I'm driving a high power load with the BTA440Z (datasheet):
line voltage is Vac = 230 AC, load power is P = 1.33 kW, thus load current is ~ I = 5.8 A.
If i swap the load with another that draws P = 2.33 kW, then I becomes ~ 10 A; in this condition the triac overheats and it doesn't un-latch due to thermal runaway.

To properly size the heatsink, I need to know how much power the triac is dissipating, and I realized that in the datasheet there is no such thing as an equivalent to what in a MOSFET would be "Rds on", and that I don't actually know how to find such a parameter.

The circuit is as follows, taken from this datasheet, without the snubber network: enter image description here

Initially R1 was 680 ohm, because i thought that it could work anyway and less current would pass through the opto; but i subsequently dropped it to 360 ohm because i figured that more current through the gate would make for a wider channel between the main terminals: is this assumption correct? Does this current also contribute to the power dissipated on the triac?

Based on what i read, what i know, (electronics-tutorials.ws, link 1 and link 2, and the Wikipedia page about Triacs), and the following image, my guess is this:
for the most part, the power dissipated by the triac is given by the current to the load I, times the triac on-state voltage VT.
Is this correct?

enter image description here
Please note that without a zero-cross detector, my microcontroller won't turn off the opto after the triac has been fired into conduction, so assume the opto is always on.

  • \$\begingroup\$ @G36 At max 8 K/W! \$\endgroup\$
    – winny
    Jul 10, 2019 at 12:50
  • 1
    \$\begingroup\$ Look at fig 10 on the datasheet. "On-state current as a function of on-state voltage" and P = I_load x VT = 10A * 1.5V = 15W So you need a heatsink Rth < (150 - 30)/15 = 8 K/W \$\endgroup\$
    – G36
    Jul 10, 2019 at 14:06

1 Answer 1


The resistor value does not noticeably affect the power dissipation. It only conducts briefly at the beginning of each cycle until the current is sufficient to trigger the gate. Using an excessively high value causes RFI from the switching too far from the zero crossing and can contribute to having some DC content in the output (important if the load has a transformer). The latter effect is due to asymmetrical triac trigger current in quadrant I vs. III.

Usually you can directly read the typical power dissipation from the triac datasheet. In your case, the datasheet you linked above contains the following graph:

enter image description here

Your conduction angle is 180° in this case. You should aim to keep the triac junction well below the absolute maximum value at maximum ambient temperature for decent reliability. Personally, I don't like to see the junction temperature much exceed about 100°C under any conditions. If it's losing gate control it's way, way above that (and any triacs exposed to that abuse should be replaced with new parts).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.