zero cross problem pls help

[takbbb]

i connect 1M resistor from main 110v ac to MCU

however, the zero cross is uneven

and i found that the waveform to mcu is a bit strange ..

as below

the point is not really drop to zero, and the scale have changed??(why??)

and this the resultant waveform fm my mcu

the falling edge or rising edge interrupt wase @ abt 2.5V

pls help and suggest thx~

 

[VVV]

This is to be expected, since the threshold of your MCU input is not going to be zero, but somewhere in the middle of the supply voltage. Variations on the threshold will also affect the duty-cycle of the waveform.

The shape of the voltage at the input of the MCU is due to the input protection diodes that the MCU has built-in to protect its inputs. This is actually why you could simply connect a resistor from 110VAC to the MCU input without destroying it. The diodes will clamp the voltage to safe levels.

If you need true zero-crossing, then you can use a comparator to detect the actual zero crossings of the waveform. The reference input can be connected to GND if you have a negative supply for your comparator.

If you do not have a negative supply, then divide down the 110VAC to some reasonable value of a few volts peak to peak and then couple that to the comparator through a capacitor. The reference input of the comparator can be biased in the middle of the supply voltage. Make sure you bias the actual input of the comparator from the same supply through a relatively large resistor. This way the two inputs of the comparator are at the same potential and since the signal is capacitiely coupled, you will detect true zero-crossings. Of course, they will still be affected by the comparator's response delay and hysteresis (which you I recommend you add to remove noise), but the results will be far superior.