Small electricity problem to solve!

[Grelinette]

Well ... I just finished reading the last comments and I see that the debate is starting to turn into pudding water

To refocus the debate, I give the application context and I pose the problem:

Mr. G. (better known by the nickname Grelinette) has equines parked a few times far from his house and it has happened that said equines are found stray on the public highway (including once in the parking lot of a supermarket!).

After careful investigation, here are the causes of these "escapes":

- fence broken by horses (or boars) frightened by lightning, a hunter's shot or stray dogs that attack them
- broken branch which fixes the fence wire to the ground (or even simply the accumulation of snow on the wires)
- walkers who opened the fence to see the horses
- ... and sometimes more worrying causes: http://www.hippoplus.com/hipposervices/protection/grev/chevaux_voles.htm )

In short, to overcome all these problems, it is a question of realizing a simple, effective, inexpensive system, easy to install and accessible to all (i.e. owners of equidae who have no knowledge or skill in electronics or DIY) and which allows:
1 °) to detect a closure problem
2 °) to warn someone from a distance

that's all !

[Forhorse]

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

It is exactly the same kind of tester that I bought in order to read the values ​​of the resistors used.
By cons cheap I would not say that. I paid mine 15 euros when there are barely 5 euros for components.


When we open it, we find this: (it couldn't be simpler: 8 resistors and 6 neon lights)


I noted the diagram of this kind of tester that here:


The problem with this kind of tester is that the very short pulses of the energizers hardly make small flashes in the lamps, which makes reading very difficult.
In the battery-powered energizer that I use they alleviated this problem by adding additional components.
One day I took advantage of dismantling it to take up the diagram of this light here. It should be noted that that is only supplied at 1000V because an intermediate socket on the high voltage transformer has been specially fitted for this indicator.


Starting from all this I therefore deduced this diagram (which I repost in this subject for the third time)

Which works perfectly apart from a small problem of too high tension at the level of the falling resistances (the 4 of 2.7kohms) which I think to replace by 8 of 1.2 or 1.5kohms

Note that personally I rather use a "digital" closing tester which indicates the voltage with a resolution of 100V but which is, when you buy it, a little more expensive (I had benefited from a promotion at the 'when I bought a mains energizer, the digital tester was offered)

This tester, as advanced as it is, also uses the principle of the resistive divider bridge to make its measurement, at least for the first stage. I have not studied it too much.


Some manufacturers, however, use the capacitive effect in some of their testers. it allows for example to have testers that work without earth (I had one in my youth, that I had paid a small fortune)
See “Digivolt” or “Control beep” from Lacmé for example.
http://www.lacme.com/Scripts/WebObjects ... ?idDep=416

Your table is interesting because it shows that, as I explained above, the current delivered in an electric fence perhaps very important is to exceed the ampere.
the data is confirmed in any serious manual of electric energizer
http://www.lacme.com/rd/613600_NU01.pdf
We note that an energizer of 0.8J is capable of delivering a voltage of 3800V on a load of 500ohms.
if we apply the law of ohm that gives? I = 3800/500 = 7.6A!
It is this characteristic which allows the "modern" energizer known as low impedance, to be able to continue to deliver a significant tension is thus efficient even if there is a lot of fault on the fence (grass which touch the wire) there or the energizers of (very) old generation, called later "high impedance" saw their output voltage collapse with the first blade of grass which came to touch the fence.
The "low impedances" on the other hand have much shorter pulses, where the "high impedances" took their time.

Yes, because what is dangerous is not only the current, but also the duration of the passage of the current. Besides when we pass an electrical habitation the gentleman trainer shows us nice curves where the danger threshold is given not only by the value of the current, but also by the duration of passage of said current.
see page 5 (and again the curve starts at 10ms, which is already very long)
http://lp-ampere-72.ac-nantes.fr/IMG/pd ... humain.pdf
With an electric fence the pulses are very very short, which means that even a large current is normally not dangerous.
In addition someone touching the fence will not be crossed by such a current because the human body has a resistance much greater than 500 ohms (especially with shoes)

To finish this little explanation we can do another calculation.
If we take the example above have at 3800V under a current of 7.6A
Knowing that the shock energy is only 0.8J anyway
according to wikipedia E (joules) = R (ohms) xI² (amp) xT (seconds)
we set 0.8 = 500x7.6²xT
T = 0.000028 seconds or 28 micro seconds
(we can also do this with E = U² / R x S because RxI² = U² / R)
This also explains why it is so hard to see the loupes of the basic fence testers, because in fact we have to "guess" them more by retinal percistance effect than to really see them.

Note that Gallagher (who is my favorite manufacturer of fencing equipment) seems to have put on the market exactly the device that we are trying to develop.
http://www.gallaghereurope.com/france/g ... tnr=055654
only its price is a little dissuasive.

[edfed]

wanting a simple, inexpensive, easy to implement, efficient solution requires a minimum of electronic knowledge.
otherwise, you must agree to pay the full price. this is the reason why you are never better served than by yourself, provided you take the trouble.

here is a schematic diagram, without component values.

we can clearly see the resistance chain to support high voltage peaks. at the price (about 15 cents per 10) of normal resistors (250v max) we will not be shy to put 40 * 250 = 10000 Volts.

on the diagram there are only 20, but it will be necessary 40, arranged in series the inside of one (or more epp thick of) insulating sheath, held as straight as possible, and whose entry will be as far as possible from the assembly.

the zener diode is used to block the negative pulses and limit the positive pulses, in order to preserve the entry of the op amp, but is not necessary, the last PDO are designed to support negative voltages compared to the negative of the input.

the measurement output is taken directly on the op amp.
on the other hand, the comparator will require a time base, to integrate the pulses, and allow relevant detection.
the comparator threshold is adjusted by the divider bridge on its + input.
the output will be high when the pulses are absent.

Normally, the input of the assembly should be galvanically isolated from the HV source.
a clipper on the input of the op amp should be enough, a 12Volts zener, for positive pulses, it will write at 12V, and negatives will be clipped at -0.6V.

the power supply of the assembly will be done directly on the battery.

the output led will be on when the HT pulses are absent for a certain time, determined by the time base of the small RC integrator at the input of the comparator.

normally, the realization of this assembly should not pose any particular problem, except the chain of resistances which must imperatively be held with the screen of the assembly, and sheathed to the maximum.

here.

[Forhorse]

The solution of the op-amp is elegant and very precise. It has the advantage of drawing very little current from the fence, the calculation of the divider bridge is therefore simpler. We can have a very high impedance on the HV side, it is the gain of the amp that does the rest.
The disadvantage is that there is no galvanic isolation between the high voltage and the rest of the assembly (and therefore its power supply)
But it tells me to try anyway after all. I would try it.

The problem with my mounting based on an optocoupler is that you have to draw a certain current from the fence (between 20 and 60mA) and that this current strongly influences the voltage present across the resistors. If we make calculations for a certain closing tension, they are no longer valid for a higher tension.
Here it is the neon indicator that acts as a surge arrester.
The advantage is that there is galvanic isolation, never the high voltage mixes with the rest of the assembly.
Behind it attacks a classic NE555, simple, current, inexpensive and reliable.

[dedeleco]

Having extensive practical experience in electronics in dangerous high voltage power supplies for fast laser or flash, radio frequencies with very weak signals, I can help you and solve the practical problem but you should summarize your exact objective clearly, with your choices, and possibilities, GSM, landline phone or walky talky (no subscription and less than a few km) because the problem is not to measure or detect high voltage (immediate at very low current with transistor without operational amplifier, like a good old system 0,1mA) but make the right choices to make it simple and cheap but functional.

Finally, the safety of electric fences is not to be looked at with your eyes closed as with your answer "the energizers of the market respect official standards", because those that I see by chance in the countryside, already, do not have enough signs informative and the day a cardiac walker pissing on it dies, it is the disaster with the insurances which try everything not to pay and make you responsible and you will have to pay millions in damages out of pocket for the rest of your life !!
I can guarantee you that insurance with an armada of lawyers always succeeds in running away and not insuring !!
It's not a blah !!
I would not trust "the energizers on the market" and I would look at how the HV power supply works (capacity, series resistance, frequency) and what the exact regulations are to verify that they are followed to the letter:
Indeed I am very skeptical when reading 15000V and 10Km considering the electrical losses with the HT and the regulation which limits to 1mA the current and the value of capacity of a wire on 10 even 100Km, and load on 50Ohms giving thousands of volts , I doubt the manufacturer and the respect for safety.
We can not draw tens of mA from the HT fence because otherwise death machine !!

It is worse, with drugs, you must not trust because by concealing the danger and selling a dangerous drug the companies earn a lot of money before it is banned !!
Read the book: Global conspiracy against the health of Claire Séverac www.editions-alphee.com, around pages 101 !!
Example Vioxx, antidepressants, neurotin, etc.
This is useful blah!

[edfed]

in terms of galvanically isolating the assembly, we can very well use several techniques, except that the HT does not come from the EDF network, it is not very serious.

1/
pass a super well insulated wire, attached and parallel to the fence wire, use it as an antenna to enter the op amp, without the recess resistance network. Galavanic isolation ensured. but precision uncertain.

ou

2/
supply the assembly via a DC / DC converter taken from the battery.
galvanic isolation. and maximum precision, inconvenient, the price of the DC / DC converter (around thirty € at Radiospares), but the latter can always supply other detectors along the fence, or other fences.

ou

3/
power the assembly with its own very small small battery + solar panel, and use the output to drive an opto, a relay or other isolator.
galvanic isolation and maximum precision
in addition, total independence of the detector from the fence, significant autonomy, if the battery falls in the harbor, it will be able to control the sending of an alert without requiring a complex and greedy logic at rest.
a forced relay to the passing state as long as it works, it is not very economical energy level.

ou

4/
we don't care about the galvanic isolation for this application.

I have a preference for solution 3, and a little 4

[dedeleco]

grelinette specifies

to warn someone from a distance

How far? fixes the technological choice and price
100 to 300 meters siren or by fixed telephone type wire, by basic amplified radio control, or cheap remote control, without subscription.
http://www.conrad.fr/systeme_sans_fil_4 ... 261_223220
5 to 10 km by simple radio type walky talky or CB (not expensive because without subscription) but not secret.
More than 10km, GSM with trading card, but expensive subscription to use with expensive commercial traps and sometimes compatibility problem.
Finally against theft, video cameras (fake well visible, but dissuasive) and real little visible can record all moving images, especially those of thieves (for the price of a two year GSM subscription) and can be sent by radio and by GSM as by ADSL.

The problem is not to detect high voltage (which excites in this forum with electronics which forget that these are very short pulses (microseconds!) and its power supply which consumes too much because it is not optimized, whereas we can reduce the power consumption to a few microwatts, as for much more complex electric watches, which work for years on a tiny battery, with the right circuits !!),
but make the right choice and make the interface that sends the signal with as much existing and inexpensive equipment as possible.

The ideal is an alarm which sends the image or the video of the cause of the incident, triggered by the drop in the closing voltage, even a few seconds, (someone or thief unplugging and reconnecting with gloves )!
By GSM or ADSL via mobile phone, sending an image does not cost more than an alarm signal, as for ADSL, where sending an email with "the closure has been forced" is the same as send the photo of the broken fence !!!!
GSM cameras and cards have all the complex circuits ready-made and just use it.
Random example (prices have dropped significantly) for house:
http://www.spiri-tech.com/fr/monitoring ... Camera.php
and more expensive but with crucial practical advice:
http://www.comodalarm.com/camera-videos ... obile.html
You can have the photo of the enclosure and count your horses for 199 € (+ GSM subscription) at the same time as the "fence problem" alarm !!!
The camera can film with a few photos the horse park and the indicator light of the energizer which if it changes triggers the sending of SMS and MMS (no electronic assembly to wire !!!) or for more security without false alarm, be triggered only through the "WINDOW DOOR DETECTOR" of the camera by a low consumption assembly measuring the presence or the value of the HV pulses of the fence.
By mounting a bistable relay which consumes current only during its change of state, the current consumption is reduced to almost nothing (at the moment of the alarm signal only).

Finally for less than the price of Gallagher
http://www.gallaghereurope.com/france/g ... tnr=055654
you can have the photos of your horses and count them.
(Gallagher does not speak of practical difficulties, sim card, subscription, etc.)
Finally lightning is a real problem, with millions of volts on the fence (microseconds) even when falling far, which if poorly studied, destroys your system and forces you to buy everything and redo, GSM etc ... !!!
With lightning I have a personal experience on my house (fire) !!
Gallagher has a built-in arrester and it is not very easy with long fences !!!
I would be very surprised that it protects every time from thousands of lightning strikes !!!!

[Forhorse]

I have a preference for solution 3, and a little 4

I also think it will be 3 or 4 (rather 4 even)
but I had thought of a fifth method, I don't know it would work but in principle I think so:
use of a small transformer.
Kind 220V-6V / 1VA
We manage so that the primary is powered by pulses between 100 and 200V (with the same principle of the resistance network of other assemblies) and we cut LV pulses at the secondary which are used to attack the assembly based on op-amp .
It is probably useless but it would tempt me to try.
I had even imagined that the energy of the pulses thus transformed are used to charge capacitors which would be used to feed the electronics of the assembly, but it is undoubtedly too much to ask some (whatever that must be possible, Lacmé makes a system which is self-powered by the fence)
This kind of small transformer is easily recovered in many small devices. I often find them in emergency stop safety modules that I change at work.

[edfed]

this 5th solution could not be too suitable because it is important not to influence the tension of the fence.

it should be possible however.

but it will require a transformer sized to receive HT input.

for simplicity:
the primary must have Ue * K turns
the secondary must have Us * Kspires

I let you imagine the misery it will be to wind 5000 * 10 turns in the primary, and recover 5 * 10 turns, respecting the isolation between primary and secondary.

but it should be possible. but super hassle, then you need a ferrite that transfers energy well etc ...
it seems to me more than difficult to do, we do it in professional electronics for SNCF or EDF, but it is for devices sold at least € 1500 ... and installed on lines with high power available.

solution 3 is my favorite because it also allows to power the radio transmitter, and will control it with TOR from the comparator, via optocoupler for example. a homemade opto coupler to have very high galavanic insulation. two LEDs face to face in a 10cm metal tube, a comparator on the output LED, and you will measure a slight voltage depending on the lighting of the input LED.

moreover, to protect the assembly from the direct passage of the HT, you can put a fuse before the resistor chain.
a fuse corresponding to the current (1mA?) which could pass in the event of a short circuit in the resistors.



already, fair the detector, then the transmitter, it will be another matter which will involve a ready-made product.

I have a clear preference for a radio transmitter like walkie talkie, or HF transmitter (Aurel module or other) and then make a small mounting to microcontroller which will monitor the voltage of the fence via the measurement on the op amp. and send the measurement in binary via the HF module.
then the receiver will show you the different parameters on an lcd screen or a 7segments display.
for the GSM solution, I don't know, I don't like on a paid network principle.
after, you can install a personal GSM antenna to create your own network, it is possible, a hacker conference explained how to do in Russia. but here, it is not for beginners or experienced DIYers, it's expert level.

in terms of monitoring the fence with a camera, why not build a wall 3 meters high instead of the fence while we are there, with barbed wire, 40kV electrification, and also a small armed militia which will rounds. the best of all worlds ...

[Grelinette]

Note that Gallagher (who is my favorite manufacturer of fencing equipment) seems to have put on the market exactly the device that we are trying to develop.
http://www.gallaghereurope.com/france/g ... tnr=055654
only its price is a little dissuasive.

I'm sure they came to poke the idea here!

Regarding the surveillance camera, it's interesting (and reassuring) but not achievable everywhere.
A wilderness park can cover several hectares scattered with rugged terrain and a camera will cover only a small area. By cons it is an interesting solution for stables or parks near the stables.