I have a requirement to drive a triac from the MCU powered by a DC/DC converter. The AC supply is typ 24 VAC @ Rc terminal with respect to C terminal. The triac is driven as follows. The MCU GND is always regulated at Rc - 3V3. The MCU output which drives the gate preferably will be connected straight, without the BJT. The gate current is II/III quadrants is typ 5 mA.

enter image description here

How can I configure the DC/DC converter and its input?

The problem is a real world design issue I am trying to solve. I can use an opto coupled triac driver to solve this. But this is a cost sensitive application, so I am trying to see if I can find a feasible solution without any isolation.

The triac drives a 24V A/C motor. When I tried the triac drive as MT1 connected to GND and MT2 to the load, I loose a feedback signal generated from the motor. I have no control over this, as this is inherent to the motor controller. So I have to wire-up the triac as MT1 to Rc (Live) and MT2 to the load. So instead of switching the 'C' (or neutral), I have to switch the 'Rc' (or Live). This is the constraint.

To further to reduce BoM cost, I am trying not to use any BJTs or FETs to drive the triac gate. To do this, and also for better switching, I want to operate the triac in II and III quadrants, where gate current is negative. I plan to use Z0103 as the triac and in II/III quadrants, I could trigger it @ -4mA (considering temperature range). My MCU (STM32F103) can sink/source 8mA from GPIO so this should be viable. This is my idea for the moment, I could be wrong.

This is where my initial design came from. I think this way might also be a viable solution. enter image description here

We can configure a standard buck as a negative output regulator, so this way the MCU GND will follow the Rc below 3V3.

The DC/DC Buck could be anything which can accept up to 60VDC and source at least 1A.

  • \$\begingroup\$ Is this homework? Is there a DC/DC converter you are required to use? Can you provide any manufacturer part numbers possible please? \$\endgroup\$
    – Electrocol
    Commented Aug 2, 2023 at 7:06
  • \$\begingroup\$ Sounds like an XY problem. Do you need to time something to the AC cycle in your MCU? \$\endgroup\$
    – winny
    Commented Aug 2, 2023 at 8:21
  • \$\begingroup\$ If your MCU needs less than 20 mA you can use a negative linear regulator like LT3015EMSE-3.3. A better solution is an isolated DC/DC converter where you can connect +VIN and +VOUT (RC in the circuit). \$\endgroup\$
    – Jens
    Commented Aug 2, 2023 at 21:46

1 Answer 1


Yes, this is an acceptable method.

Note that:

  • You generally want to drive with as much current as possible (given device ratings, thermal limits, driver capacity, etc.), to ensure rapid and complete turn-on.
  • Gate is the base of some back-to-back transistors that make up the thyristor (effectively); as such, it has voltage drop (a VBE) while conducting, in the direction of load current. So, I would put a diode or two in there, to introduce a little voltage drop, ensuring the MCU isn't reverse-biased while the load is reversed.
  • Gate drive doesn't need to be maintained (continuous) while an on-state is desired, but it should be pulsed to ensure the TRIAC doesn't turn off due to unexpected load current phase (which, a motor will be notably inductive, with phase depending on torque). A modest duty cycle (say, 10-20µs on-time, 10% duty) might help save 3V3 current consumption, say.
  • \$\begingroup\$ Thanks Tim, I have prototyped it and seem to work. I have not yet tried full temp range yet, but I think that should be ok. Sorry for my delay in replying. \$\endgroup\$
    – Kaush
    Commented Aug 27, 2023 at 2:55

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.