Voltage Leakage

It's difficult to understand what you mean by "leakage".

If you're measuring this voltage on conductors which are not terminated - i.e. not connected to anything - and you're measuring with a high impedance voltmeter (e.g. a DVM) , you will inevitably measure large AC voltages - picked up by inductive and/or capacitative coupling.

If these conductors are not used, connect them to earth/ground at one end. If your measured voltage still doesn't disappear, try measuring it with an old-fashioned AVO, or put a 1K resistor across the test leads.

N.B. If the resistor is really seeing 120V, it will be trying to dissipate about 14W, so (unless it's a big'un) it will give up the smoke (which is its electrical life) and die.

<...put a 1K resistor across the test leads...>

Or a filament lamp.

we measured both conductors terminated and not terminated. those already terminated, we disconnected it and taped the other end and measure the voltage between conductor and neutral (earth). the neutral wire s is not part of the 15 cores cable. we already did like what you suggested to earth the spare conductors when we first encounter the problem. it reduced the voltage reading by 40 - 50% only and the relay still rattle if we connect to coil any wire and the neutral wire if one of the wire is connected to the supply source which is 380/220 control transformer. btw thanks for the knowledge you shared.

This is an "I'm Desperate - Get Me Out Of It" solution. With a connection to a relay coil which is rattling, put a resistor in parallel with the coil. Make it about half the resistance of the coil (which you can measure with a meter) and make sure it can take a bit of power without going puff. 1W should do it - but do some sums to check (or make it 4X the VA rating of the relay coil to be sure).

The proper solution is to use suitable wiring (individually shielded twisted pairs, with the screens grounded at one end only).

N.B. You could use a filament lamp instead of the resistor!

Check for all terminal connections.Is it as per the circuit?

These voltage 120vac you measure from which terminal, do you measure from live to ground?

You should go back to school and learn a bit about electrical stuff . . .

Voltage does NOT leak . . . it is current that does the 'leaking' and voltage is the 'indicator' of the problem.

Your problem is caused by having power and signal wires running in the same cable. The "shielding" you are talking about is likely on the outside of the cable - YES? This does not do anyone any good . . . .

You need a cable wherein EACH PAIR OF WIRES will be shielded separately . . . to separate the AC power circuits from the signal circuits.

voltage is not the indicator . . . . voltage is the cause . . . when the isolation is breached (broken) then the resistance offered is lowered and the voltage is able to overcome the resistance to establish the current . . . the current is the effect . . .

also - current in one conductor can create a magnetic field in the neighboring region leading to an induced emf (voltage) across other conducting elements . . . resulting in currents being established in those conductors if a closed path is available . . . . in such a case no breakdown of isolation or reduction in resistance of the insulator may be in evidence . . .

I would suggest that you change the system to 24VDC over the cable to elimate impedance caused over this 50m cable.

Joe

yup we did it already and its working well. I only want to know the cause of leakage.

thanks a lot for advised.

Perhaps you missed the point about the digital volt meter. Your standard electronic volt meter has a very high input impedance. This makes it sensitive to the slightest inductive link which you are calling leakage. Long parallel wires link magnetically the same way as they do in a transformer. This is why you should separate power feeds from control lines. The longer the run of wire, the better the linkage.

Realistically you can not transfer much power this way as it is very inefficient. But you will detect the presence of a voltage with certain meters. An analog meter may help but it really depends on linkage to power lines.

If you live near overhead power lines that run parallel to a nearby fence, you will be able to detect a voltage drop along the run of the fence.

You are using shielded cable. Although this signaling cable should be separate from the power cables, properly installed shielded cable should eliminate induction from power cables. There are two likely causes:

1. Does the cable have one common shield around all the cores? If so, the one live wire is probably inducing voltage into the others. Shielding does not help if the the source field and the affected conductor are both inside of the shield. The ideal cable for this installation is twisted shielded pair. Each signal supply and return pair is shielded separately, protecting it from the others.

2. Is the shield connected to earth at only one end? If the shield is not connected to earth it is useless. If it is connected at both ends it can actually make things worse. Voltage induced into the shield causes a current to flow through a loop consisting of the shield and earth. That circulating current is right next to the conductors and will induce voltage into them.

Use a 24 volt system instead of the 220 VAC control relay system. See to it that all shielded wires, the shields are terminated to ground /earth. Then run all signal wires through or contained in a separate metallic conduit (do not use PVC) which should also be grounded. Be sure all ground wires / shielded wires share the same ground plane.

cable shield is used to protect interference cost by EMC.

take note not only inductance is present in the circuit we do still have CAPACITANCE.

The voltage that you have measured is caused by cable capacitance due to a long distance. We have line to ground capacitance and line to line or cable to cable capacitance. Theoretically, the current can transfer to other cable by routing on the capacitance of 2 cables. This capacitance completes the circuit that's why you can measure voltage even though it should't have.

Solution is to reduce this current cost by capacitance by placing resistance.

Actually reduce the voltage across the coil so that it will be less than the hold in voltage.

Design of a shunt resistor to limit the voltage across the coil. That's the solution.