Self triggering Triac

[hamdi.tn]

hi everyone,

i build a circuit that power up an AC load (a pomp) and it works fine, until i tried to inject some 4KV parasites on the main power supply and i noticed that the  pomp will start without being switched on by the MCU (3.3V on R187).
The MCU itself is powered from an SMPS on the same board, but with all the tests performed on the board , the MCU is fine and immune to the 4KV burst and everything on the board continue to work quite well.

am not sure if Q30 is being switched on by those 4K burst and i should put a cap on BE junction or something else can be done with the triac itself.

any help is appreciated 

[Andy Watson]

I might have misinterpreted what you are doing but where does the datasheet(s) say that the MCO3063 or BTA..600 will block 4kv?

[hamdi.tn]

They don't, am looking for way to protect those devices.

[matseng]

The MOC3063 should be fine with a 4KV pulse I think....?

[hamdi.tn]

that isolation between emitter and output triac

[ruffy91]

1. Most likely it's the slew rate of the pulse which is triggering the TRIAC. Try improving your snubber.
It's also possible the transient is inductive coupled into the gate. Please show us your PCB layout.
2. Your MOV is limiting to 710V but you use a 600V TRIAC, switch to the 800V variant.

[Ian.M]

Your MOV max. clamping voltage of 710V@10A exceeds the 600V nominal breakdown voltage of both the TRIAC and opto-TRIAC.  (see datasheet: http://www.spiratronics.com/data/6815.pdf)   You are also probably exceeding the 600V/us dV/dT limit.   Try a choke in series with the line feed to your circuit to limit dI/dT and thus dV/dT in conjunction with your snubber, and reduce the peak current through your MOV, reducing its clamping voltage.  Worst case: You may also need to replace the MOV with a suitable bidirectional TVS diode, which can be chosen to have a considerably faster response and more closely defined clamping voltage than a varistor.

[matseng]

that isolation between emitter and output triac

That so basically isolation between the triac-side and the transistor-side then - since the transistor itself doesn't have any "internal isolation".

So this means that you can put 5KV AC at the triac side for one minute without affecting the transistor side at all - or else it wouldn't be called isolation.

[SeanB]

You would need to add line chokes to the supply, and put some MOV units, 7N251, along with a common mode choke and capacitors, on the supply as well. Dropping R185 down to 100R will help slightly by removing charge from the trigger, but it will not help much. R182 and C80 are also wrong, drop R182 to 22R 1W and C80 to 100n 275VAC Class X, and duplicate again after the choke. That should reduce triggering, though the line chokes and input filtering and MOV units will do most of the work.

[hamdi.tn]

Thanks everyone , this is a part of the whole board, a common mode choke and X capacitors are already in place between this part of the schematic and the power feed. so it's supposed to be a filtered power source, will try to modify the circuit based on your suggestions