Non contact mains voltage presence sensing techniques

[MadScientist]

I have several applications where I need to sense the presence of mains voltage ( 230VAC) any good ideas on non contact sensing suitable for microcontroller io.

Thanks

Dave

[tooki]

Following, since I want to do the same thing in a project.

[DavidAlfa]

Capacitive coupling, as DMMs do, like this.

[cdev]

Must be very high impedance input for it to work, I am guessing?

[FlyingDutch]

Hello,

use optocouplers with properly matched resistors on LED diode side.

Best Regards

[Haenk]

Hello,

use optocouplers with properly matched resistors on LED diode side.

Best Regards

Why? This (if commercially bought) is fully insolated, the coupling coil is inside a plastic tip.

[Gyro]

Must be very high impedance input for it to work, I am guessing?

Highish. The link that DavidAlfa posted shows a bipolar transistor front-end.

[mawyatt]

Believe you want to detect the E field from the mains. A synchronous type detection coherent with the mains (50 or 60Hz) should provide an acceptable degree of filtering to help eliminate unwanted signals.

Best,

[floobydust]

I have a cheap knockoff non-contacting voltage probe "1AC-D" that didn't work very well. It usually didn't trigger. Maybe to kill capitalist dogs. I'm not sure if it's a copy of the Fluke 1AC probe.
Found it was using a single resistor lead for the probe, not a metal flag. So I fixed that and it works very good. 2xAAA 3V power,always on and it picks up me walking by it with a high ESD charge sometimes which is a bit weird.
Inside is a single 74HC14 and a couple transistors. I can draw the full schematic if needed.

It uses a series string of qty. 12 of 22MEG (=264MEG) resistors from the probe to ground, and a 10MEG to the 'HC14 input. It looks like the AC picked up there retriggers a one-shot, and a few gates make the beeper oscillator.
edit: added pic of pc board to show resistor array

[MadScientist]

I have seen a few circuits using  mosfet input fed from a wire wrapped around the mains live. Any comments ?

[MadScientist]

Hello,

use optocouplers with properly matched resistors on LED diode side.

Best Regards

Why? This (if commercially bought) is fully insolated, the coupling coil is inside a plastic tip.

And optocoupler isn’t non contact , the led is in circuit

[mawyatt]

Since you are considering a microcontroller, use an output pin to commutate (toggle) the sense amplifier gain in a +- fashion, then low pass filter the resultant and follow with a level detect (maybe just a Schmidt Trigger (for hysteresis) type microcontroller input pin). This produces the well known Synchronous Detection scheme which has benefits of good noise (unwanted signals) rejection, and should help with false sense indications.

Best,

[joeqsmith]

Depending on your requirements,  if you want to measure the voltage you should have a look at the patents for Fluke's non contact volt meter.   

[Gyro]

Didn't Fluke have problems with that and stop offering the products? My memory might be defective though.

It would be much easier with a stable coupling capacitance though.

[floobydust]

The first generation Fluke 1AC-AI VoltAlert testers (released 10/2009, sold through 2/2010) were recalled "The testers can fail to give an indication of live voltage..."  Second generation Fluke 1AC-II VoltAlert testers are OK.
FYI latest Klein Tools Non-Contact Voltage Tester Model NCVT-1 is recalled (almost 1.8M units 6/2021) on/off button gets stuck. Low end manufacturing in china is surely worth it.

What I got for $1 is a (defective) Fluke knockoff which needed the probe fixed. I'll use it when in electrical panels that have multiple sources, a quick check for if power is off everywhere. Of course I use a multimeter when in there.

OP can calculate (or prototype and measure) how much capacitance is required for the sensor. I imagine a few turns of insulated wire wrapped on a mains live wire would be good.
The approach using an MCU (like multimeters with non-contacting alert) are likely just a CMOS gate input looking for transitions.

[joeqsmith]

Didn't Fluke have problems with that and stop offering the products? My memory might be defective though.

It would be much easier with a stable coupling capacitance though.

It appears they still offer them.   Patent is a good read.
https://www.myflukestore.com/category/fluke_electrical_testers_us

Videos showing it in operation.