This is the interesting part of a relatively simple NRF24L01 based mains door bell repeater project.
In principle, it is nothing more than a "home made" opto-coupler consisting of a neon bulb, a photo transistor and two 120K resistors in series. It fits into an insulating tube to isolate it from the low voltage electronics.
The components are probably non-critical in this application but these are the ones I chose:
neon part number 5AB-BT (manufacturer: VCC)
phototransistor BPW85
I used a plain (no copper) perforated board and made a cutout to (hopefully) prevent arcing across the surface.
Completely forgot about neon bulbs. Goes to show that some of the old technology can provide the simplest solutions.
Do you get pulses out of the phototransistor?
Could be useful for other purposes, like providing a clock signal.
Good question. I didn't put a scope across it so I can't say at the moment and it is now installed. However, I have a test version of the detector but minus mains lead so, I'll fix that up, do a test and publish the results since I'm also curious about that myself now. Having said that, a clear neon bulb (no phosphor coating etc.) should not have much persistence and neither will the photo transistor so I guess that mains pulses will get through.
Most of the published circuits for mains voltage detection do seem to be led opto-coupler based and usually quite elaborate, some even with bridge rectifiers. This application is intermittent but, in some applications, say mains failure detection, the lifetime of neon bulb could be an issue. The strike voltage can increase with age and the brightness diminish.
This, incidentally, is the circuit I used to interface to it. It is powered by a 1.5v AA cell so contains a boost converter and latch.
I got a few minutes to do a trace with just a detector part similar to that described in the OP.
I used a 3v battery pack wired as so:
3.0V --> 47K --> PT collector PT emitter --> 0V
The measuring was across the 47K resistor.
Our mains frequency is 50Hz.
I used averaging on the trace otherwise it was very smeared.
(I've just seen after putting the scope away that the scale is barely visible on the picture. Anyway, the screen is 12 horizontal divisions by 8 vertical divisions)
EDIT
Scope picture improved and resistor value changed to 47K
I think the problem is due to the resolution of the picture.
If you left click your mouse on the picture, then the image gets displayed full size and the scale becomes clearer.
In what way is detection of the mains related to doorbell repeating? Does your doorbell work off of a 16 to 24 volt transformer? I’m just wondering how the project all comes together.
It works like this. The door bell is powered by 230v. The bell push button delivers 230v to the bell. There is no transformer. This arrangement complies with the regulations here. I wanted a remote repeater to indicate that the bell push had been operated so this meant detecting 230v across the bell terminals. Repeater in this context means driving a remote audio device. Hence the circuit that I published and it works perfectly so far.
I use cheep wall wart (similar to a phone charger) and use that output. Most have a capacitor so that will hold the pulse for a bit making it easier to detect.
