proposed zero voltage crossing switch

Thread Starter

denison

Joined Oct 13, 2018
328
Hi All, I am trying to design a mains zero crossing switch. Refer to the attached schematic. I don't have a Ltspice model for the moc3043 zero voltage crossing triac optocoupler so consider that the transistor opto is a moc3043. There is no simulation of this circuit.
According to the specs 20 volts is the maximum allowed by the zero crossing circuit. I attached the capacitor to read the maximum zero crossing voltage.
Both resistors blew immediately when 5v was applied to the emitter. The zero crossing circuit on the opto should have stopped higher voltages but this appears not to have happened.
Can the result be explained?
 

Attachments

AlbertHall

Joined Jun 4, 2014
12,345
The MOC3043 triggers a triac at the start of the cycle and it will continue to conduct until the end of the cycle. The full, peak, supply voltage will be applied across the two resistors, bridge, and ultimately to the capacitor. If the capacitor is rated for at least 400V then it would be charged to the peak of the supply voltage. If its voltage rating is lower then probably the capacitor will fail short circuit leaving the resistors connected across the supply and the magic will be released.
 

Alec_t

Joined Sep 17, 2013
14,280
The MOC3043 may need as much as 5mA input trigger current (but no more than 20mA), so the input resistor R1 should be in the range (5-1.4)V/20mA = 180Ω and (5-1.4)V/5mA = 720Ω.
According to the specs 20 volts is the maximum allowed by the zero crossing circuit.
? I don't see that in the datasheet. The triac can withstand 400V. The input diode has a reverse voltage limit of 6V and a forward current limit of 20mA.
 

AlbertHall

Joined Jun 4, 2014
12,345
I believe that the TS is referring to the part of the specification copied below but is misunderstanding - it does not prevent voltages above 20V. It means that if the supply voltage is above 20V when the opto is turned on, it will not trigger the triac until the start of the next cycle.

1622284512294.png
 

Thread Starter

denison

Joined Oct 13, 2018
328
The MOC3043 triggers a triac at the start of the cycle and it will continue to conduct until the end of the cycle. The full, peak, supply voltage will be applied across the two resistors, bridge, and ultimately to the capacitor. If the capacitor is rated for at least 400V then it would be charged to the peak of the supply voltage. If its voltage rating is lower then probably the capacitor will fail short circuit leaving the resistors connected across the supply and the magic will be released.
Thanks Albert. I have built many circuits using the moc3043 but always with the triac attached. You can see that the resistors are not going to blow then because when the triac fires it puts equal voltages across the input and output of the moc3043.
I misunderstood the 20 volts in the specs as you said. For anybody watching this thread the best triac for ease of firing is the sensitive gate triac (isolated tabs) BT138X-600E. I had a lot of trouble with other triacs. Some would only fire ,for instance, to turn on a half cycle of the mains.
Again thanks. I completely misunderstood the working of moc3043. Still looking for a good zero crossing switch to time the pulses for phase control of the mains. I have seen quite a few examples using a 240vac to 24vac transformer. I would like it simpler than this without the transformer if possible.
 

eetech00

Joined Jun 8, 2013
3,859
Hi EETECH, Where did you get your moc3043 model? I could do with this for my Ltspice simulations. I could also use a triac.
The model is from the LTspice group at groups.io. I modified the symbol to include pin numbers and nomenclature.
I've included it in the attached file along with a Triac library file. The Triac library includes a 12A 600V triac equivalent and is used on the included example Test circuit.
:cool:
 

Attachments

Thread Starter

denison

Joined Oct 13, 2018
328
This zero crossing circuit has been discussed here before and to my knowledge has not been bettered.
View attachment 239984
Hi Albert, I have modified your circuit a little so I could run it on a Ltspice simulation. I cut out the 1n capacitor as the mains needs a path through the diode bridge. It seems to serve no purpose. I used a bc337 npn. I have also designed my own zero crossing detector but it has more parts. I didn't like the negative going voltages in your circuit. In the simulation of your circuit the ZCV appears to to be 1ms wide.
Anyway surely there is a single IC containing all these parts so that we can reduce our component count. I believe I may have found one from Fairchild . It is their UA742. Or if you know of a good one please let me know.
 

Attachments

Ian0

Joined Aug 7, 2020
9,671
Cutting out the interference filter to save money is the first thing people do. It usually has to go back in, after the first units get to the customers.
 

Thread Starter

denison

Joined Oct 13, 2018
328
I used to use the CA3059
https://www.onsemi.com/pdf/datasheet/ca3059-d.pdf
but it is obsolete as well.

I'm not sure why you need a zero-crossing detector when you are already using an opto with a zero-crossing detector?
It is for phase control. You can switch the triac on at zero crossing of the mains but this is not phase control. You have to know when zero crossing occurs to time your pulses for phase control of the mains. So far I have not been able to find one that I can use to pick up the zero crossing with software. The pulses are too short. The circuit by Albert may do it as the simulation gives a zero crossing pulse of approximate 1ms long.
 

Thread Starter

denison

Joined Oct 13, 2018
328
Interesting that it has a minimum specified frequency of 60Hz but I can't see any reason why it wouldn't work at 50Hz.
Interesting that it has a minimum specified frequency of 60Hz but I can't see any reason why it wouldn't work at 50Hz.
So far I have not been able to find a supplier of the UA742. I have contacted Fairchild to find out where I can get one. Fairchild has been taken over by OnSemi. You circuit may be just the thing for picking up zero crossing with software as the simulation shows a pulse length of 1ms for the zero crossing voltage.
 

Thread Starter

denison

Joined Oct 13, 2018
328
Just to say this is not my circuit and I don't know who designed it but it has often come up in this forum.
Hi Albert, I am going to use the random phase triac opto moc3023. File attached for recommended 240vac application. Just one problem I can see. If the triac gate conducts current before firing and the firing is set when the mains is at high voltage. Wont the resistors blow just as they did with my moc3043? If the triac gate doesn't sink any current before it fires then no problem.
I guess that must be the case as the manufacturer has suggested the circuit shown.
Am I missing something here again?
The Ltspice simulation of the circuit you suggested showed a zero crossing pulse of 1ms long. I made my own circuit and found using Arduino software the pulse was 0.28ms. Haven't built the circuit you suggested yet.
 

Attachments

Ian0

Joined Aug 7, 2020
9,671
A triac is pretty slow - it takes about 1us to switch on. During that time your 360Ω resistor will have full mains voltage across it, and will dissipate 150Watts, but 150 W for 1us, 100 times a second is an average of 15mW, and that's the worst case scenario if you were triggering it at mains peak. If you are triggering it at zero crossing, it will not switch on until Vmains/360Ω exceeds the trigger current (35mA). The worst case dissipation is then 440mW, which will last for approximately 175us - average power dissipation about 4mW.
If your triac MT2 terminal becomes disconnected from the load and the opto-triac remains connected, the opto-triac will blow up the resistor, so make sure you track the pcb so that it can't happen.
Also note that the 360Ω resistor must withstand the full mains voltage, so make sure you use a component rated for at least 375V (and that excludes most surface mount resistors smaller than 2512).
To use a 1206 resistor (most of which are rated at 200V), put two 180Ω in series.
 
Top