Triac switching 220v@80W = funny smell

[edit]This is the corrected schematic.[/edit]

Legend:
T1 - BT139-600
X1 - mains input
X2 - output to external device

The problem is that it's burning the resistors on the 220v side of the optocoupler when I increase the load to 80W but that's working as expected since the resistors I'm using are rated at 1W.

Taking the context that I want to switch on/off a light bulb of ~80W, my questions are:

  • Are those resistors (R19, R20, R21) really needed there ? Yes they are, BT139 datasheet clearly states that.
  • R22 and C13 are also needed ? Acts as a Snubber - Wikipedia, so it's required.

Yes I read it.. otherwise how would I figure out the pinout ? :smiley:

For instance the 360R resistor: v=220 r=360 so wattage should be ~135w.. 2x thats 270w.
Looking at Farnell and Digikey the max wattage I see for a 360R resistance is 6W (Farmell) and 10W (Digikey)..

And I know how dangerous it is, don't worry with that part.

For the BT139 I found:

  • Peak gate current 2A
  • Peak gate voltage 5V

Are those the values to look for ?

[edit]Changed the initial schematic to correct the issues.[/edit]

OK, so 1W resistance should handle it, 2W should be ideal.
The gate is connected to pin 4 of the MOC, so R20 should be 1W/2W.. how about R19 ? That one is connected to T2.

We're talking about 360R R19.. but when you talked about gate..

Correct me if I'm wrong, from the MOC POV:

  • PIN 6 connects to BT139 PIN 2 (T2)
  • PIN 4 connects to BT139 PIN 3 (GATE)

You're right, the brunt resistor was the one directly connected to T1 thus having 220V passing trough it.
A five minute test with an 80W load turns out OK, more testing needed tough.

Another lesson learnt, ALWAYS trust the datasheet's examples. :slight_smile: