for that kind of voltage is difficult and expensive. If the voltage pulses are long enough to fire a small neon lamp with a suitable string of series resistors, maybe a lamp/phototransistor might be a safe interface to an Arduino. I can't see a simple way for the Arduino to actual measure the voltage peaks, but detecting that voltage pulses are present via the neon/detector might accomplish the goal.
It will be interesting to see the responses to this posting.
Zoomkat: you posted the same idea as I was typing, must be the best way to go. ;D
You probably don't have to actually touch the fence wire, just get close enough such that ~100v potental is generated by the pulse to flash the bulb. If you get in the dark on a dry winter night with with a neon bulb, just the static on a wool sweater or the house cat will make it flash. A search of farming supplys may actually find neon bulb setups for indicating that an electric fence is operating.
Thanks guys, but while the neon bulb will indicate that there is a voltage on the fence it may only be 4000 volts (half the desired voltage) because the fence is arcing to a tree, for instance. The 'shorts' are often arcing to a nearby earth and therefore not like a complete pull down to earth.
Those farmers like to 'light up' the potential escapees with the full 8k to get the message across!
Thanks guys, but while the neon bulb will indicate that there is a voltage on the fence it may only be 4000 volts (half the desired voltage) because the fence is arcing to a tree, for instance.
Ever think about increasing the distance between the bulb and the fence until the bulb stops flashing, then moving it back a little closer?
I thought it sounded like an australian unit. I've worked with some cattle ranchers that have gotten them, but here in the US if somebody's kid happens to get on the fence, they might lose their farm in a law suit.
You could have a lot of 'fun' with this. My first thought was using a static electricity meter- I recall a physics professor having a very simple device that was two metal rod joined at one end with a pin to allow the rods to swing. The rods were thin and light weight, and once electric charge flows onto the rods, they will want to repel each other and will swing some distance apart. You could put a piezo strain sensor to measure the force of the rods repelling as a function of voltage. You could also experiment with taking two metal plates to make your own capacitor, and install a piezo strain sensor to measure the force between the two plates. Put some thin plastic, or rubber, or wax paper between the plates (I suggest Aluminum).
Another type of idea. I believe that this voltage sensor would not need to refresh often....maybe somewhere between 4 times to 24 times a day is good. If this is the case, then you could setup a bank of capacitors and diodes to handle the voltage to charge at the high voltage, and then discharge at a much lower voltage. Look into 'voltage multipliers'.
You could maybe most easily use a shunt that is switchable from programming in Arduino- say activate the shunt once an hour and then read the voltage on the line using common meters. Use long grounding rods into the soil for both the electric fence ground and another for the shunt ground. The voltage will fluctuate some with weather and moisture in the soil.
another one: radio tuned to a blank AM station. The RF noise generated by the pulsing high voltage should be detectable; this is essentially how a lot of tachometers get their signal for measuring engine rpm. I know you want the actual measurement of the voltage, rather than simply know if its there. You can spend some time and creativity and find a way to calibrate any sensor you make.
Neons have a trigger voltage of around 90 volts or so and were once used as a form of constant voltage source, rather like zener diodes are currently used, so they'd be no good at providing an analogue value.
Assuming you want a drop from say 20,000 volts to 5 volts (to allow for the pulsar giving more than 8Kv) you would need a division of 4,000:1 (OK not exactly but near enough) with say 1000 ohms as the input or sense resistance, you'd therefore need 4000 x 1000 or 4 megohms as the main dropper resistor. To ensure against track-over this could be made from 4 off 1meg resistors each rated at 10kv wired in series.
In effect you'd end up with a chain of 5 resistors, the lowest of which provides the ground and +5 volt sense to be input into the arduino.
Also place a diode across the 1k ohm detection resistor reverse biased to ensure you only detected a positive going pulse.
Cheap and cheerful with a known calibration factor.
