Zero crossing detector for Arduino using 4N35 optocoupler

[Wrichik Basu]

TL;DR Summary

Can the 4N35 optocoupler handle 6 VAC in its input, or do I need to rectify it to DC?

I am designing a zero crossing detector for Arduino using the 4N35 optocoupler. This is just a small part of a large project. I haven't bought the components yet, so do not ask me for readings. Everything is in the planning phase now.

My input will be a 6-0-6 transformer (220V to 6V step down).

I was having a look at this website where they have controlled an AC fan using an Arduino and a TRIAC. They have fed the 9 VAC output from their transformer directly to their 4N25 optocoupler:

In https://forum.allaboutcircuits.com/threads/zero-crossing-with-4n25-issue.138690/, they have first rectified the AC to DC, and then fed it to the optocoupler (don't look at the input; I will be using 6 VAC):

The above circuits use 4N25, but I believe using a 4N35 won't create any problem. Do I need to rectify the input to the 4N35, or can I directly feed it 6 VAC?

 

[Rive]

Well, 4N24 and 4N35 both are with low reverse voltage for this, I think.
I would keep the transformer only, but put a LED in reverse, parallel with the opto.
Or a different opto, maybe with dual LEDs on the input side (something from here, for example).

I guess you are aware that it'll detect a time range around the peaks, not the crossings?

 

[Wrichik Basu]

I guess you are aware that it'll detect a time range around the peaks, not the crossings?

Yes, I am aware of that. I need the time range around the peaks because I will be using that information to switch on a TRIAC for a short period of time, thereby dimming the appliance connected to the TRIAC. Similar to what has been said in the first website linked in the OP.

 

[alan123hk]

It seems that if there is a safety isolation transformer, it may not be necessary to use an optocoupler, but certainly the optocoupler can serve as a secondary fault protection.

As mentioned in response 2, please make sure that the reverse input voltage does not exceed the rated value.

Do I need to rectify the input to the 4N35, or can I directly feed it 6 VAC?

The rectified waveform is very different from the original sinusoidal wave waveform, so the two corresponding output waveforms are also very different. It depends on your design and the processing power of the Arduino hardware and software.

 

[Wrichik Basu]

As mentioned in response 2, please make sure that the reverse input voltage does not exceed the rated value.

Yes, I will see to that. Maybe I will use a 5V transformer instead of 6V.

The rectified waveform is very different from the original sinusoidal wave waveform, so the two corresponding output waveforms are also very different.

That's what I was thinking about, and the two designs are basically not interchangeable. Thanks for the clarification.

 

[Rive]

Maybe I will use a 5V transformer instead of 6V

You have to work with the peak voltage instead of RMS
5V is still too much if you don't limit the voltage by some other means.

 

[Wrichik Basu]

You have to work with the peak voltage instead of RMS
5V is still too much if you don't limit the voltage by some other means.

Fine. Resistors are always handy in such cases.

 

[Rive]

Be careful, serial resistors are no good for limiting maximal reverse voltage load.

 

[Wrichik Basu]

Be careful, serial resistors are no good for limiting maximal reverse voltage load.

Then how do I limit the voltage? (I can't seem to find a 5V transformer, so I have to use a 6V transformer.)

 

[DaveE]

The absolute maximum reverse voltage spec for the input LED is 6V. You can put in an anti-parallel diode to clamp this at about 0.7V. Read the data sheet for any IC you might use.

 

[DaveE]

You might consider bi-directional opto-couplers like these too:
https://www.vishay.com/docs/83608/h11aa1.pdf

I think the whole HV side could be a resistor and a bidirectional coupler. If the pulse is too short you could insert back to back zeners in series with the resistor to increase the off time. Because there is always a current path, the only HV will be across the resistor.

You may find some with a darlington output which may allow you to have less power dissipation in the HV resistor. Like this: https://datasheet.octopart.com/IL755-1-Vishay-datasheet-10103.pdf

 

[Wrichik Basu]

You might consider bi-directional opto-couplers like these too:
https://www.vishay.com/docs/83608/h11aa1.pdf

I am actually buying both types. I have to do some testing to see which one actually suits my case. My final aim is to use a triac as a dimmer — I will use a MOC3021 triac driver, and I need to detect the zero crossings because I will fire the triac driver accordingly for a dimming effect.

You may find some with a darlington output which may allow you to have less power dissipation in the HV resistor.

The two websites from where I buy things don't have an optocoupler with darlington output.

 

[DaveE]

The two websites from where I buy things don't have an optocoupler with darlington output.

Not even 4N33 et. al.? I thought those were everywhere. You could use two in parallel for full wave.

 

[Wrichik Basu]

Not even 4N33 et. al.? I thought those were everywhere. You could use two in parallel for full wave.

4N35 is available, and it's already in my cart. H11AA1 is also available.

 

[maxwells_demon]

The following snippet is from a 4N35 Vishay Datasheet (https://www.vishay.com/docs/81181/4n35.pdf)

Note that 6Vac is an rms value. The peak voltage would be at 6√2 = 8.485 Vpk. Note that we are talking of ac voltage, and based on your schematic:

https://www.physicsforums.com/attachments/273618

you will be subjecting your 4N35 to 16.97Vpk-pk. I suggest you put a bridge rectifier circuit at transformer secondary to limit the voltage to 8.485Vpk, and further lower the secondary peak voltage to less than 5V by doing a resistive voltage divider. The good news is that even with these modifications, the integrity of your pulse output is not compromised. it will still fire at the zero crossing.

P.S. i implemented a similar circuit to this in my college days for a friend.

 

[DaveE]

Note that 6Vac is an rms value.

No! It is an instantaneous value. You won't see RMS voltage spec on a diode (maybe on a diode circuit, or application). RMS is only useful for AC signals and specifying an AC voltage limit on a diode is, at best, confusing. In any case the issue is damage to the PN junction which can happen very quickly.

 

[maxwells_demon]

No! It is an instantaneous value. You won't see RMS voltage spec on a diode (maybe on a diode circuit, or application). RMS is only useful for AC signals and specifying an AC voltage limit on a diode is, at best, confusing. In any case the issue is damage to the PN junction which can happen very quickly.

I meant 6Vac as the output of the transformer in the OP's schematic.

though now that i read it more carefully, OP did say:

My input will be a 6-0-6 transformer (220V to 6V step down).

so it's basically a transformer with two secondary windings, only one of which would be connected to 4N35. So scrap what i said about 16.97V pk-pk. Though it still is 6Vac, and therefore 8.485V pk-pk.

 

[maxwells_demon]

Scrap the bridge rectifier, but the voltage divider is still necessary to bring that 8.485Vpk-pk down to 5Vpk-pk.

 

[DaveE]

I meant 6Vac as the output of the transformer in the OP's schematic.

Yes, I see that now. Sorry, I misunderstood you! That was covered pretty thoroughly earlier in the thread.

 

[Wrichik Basu]

I am actually thinking of switching to H11AA1 instead of 4N35. My circuit is not yet complete, but am I correct in saying that the H11AA1 will detect a time around the crossing rather than the peaks?

 

[Tom.G]

I am actually thinking of switching to H11AA1 instead of 4N35. My circuit is not yet complete, but am I correct in saying that the H11AA1 will detect a time around the crossing rather than the peaks?

Either will work, it's just that the programming will be slightly different.

The 4N35 will be 'On' whenever the voltage is above about +1 Volt.
The H11AA1 will be 'On' when the voltage is above about 1 Volt of either polarity.

That's about 7 to 10 degrees from zero crossing.

For the 4N35 the coding would be:
Detect a change of input
If changed from 'Off' to 'On', a positive going Zero crossing just occured
If changed from 'On' to 'Off', a negative going Zero crossing is about to occur

For the H11AA1:
If is 'OFF' you are within 10 degrees of a Zero crossing in either direction

Cheers,
Tom

 

[MalcolmB]

Have you thought of using a MOC3041 zero crossing opto triac?

 

[DaveE]

Have you thought of using a MOC3041 zero crossing opto triac?

I'm not sure this would help much. IIRC, he wants to make a phase angle dimmer circuit and needs zero crossing information to know when to turn on a thyristor, which will then self commutate at the zero crossing.

 

[Wrichik Basu]

I'm not sure this would help much. IIRC, he wants to make a phase angle dimmer circuit and needs zero crossing information to know when to turn on a thyristor, which will then self commutate at the zero crossing.

Exactly, you are right.