zero crossing detector

Take the period of a 50Hz sine wave. Now convert your "nanoseconds" into the change in angle that it represents for a signal with that period.

Now take that change in phase angle and calculate the maximum change in voltage that can occur during that change in phase.

Now tell me how you are going to keep noise on that signal well below that change in voltage..

How then to do it is fairly simple. Connect your signal through a resistor to base of a transistor connected as common emitter, with a diode between base and emitter to protect against reverse voltages. Take the output pulse train from the collector, with a resistor going from collector to your desired dc supply.

That will give you your pulse train of zero crossings. If the lag, before the mains voltage rises enough to switch on the transistor, is too great, include another resistor to make the base a summing point and add an offset voltage to decrease the lag to "zero".

You can achieve zero lag time with that circuit - although noise and thermal effects *will* cause errors - including causing the zero crossing point to wander. So, if you want to improve that, you will need extra complexity. If you can hold the temperature of the transistor reasonably constant, it won't wander much.

Simple enough?

with an operational amplifier, taking the - input to the ground. If your signal has too much noise you should use a Schmitt-Trigger circuit, it comes on any book.

Best regards.