# Zero Crossing Signal Timing

I have a zero crossing circuit for dimming lamps exactly like the below

According to my calculations:

$$\ I_{R7} = \frac {V}{R} = \frac {3.3}{10k} = 0.33~mA\$$.

Considering TCMT1109 transfer ratio equal 200% (worst case), when the input current get lower than $$\0.165~mA\$$ the ZC signal should start rising.

$$\ V = {R}*{I} = {200k}*{0.165~mA} = 33~V\$$.

Adding the optocoupler and the bridge voltage drop, let's assume $$\35~V\$$ the minimum input voltage to detect the zero crossing.

The mains voltage is $$\220~Vac, V_{peak}=311,12V\$$, and the frequency is $$\60~hz\$$.

$$\ V(t) = V_{peak}*sin(wt) \$$

$$\ 35 = 311,12*sin(2*pi*60*t) \$$

$$\ t = 0,299 ~ms \$$

After calculations, I was expecting the ZC signal start rising at least at 0.299 ms, but after measure at oscilloscope I got the ZC start rising around 0.450 ms.

I know this is a very low time, and I can manage that at the microcontroller firmware. I'm just worried because I can't get the 0.450ms (50% higher than 0.299ms) at any calculations I make.

Where am I missing at calculations? I considered the lower CTR for worst case.

Where is my analysis wrong?

• Any particular reason you chose that side of the bridge to put the resistors? Mar 24, 2019 at 0:58
• @SpehroPefhany I chose the resistors after bridge because of my pcb layout. If you tell me is "very very much better" before the bridge I can study in my pcb if that would be possible. Mar 24, 2019 at 18:31
• Consider behavior with a 1kV transient on the mains. Mar 24, 2019 at 19:27

Where am I missing at calculations? I considered the lower CTR for worst case.

No you did not consider your worst case CTR.

Look at the datasheet for your opto-coupler TCMT1109

The CTR varies with I(f)

Your CTR at very low I(f) will plummet to only 0.1 - 0.2 of the expected value.

I work on frequency meter and i make board for sampling, those 4 images all from my work. in this case i read signal from line with PIC18 microcontrol . You most use capacitor for charging energy and use this at . by this way you can use low power resistor and fast zero crossing detection (for good accuracy, use falling edge) also with change R1 R2 R3 use more input voltag.

also this output

I offer this Practical circuit

for best answer use gap under opto-coupler