Triac Driver using Optocoupler

[gauravkothari23]

Hi all.

I am trying to control the 24 watts LED Bulb using a Triac and microcontroller. (Circuit Attached)
i am Using PC817 to control the triac On and OFF.
I am using 2 Triac to control 2 LED Bulbs. and to make it isolated i am using optocoupler PC817.
I have two buttons on circuit which are not shown.
The two buttons are used for switching the BULB 1 and BULB 2 ON or OFF

My problem is when i switch BULB 1 ON, BULB 2 also goes ON, or when i switch BULB 2 ON, BULB 1 also goes ON.
As i have noticed that when i remove the load BULB 1 and then switch BULB 1 ON, the BULB 2 does not GLOW. but when both the BULBs are connected and i switch any one BULB, then BOTH the bulbs goes ON.

can anybody please suggest me where the problem is.
I have even checked for any track shorts. but i did not find any.

 

[KlausST]

Hi

there are dedicated "optocoupler triac drivers". Named MOCxxx for example.

They do what you need, easy to use, are made for AC
I´d definitely use them instead of playing around with optocouplers made for controlling DC signals.

Klaus

 

[gauravkothari23]

Hi

there are dedicated "optocoupler triac drivers". Named MOCxxx for example.

They do what you need, easy to use, are made for AC
I´d definitely use them instead of playing around with optocouplers made for controlling DC signals.

Klaus

agreed with you.
But even on the collector side of optocoupler, its a DC which i am switching to trigger the triac.

 

[KlausST]

Hi,

While it´s not wrong ... I still call it "handstand".

Why taking the effort to generate a DC supply to use a DC_switch to control an AC current?
I see 13 parts vs 5 parts.

If you ask me how to transport a 5 tons of sand, I´d definitely prefer a truck instead of a car.
The truck simple is made for this task ... and does it easily, and needs to drive only once.

For sure there may be reasons for taking a car: truck not available, road not suitable for a truck....

But form your posts I don´t see a reason for not using the straight forward solution.

Klaus

 

[gauravkothari23]

Hi,

While it´s not wrong ... I still call it "handstand".

Why taking the effort to generate a DC supply to use a DC_switch to control an AC current?
I see 13 parts vs 5 parts.

If you ask me how to transport a 5 tons of sand, I´d definitely prefer a truck instead of a car.
The truck simple is made for this task ... and does it easily, and needs to drive only once.

For sure there may be reasons for taking a car: truck not available, road not suitable for a truck....

But form your posts I don´t see a reason for not using the straight forward solution.

Klaus

I agree with your point and would surely implement MOC3021 opto triac in place of pc817. I think using moc3021 circuit would be a easy part.
Actually in one of my project, where I have to control a single AC motor, I have used the same circuit which worked properly.
But again when I am using 2 bulbs, in my existing circuit, I am facing such issues.
But just for a knowledge, can you share the reason,why I am facing an issue with my existing circuit.

 

[KlausST]

Hi,

to the problem:

The first I´d check is PO.3 signal and PO.4 signal.
You need to do this with a scope and try to find any capacitive coupling.
You surely know that a triac just needs a short pulse to switch ON.
So even a short pulse on the port pins may cause a problem.
(I´ve noticed your extra capacitor on the gate ... to make it less sensitve on glitches)
After this you know whether the problem is on the microcontroller side or the mains side.

The next possible problem is PCB layout.
the signal LINE needs some kind of star point wiring (to avoid problematic voltage drop).
The star point should be at pin 2 of the triac ... for each channel.
In other words: The trace from Triac_pin_2 to according capacitor C1 (C3) should not see any load current. This way you guarantee that the C1_LINE voltage is the same as triac_pin2 voltage.
Depending on the inner circuit of the LED bulb (did you check if it is designed for phase control?) there will be a high inrush current. This inrush current may shift the voltage levels on your PCB (if not routed correctly)

Please upload your PCB layout. (Mind a useful file size)

***
BTW: I dont understand the waste of bandwidth. Why is your schematic that huge?
This (the expanded! if you click on the preview) is less than 1% of your filesize. Thus it loads much faster .. especially in regions with low internet bandwidth.


Klaus

 

[FvM]

A multimeter should be sufficient to clarify the problem. My guess is short between both A2 nodes, e.g. noninsulated triacs on a common heat sink.