I need to (preferably non-invasively) reliably detect whether a wire is live (230v AC mains), when it has minimal (maybe even zero) current flowing. It's a 'signal' wire rather than a 'load' wire. Detection is for use as input to a microcontroller.
Most solutions seem to require flowing current, but I cannot assume I have any appreciable current.

 

You don’t need appreciable current, a micro input draws only nanoamps.

I will defer to others to tell you how to do this safely as I don’t have the experience to feel comfortable recommending a circuit.

Bob

 

Have a look at this non-contact line voltage sensor project.
https://www.jameco.com/Jameco/workshop/JamecoFavorites/non-contact-AC-voltage-detector.html

 

You can put a 6.8M resistor in series with a 12k resistor. Connect the 12k to ground (common) and the other leg of the 6.8M to your load. You'll have about 100 microAmps through the pair of resistors. Measuring across the 12k resistor will show 1V p-p when 230VAC RMS is on the line.

 

You can put a 6.8M resistor in series with a 12k resistor. Connect the 12k to ground (common) and the other leg of the 6.8M to your load. You'll have about 100 microAmps through the pair of resistors. Measuring across the 12k resistor will show 1V p-p when 230VAC RMS is on the line.

In order to do that, your 6.8M resistor will have to survive a flash test, possibly as much a 4kV.
How long is the conductor in question? Could you wrap another wire around it? If you put 22M on the input to your microcontoller, you will need about 100pF of coupling capacitance to detect a signal reliably.
The other question is - how reliably do you need to detect voltage in this wire?

 

Microchip says you can connect an input to the AC line through a 5Meg resistor. I believe you would need to connect your micro circuit to earth ground, NOT neutral to make this safe, and use 2 resistors in series instead of one.

Detect AC line voltage

Bob

 

Microchip says you can connect an input to the AC line through a 5Meg resistor. I believe you would need to connect your micro circuit to earth ground, NOT neutral to make this safe, and use 2 resistors in series instead of one.

Detect AC line voltage

Bob

It would have to be connected to neutral to make it reliable. It has to survive a flash test between live and earth to be declared safe enough to sell. Not many 5M resistors will survive 4kV.

 

In that case, how about a smaller resistor connecting neutral to the circuit ground. Maybe 1 Meg?

Edited to add,I would really hate to see the circuit live if the AC line was connected wrong.

Bob

 

In 230V-land, live and neutral are regarded as interchangeable (the Schuko plug widely used, mainly in Germany can be inserted in the socket either way round), so the AC line can never be connected "wrong(ly)", because it is permitted to be connected either way round. To qualify as safe, products have to pass a flash test where live and neutral are connected together, and a test voltage is connected between live/neutral and earth.
One can never assume that neutral=earth.
So, if the TS's product is a "portable appliance" i.e. connected to the supply with a mains lead, one cannot assume that the wire he claims to be "live" is 230V AC above earth potential. The product may be connected with Shuko lead, in which case there is a 50:50 chance that his "live" wire is actually neutral.
If the product is permanently wired to the mains, then he can be reasonably sure that "live" really is live, so a capacitive technique could discover if the wire was at an AC potential above ground.
To be absolutely sure that the wire is "live", there are only two certain-to-work ways of confirming it
i) a transformer to neutral
ii) an opto-isolator with the other side connected to neutral through a resistor.

 

In 230V-land, live and neutral are regarded as interchangeable (the Schuko plug widely used, mainly in Germany can be inserted in the socket either way round), so the AC line can never be connected "wrong(ly)", because it is permitted to be connected either way round. To qualify as safe, products have to pass a flash test where live and neutral are connected together, and a test voltage is connected between live/neutral and earth.
One can never assume that neutral=earth.
So, if the TS's product is a "portable appliance" i.e. connected to the supply with a mains lead, one cannot assume that the wire he claims to be "live" is 230V AC above earth potential. The product may be connected with Shuko lead, in which case there is a 50:50 chance that his "live" wire is actually neutral.
If the product is permanently wired to the mains, then he can be reasonably sure that "live" really is live, so a capacitive technique could discover if the wire was at an AC potential above ground.
To be absolutely sure that the wire is "live", there are only two certain-to-work ways of confirming it
i) a transformer to neutral
ii) an opto-isolator with the other side connected to neutral through a resistor.

If live-to-live voltage is 230VAC in your land, what is a live to neutral voltage? And will taping into a neutral cause the GFCI to trip? Or, is one of the "live" actually neutral but you have to account for the plug being inserted either way?

 

If live-to-live voltage is 230VAC in your land, what is a live to neutral voltage? And will taping into a neutral cause the GFCI to trip? Or, is one of the "live" actually neutral but you have to account for the plug being inserted either way?

In Europe live to earth is 230V and neutral=earth, connected together where the cable enters the property, followed by a GFI/RCD (or whatever you like to call it). Earthing is generally T-C-NS or TT, though anecdotally, I have heard that some of rural Spain is IT.
Britain, Ireland, France, Belgium and Denmark have non-reversible connectors, so you can be sure that the brown wire is live within a product. Everywhere else has reversible connectors, so live and neutral can be interchanged.

 

You can put a 6.8M resistor in series with a 12k resistor. Connect the 12k to ground (common) and the other leg of the 6.8M to your load. You'll have about 100 microAmps through the pair of resistors. Measuring across the 12k resistor will show 1V p-p when 230VAC RMS is on the line.

Potentially VERY dangerous, as others point out

 

If live-to-live voltage is 230VAC in your land, what is a live to neutral voltage? And will taping into a neutral cause the GFCI to trip? Or, is one of the "live" actually neutral but you have to account for the plug being inserted either way?

Live to Neutral is 230v. Neutral to ground is a few volts (neutral is grounded at the substation in UK, so you just see voltage drop across the supply line (low) resistance varying with the total current drawn by neighbours)

 

Live to Neutral is 230v. Neutral to ground is a few volts (neutral is grounded at the substation in UK, so you just see voltage drop across the supply line (low) resistance varying with the total current drawn by neighbours)

UK (apart from rural properties) has a TC-N-S earthing system. Neutral is connected to earth where the cable enters each property. Is has been this way since PME (protective multiple earthing) was introduced in the early 1970s.

 

In 230V-land, live and neutral are regarded as interchangeable (the Schuko plug widely used, mainly in Germany can be inserted in the socket either way round), so the AC line can never be connected "wrong(ly)", because it is permitted to be connected either way round. To qualify as safe, products have to pass a flash test where live and neutral are connected together, and a test voltage is connected between live/neutral and earth.
One can never assume that neutral=earth.
So, if the TS's product is a "portable appliance" i.e. connected to the supply with a mains lead, one cannot assume that the wire he claims to be "live" is 230V AC above earth potential. The product may be connected with Shuko lead, in which case there is a 50:50 chance that his "live" wire is actually neutral.
If the product is permanently wired to the mains, then he can be reasonably sure that "live" really is live, so a capacitive technique could discover if the wire was at an AC potential above ground.
To be absolutely sure that the wire is "live", there are only two certain-to-work ways of confirming it
i) a transformer to neutral
ii) an opto-isolator with the other side connected to neutral through a resistor.

I appreciate your sincere attempts to help, but there are a couple of errors in the above.
Reversible plugs are used across most of Europe (not just Germany), but only on low power devices which are 'double insulated'. Typically, the only thing connected to mains is the 230v primary of a 5v or 12v or 19v transformer (eg phone or laptop charger, or similar). Substantial devices always have non-reversible plugs, including a ground wire; that applies in all countries including Germany.
There is ALWAYS a risk that power at the device has Live and Neutral reversed (either because wrong at wall socket or plug wrongly wired)

 

To answer questions raised:
The wire I'm needing to detect is long and permanently wired in, so I know which conductor might be live
I can separate the L and N conductors (eg to use a clamp-type sensor) but prefer not to connect to them
There may be no current flowing. Consider it like an extension lead with no appliance connected: I want to know if it is plugged in and switched on at the wall socket end.
I do NOT need to measure the voltage or current, just know if it is 'on'
The most sensitive clamp sensors I have seen give 1v output if 30A flowing. Microcontroller reads up to 3.3v. so if only 1 microamp flowing, the sensor measurement may not be reliably differentiated from zero

 

I appreciate your sincere attempts to help, but there are a couple of errors in the above.
Reversible plugs are used across most of Europe (not just Germany), but only on low power devices which are 'double insulated'. Typically, the only thing connected to mains is the 230v primary of a 5v or 12v or 19v transformer (eg phone or laptop charger, or similar). Substantial devices always have non-reversible plugs, including a ground wire; that applies in all countries including Germany.
There is ALWAYS a risk that power at the device has Live and Neutral reversed (either because wrong at wall socket or plug wrongly wired)

Unfortunately not.
The 16A Schuko plug, which has sliding earth contacts, can be inserted in the socket either way round.

 

Microchip says you can connect an input to the AC line through a 5Meg resistor. I believe you would need to connect your micro circuit to earth ground, NOT neutral to make this safe, and use 2 resistors in series instead of one.

Detect AC line voltage

Bob

That’s an old ap-note. FWIW, Microchip no longer endorses the practice described in the note. Besides the safety issues noted, even small currents injected into the substrates of ever smaller geometries have caused problems in the chip’s operation.

 

Unfortunately not.
The 16A Schuko plug, which has sliding earth contacts, can be inserted in the socket either way round.

Apologies - I find you are correct. In practice though most sockets are the French style (type E) which have the male ground pin which makes the plugs irreversible. Despite living there, I've never seen a Type F socket