Small electricity problem to solve!

[Forhorse]

An efficient electric fence is generally over 3500V, it can go up to 6000V with powerful energizers.
Below it loses its efficiency. Old used horses that will experience lower tension will understand this warning. A young horse who does not know what it is or a panicked horse will make fun of the little chestnuts it will give him. This is why I personally consider that a good fence should not be below this value.

The frequency of the pulses should not normally be lower than 1s, to see higher in this subject, I explained why it seems to me.
As for the intensity, it is difficult to say. depending on the energizer it can exceed 1 amp (hey yes!) because the pulses are very short, if we want a sufficient shock energy the intensity is sometimes more important than we think.
The maths will make the calculation, knowing that a battery energizer generally delivers 0.8 joules per pulse, and that those on sector deliver rather in 4 joules, take the duration of a pulse (a few miliseconds at most) and deduct it the current of this pulse ...

I do not believe in detection by capacitive effect, it seems to me diffiicle by this means to know if the fence at a sufficient voltage or not. We can actually detect whether or not pulses are present, but not the actual voltage of these pulses and therefore if the wire has fallen to the ground or not.
For info, just by capacitive and inductive coupling, a "powerful" fence (4.5j joules) manages to power up another fence parallel to this one located at 50cm over a length of 100m. The induced impulses are more than enough to start a man touching this fence which he then believed to be off (my neighbor came to inspect the fence of his cows )

My mounting based on resistances it works. I know that when the voltage is high enough the LED lights up, and I also know that if the fence is of poor quality and that the voltage is no longer high enough it doesn't light up anymore (I did a lot of damage to be sure)

But if you have a diagram based on the capacitive principle to offer us I am ready to try it too.

[dedeleco]

For men we officially limit to 1mA whatever the low or high voltage especially for the 50Hz, even at 300000V!
For defibrillators we put about 100mA in pulse to resynchronize the heart!
But the same pulse with no luck compared to the heart cycle, blocks the healthy heart and we die !!
It is said that in the 1920s the first who studied this, tried on himself and died!
But horses have more resistive hair than men and the voltage is increased but the current hardly exceeds mA at 6000Volts even for 1joule.
With the capacity of 1/18 microFarad storing 1 Joule at 6000Volts this current depends on the resistance in series. To limit around 1mA (safety standard) a resistance of 6 MegaOhms is required, otherwise the shock is excessive and the time constant is 1/3 of a second.
This explains the time interval between pulses so that they are well separated.
At an ampere one kills the men and the safety standards are not respected !! In the event of a human accident, insurers will raise this problem to insure nothing !!!.
The peak current is fixed by the series resistance and that of the animal or of the man very variable according to its humidity (around several tens of Kohms in general but fear to be weaker).
So this current without series resistance reaches 1 / 10A at most, a real machine of death on man if bad luck!
With a much lower series resistance we go to the defribillator and the risk of human death !!


Finally, the capacitive detection by capacitive divider bridge is as good as a resistive divider bridge especially with pulses without having the problem of finding the high solid value resistors, which dissipate and reduce the voltage.
The capacitive bridge is formed by the capacity of a piece of high voltage wire a few cm and the parasitic capacity of the input transistor (in some picoFarad);
I am sure of that because it is a basic presence detector at short distance, even with ordinary transistor (not CMOS too fragile), because the 50Hz in the house by capacity is enough and it is very visible on an oscilloscope !! Here the high voltage repeated at 1s is enough to replace the 50Hz
The proof is that a close fence by capacitive effect is found with a high voltage !! (not inductive because current too low and frequency too low).
In addition in case of lightning the resistors will be dead but not the distant capacity.

[Forhorse]

The problem is that we are not looking to make a presence detector, but a way to detect the presence of certain high voltage pulses.
A capacitive assembly because of its high input impedance will be able to tell the difference between the 3500V well present and correct power pulses of a fence in good condition and the vague pulses of a few hundred volts, sometimes induced simply by capacitive coupling, of a fence on the ground over several ten meters?
If yes then what would be the diagram of this detector?

[edfed]

don't bother, put:

a voltage divider with HV resistors; 10Mohms / 10kohms => Ue / 1000 approximately, Ue / 999 exactly.

a very high input impedance op amp, Ze >> 10Mohms, configured as a follower, to take the measurement above with a voltmeter, without disturbing the measurement.

possibly a comparator with Vseuil = 5V; 5 * 1000 = 5000V, to light a led above 5000V

and you will have your measure. it is the simplest and the best. believe in my experience my child.

[Forhorse]

don't bother, put:

a voltage divider with HV resistors; 10Mohms / 10kohms => Ue / 1000 approximately, Ue / 999 exactly.

a very high input impedance op amp, Ze >> 10Mohms, configured as a follower, to take the measurement above with a voltmeter, without disturbing the measurement.

possibly a comparator with Vseuil = 5V; 5 * 1000 = 5000V, to light a led above 5000V

and you will have your measure. it is the simplest and the best. believe in my experience my child.

This is more or less what I intend to do. The diagram of my thing I had already posted paged 4
https://www.econologie.info/share/partag ... SpodZV.gif
it works well as well.
The only problem is that the tension that each resistance can withstand is too great. Instead of putting only 2 resistors in series at the head, I must put the number which allows not to exceed the maximum voltage of the resistors.
There it works well for about 3000V, but if it exceeds it starts to initiate resistance inside. But the component values ​​are good.
Then, to have a galvanic isolation between the fence and the rest of the assembly the led will be replaced by an optocoupler, and this will be used to drive a beast NE555 which will be responsible for checking the regularity of the pulses.

[dedeleco]

Between 3500V and 350V a resistive or capacitive divider bridge divides the pulsed voltage at the frequency of 1s like this !!!
The 50Hz AC for the capacitors and their impedance is replaced by the 1Hz of the HV pulses and its harmonics at 2,3, 4 Hz, to give the final voltage.
The parasitic wire-to-wire capacity over a few cm is low on the order of 1/10 pF and therefore the capacity to input the detector transistor is calculated by multiplying it by the division ratio at least 350 or at least 100pF at input. put, but we can put more 1000pF (protects the transistor more better than a high resistance) and put a length of HT wire opposite longer (it's free !!).
Do you have an oscilloscope, because then you can directly see the signal of the fence according to the length of wires, distances, plates, and capacity and resistance put in parallel on the input.
Otherwise, you can mount a transistor in detector mounting with diode and capacity for 1Hz to measure your signal with voltmeter.
It is good to put two 15V zener diodes for head to tail protection to protect against overly close HV and destructive overvoltage by bad luck (the varistors have too many capacities unless they are used as capacity).
Even with resistors, the capacitive effects exist in addition to the high resistances and can be very strong without suspecting it !!!

Finally, considering that tomorrow afternoon I am 900km and that I am preparing, I cannot give a precise diagram before Sunday.
But on the net with google there are various presence detector diagrams for alarm or other (for example on the National Semiconducteur website, instructive examples of use with various CMOS 4007 or LM324 operational amplifiers, often too complex).
the final diagram in peak detector limited to the first stage with diode
http://www.national.com/ds/LM/LM124.pdf

http://ics.nxp.com/products/hef/datasheet/hef4007ub.pdf

[dedeleco]

There it works well for about 3000V, but if it exceeds it starts to initiate resistance inside. But the component values ​​are good.

Normal and fun, why keep it simple when you can make a mess !!
In addition the inevitable capacitive effects distort you all you realize it with precise resistances !!
But in France we sometimes teach poorly the parasitic capacitive effects to the point of sticking BTS electronic teachers with the question on the signal seen on a wire in the air at the entrance of an oscilloscope with the question: where does that unwanted oscillating signal?
It is the proposed capacitive effect and more resistance breakdown !!!

[Forhorse]

The fence testers at 15 euros work exactly on the same principle as my diagram, perfectly and this for years. And again, the most basic have only resistances.
Except that there are much more than 2 resistances precisely because of this problem.
I was inspired by it because I am not the type to reinvent something that already exists. The only adaptation is the addition of an LED, to simulate the future optocoupler of my assembly.
If I am offered another scheme I would try it but any future verbiage on this subject without me having anything concrete to try would be a waste of time.
If you have nothing but blabla and big principles to state then I would stick to my idea.
Cordially...

[dedeleco]

It's not a blabla but the reality, you pay € 15 with resistance HT for what is worth € 0,5 with a diode and a transistor and a capacitor mounted according to the same principle which explains that another fence next gives discharges to walkers who do not need a € 15 tester to test at 50cm !!.

It makes me laugh !!

[Grelinette]

Houla I was late in reading the answers.

This little problem of electricity arouses interest ... but for the moment we do not yet have a system developed. Like what is not so simple!

I hope everyone's advice will allow Forhorse to make effective material. (Forhorse seems to me the most competent to plunge his hands into the sludge and achieve an operational system! )

Here is a summary and description table of several market energizers:


I add a remark concerning the dangers of too strong a current: it is useless to dwell on this detail because the electrifiers of the market respect official standards.

The problem is simply to realize a control system which can be directly connected to any fence installed in the state of the art and which warns if the fence is broken or open.

For the rest (branches, lightning, grass, butterfly, mushroom picker, ... it's detail!

I add one more precision (which comes to me as and when reading the last comments): there is already a very common fence tester model which lights up diodes according to the power of the current which passes in the fence:

Perhaps this reliable, efficient, easy to plug in (and tinker) and inexpensive hardware can serve as the basis for the system? ...