using moc3023 triac optocoupler to turn on triac

Thread Starter

denison

Joined Oct 13, 2018
328
I have been using the moc3023 random phase opto to try to turn on a triac. It has been turning on the triac at full power. I have tried the snubberless bta16-600cw triac . Still getting full power on the load.
I have been turning on the emitter side of the opto at 5ms after zero crossing. I am using 5 volts here. I measure a dc voltage of about 2.5v which would be the correct average if the 5v signal is applied at half way through a half cycle of the mains.
The recommended triac snubber is a 39 ohm resistor and .01uf capacitor. I used a .047 uf capacitor as I do not have the .01 at the moment. This didn't work. I then tried the opto snubber of a 470 ohm resistor and .047 capacitor. I blew a 1/2 watt resistor. Probably need a 5w here.
Before I get the correct recommended parts does anybody have a PROVEN circuit with the moc3023 for turning on the triac with no false triggering? This would include a suggested triac and snubber components and their ratings.
The capacitors of course have to be mains rated.
 

Ian0

Joined Aug 7, 2020
9,671
First, get it working on a resistive load. Don't use a sensitive gate triac. Use something like BTA12-600BW. The circuit will work with 100Ω between the output of the MOC3023 and the triac gate (the other MOC3023 terminal goes to MT2) and nothing else (no gate-MT1 resistor and no snubber). The most important thing is to turn off your trigger pulse at least 500us before mains zero crossing. Triacs are slow to turn off, and if you don't switch it off soon enough it will remain on into the next cycle. Accuracy of the zero-crossing detector is important.
You may need a gate-MT1 resistor if you use a sensitive gate triac, but if you are driving it with a MOC3023, a sensitive gate triac is NO advantage.
Then you can add a snubber if it needs it (100Ω/100nF or 47Ω/47nF, use half-watt resistors and class-X capacitors) it deals mainly with the switch-off on an inductive load. It should also have a choke in series with the live output because otherwise the leads will radiate interference. Use something like 150uH to 1mH on an iron-powder toroid.
 

AlbertHall

Joined Jun 4, 2014
12,345
I have been turning on the emitter side of the opto at 5ms after zero crossing.
You must remove the trigger signal before the end of the cycle. Once the triac is triggered it will stay on until the end of the cycle so you need only a brief trigger pulse. If the trigger is still present at the end of the cycle then the triac will continue conducting into the next cycle.
 

Thread Starter

denison

Joined Oct 13, 2018
328
First, get it working on a resistive load. Don't use a sensitive gate triac. Use something like BTA12-600BW. The circuit will work with 100Ω between the output of the MOC3023 and the triac gate (the other MOC3023 terminal goes to MT2) and nothing else (no gate-MT1 resistor and no snubber). The most important thing is to turn off your trigger pulse at least 500us before mains zero crossing. Triacs are slow to turn off, and if you don't switch it off soon enough it will remain on into the next cycle. Accuracy of the zero-crossing detector is important.
You may need a gate-MT1 resistor if you use a sensitive gate triac, but if you are driving it with a MOC3023, a sensitive gate triac is NO advantage.
Then you can add a snubber if it needs it (100Ω/100nF or 47Ω/47nF, use half-watt resistors and class-X capacitors) it deals mainly with the switch-off on an inductive load. It should also have a choke in series with the live output because otherwise the leads will radiate interference. Use something like 150uH to 1mH on an iron-powder toroid.
Good advice Ian. I can see that if I don't switch off trigger pulse soon enough it will remain on into next cycle. I am using a BTA16-600CW which is a snubberless type as is the one you suggest. No disadvantage in switching off the pulse earlier as the triac continues to conduct until the next zero crossing.
I have previously used zero crossing circuits recommended by others that didn't work. The Arduino software couldn't pick up zero crossing. I had to design my own which works great. The pulse is only .28ms long. This is picked up on the Arduino software which gave me the pulse length.
There seems to be no IC available that gives good zero crossing detection. My detector has quite a few parts which could easily be integrated onto an IC cutting down on the real estate required.
 
Last edited:

Thread Starter

denison

Joined Oct 13, 2018
328
You must remove the trigger signal before the end of the cycle. Once the triac is triggered it will stay on until the end of the cycle so you need only a brief trigger pulse. If the trigger is still present at the end of the cycle then the triac will continue conducting into the next cycle.
Hi Albert, Ian o has suggested the same thing. I will keep the trigger pulse brief. This is obviously what has been happening.
 

Ian0

Joined Aug 7, 2020
9,671
There are a lot of different zero-crossing detector circuits - it depends on where you start from!
If you have a transformer with a single winding, or a transformer with a split-phase winding, or if you are working it directly from the mains - each needs a different circuit. The only IC I have found that incorporates it is the ATM90E26 energy meter IC (and other similar devices), but no use buying one of those just for its zero-crossing detector! (even though it is cheap)
I usually aim for a 500us pulse, but I turn the trigger pulse off after 9ms. You can use a very short trigger pulse, but only if your load current exceeds the holding current (see datasheet), otherwise the triac will switch off again at the end of the pulse!

My least successful zero-cross circuits have been when I used the comparator built into a microcontroller. I concluded that most micros don't have very good comparators. (You can quote "bad workman blames his tools" back at me if you want)

I have used the "simple" circuit (BTA16-600BW + MOC3021 + ONE resistor) for many years, built it thousands of times in production - never had a problem with it; no snubber but an LC filter for EMC purposes.

A further note - just for interest - the zero-crossing detector type optos (MOC3041 etc) don't work with half-wave rectified loads and snubbers - the snubber cap charges up and keeps the opto OFF. Half-wave rectified loads are generally prohibited by EMC regulations, but occasionally you'll find one which causes trouble.
 
Top