# Separating Phase and Neutral in AC?

I need determine if a cable is connected to a wall-socket (230V) on the other end with a timer in-between.

230VAC---TIMER---CABLE


Depending on how the timer is plugged in, it can either cut the phase or the neutral. If the timer is "ON", then it's quite simple by just putting a relay on the 2 AC-cables that shorts/closes another circuit. However, if the timer is "OFF", then it can cut either the phase or the neutral.

In case of timer cuts the phase, then there should be connection between zero and ground (making a continuity-test will show if cable is connected to the socket or not)

In case of timer cuts the neutral, the phase should provide voltage if cable is connected. To detect voltage I would simple use a "non-contact voltage detector".

Brown is phase Blue is neutral Yellow/green is ground

My main problem is how to detect which wire at the end of the cable that is the phase or the neutral. Until that I can not perform a continuity test on the correct wires (neutral and ground) and not a voltage detection on the phase.

My first thought was to rectify the AC voltage and when given DC+ and DC-, I would make a continuity-test between ground and DC-, and a "non-contact voltage detection" on DC+. If one of them is "true", then the cable is connected. If non is "true", then the cable is not connected.

I know how to build a "non-contact voltage detector" and a "continuity-tester", but when I made a simple rectifier-bridge with 4x 1N4004 diodes I measured voltage around 100V between ground and DC- and I don't want to continuity-test on that!

So what am I missing? How can I safely make a continuity test between zero and ground on AC voltage?

Best Regards Niclas Rådström

• sorry, can't quite understand... could you be more specific: from which point do you need to determine it? From the timer? or from the other end of the cable? Or from the wall?
– PkP
Commented Oct 22, 2016 at 21:04
• From the other end of the cable Commented Oct 23, 2016 at 6:49
• Hmm, so basically you'd need to test if either of the phase or neutral wires is not floating... hmm...
– PkP
Commented Oct 23, 2016 at 6:53
• I've tried to clarify my question with images in the main-post. Commented Oct 23, 2016 at 8:27

## Without a ground wire

If I correctly understand the question, it's essentially like this:

simulate this circuit – Schematic created using CircuitLab

In this case, it is possible that only one of SW1 and SW2 are closed (because the timer may only interrupt one line). The simple way to detect only the situation in which both lines are closed is to simply put a (safe!) load across the line and see if current flows. A voltmeter, as shown, or neon lamp are both simple way to detect this situations.

If these are the only wires, however, there will be no reliable way to detect which switch is open. All you will be able to detect is whether both switches are closed or not.

## With a ground wire

simulate this circuit

If you also have a safety ground wire, there is another option because the ground and neutral wires should be at the same potential. In the case that SW1 is open, neither D1 nor D3 will be lit. If SW2 is open, only D3 will be lit. If both SW1 and SW2 are closed, both D1 and D3 will be lit.

• In the cable (and all the way through the timer) to the wall socket I also have a ground wire. The ground wire is the one I'm planning to use for a continuity test together with neutral (when I can identify it safely). Commented Oct 23, 2016 at 8:38
• The usual way to do that is to use two high-CTR optocouples with about 100k current limiting resistance each, connected between L(?) and PE resp. N(?) and PE. (Please take care both LEDs may be lit in some countries which have split-phase power. Just had someone asking about Spain.) Commented Oct 23, 2016 at 12:24
• Since there's a ground wire, I've added another schematic to show how that might be detected. I've shown a simple LED circuit, but of course one could still use a voltmeter or neon bulb as well. Commented Oct 23, 2016 at 12:57
• Antiparallel diode is a good idea but 330nF/350V are pretty bulky. And give Z=10kΩ->23mA, which isn't really necessary with a high-CTR optocouple. It may also make RCD switches trigger. Commented Oct 23, 2016 at 13:24
• @janka: Good point about RCD switches; I've changed cap value to 0.18uF. Your mention of optocouplers seems to assume that another circuit will be attached, but this answer assumes that a human is directly using the indication. Not clear from the question which was intended. Commented Oct 23, 2016 at 13:49

Hmm, so basically you'd need to test if either of the phase or neutral wires is not floating... I don't know of any ready make equipment that would do it, but let's think about it for a bit. Maybe you could tie them together with a large (100k...1M resistor) then feed a constant voltage to the net through another 1M resistor, then measure if you get 230VAC, the constant voltage (say, 5 volts) or no voltage at all, when you reference it to ground.

The results would be interpreted as:

• 230V: Line wire or both wires are connected
• Constant voltage : neither wire is connected (cable disconnected)
• 0V: neutral wire is connected