The following is the schematic of a security system designed to work on the break-wire principle and intended to be used on a perimeter fence. The input 2 of logic gate U-3.1 is kept low via a long thin (60 ft) bifilar wire. Upon breakage of the wire, R4 makes the aforementioned input 1 high, which turns Q2 on. Although I have dedicated filter cap C2 (100nF ceramic) and ferrite bead on the break-wire near the circuit, Do you think that would be enough? Please post your suggestion to improve this filtering and to make this circuit more immune to false alarm induced by RF, EMI etc. The whole circuit board will be placed in an aluminium box except for the break-wire. The power source is 9v battery. Kindest regards
I would make the tripwire input like this:
This way you keep the low quiescent current but improve the filtering. if the wire breaks.
Also: you can leave out the ferrite bead, it will not help to improve anything.
3.3 MOhms - can't be serious! A drop of water bypasses your wire and nothing is detected. Change the pullup resistor to 500 Ohm. If that's too much in current consumption, consider to have pulsed operation - say 1 ms in every half second.
No test? Again: Cant't be serious! You should have a relay that makes a test breakage somewhere, preferably in the middle of the loop and a checking plan, too, preferably automatic.
A lowpass RC filter at the input of your nand gate kills the RF, have a time constant 1...100ms.
Consider more specific detection. ON-OFF does not see the bypassing. At least make total bypassing difficult by keeping the outside ends of the loop unreachable at the same time.
The test relay: A normally closed switch is inserted as a part of your loop. The switch is remotely controlled.The simplest form of it is a relay. If opening the switch does not cause alarm, the current has a bypass way somewhere (=a bypass circuit is inserted, broken insulations and the vire is in the water)
Before any further circuit analysis a threat analysis should be performed. The most important things after deciding that the intrusion trial is probable and a tripwire is useful:
- does the intruder know beforehand that there's a tripwire (=by having some inside knowledge)
- can the intruder survey freely the systems as long as he wants (=no dogs, no cameras, no motion sensors)
- if no foreknowledge, can the intruder see the tripwire before it's too late
- what's the workaround skill level of the intruder
I spotted word "BIFILAR" in your text. Your loop seems to have 2 parallel wires. Thus it can have 4 wire ends, all available where your electronics reside. See the illustration:
The bifilarity (=2 parallel identical wires) gives a lot of new possiblities that you should explore
- In heavy disturbing fields (=powerline, radio transmitters, high current works around) your system can feed the test voltage between the 2 wires. Any disturbing signal that affects equally to both wires can be cancelled
- If there appears a resistance difference between the wires, something is added to the circuit
- if there appears a connection between the wires, something is added to the circuit
- a voltage or intentional signal between the wires can be used to do something useful. The wires can be at the same time for power supply, signal transmission and your original tripwire loop
- by measuring pulse reflections the place of the breakage can be deduced (not useful if the max distances are only a few meters, but very useful if the distance is beyond the field of the vision)
I have not written about how to protect against the lightning (not direct hit, but nearby). For that the methods can be borrowed from the landline telephony.
A way to reduce induced voltages - not much better than a twisted pair as a single wire loop and makes bypassing easy. Gathers common mode (=against the earth) voltages as easily as a single wire loop => NOT RECOMMENDED. To utilize properly the 4 end loop, you must have more complex circuit.
Use a large value resistor tree, and put the capacitor on the same side as the pull-up not the ground side. It will take a little longer to switch, but not long enough to care about.
I'd also add series resistors and a protection diode as shown to give you some protection from static and maybe even a Transorb go give you some protection from static discharge, though it won't protect you from a lightning strike directly on the fence.
Both the ground side AND the signal side need to go through the bead with a turn or four. Though you may not need it.