I'm familiar with simple forward phase dimming using a triac. The circuit is stupid easy with only a handful of components. enter image description here

Of course this simple example can be improved upon to make it safer or improve performance and efficiency. But the basic principles remain the same.

What I'm having trouble finding is useful information on reverse phase dimmers. Everything I can find describes the difference between forward phase and reverse phase in principle and how they each affect different kinds of loads, but I'm unable to find any technical information on how to actually accomplish reverse phase dimming.

How do you make a reverse phase dimmer and what would a simple example circuit look like?

Note: I'm guessing that reverse phase dimming has to be digitally controlled. I don't need any code for an MCU, but I would like to know what kind of signal goes into an MCU that it can recognize when the phase begins and what kind of signal it sends out to cut the phase at the right time (presumably using a MOSFET?). If there's a simple way to make a reverse phase dimmer without digital control, that could be useful too.

Some additional information that might be useful. I'm researching this because I want to build a home-made wifi controlled light switch dimmer. I've got a 3.3v power supply and a wifi MCU with a dozen GPIO pins and a handful of PWM pins (a PADI IoT stamp based on the Realtek RTL8710AF).


2 Answers 2


enter image description here

A reverse phase dimmer does not have a triac but an IGBT or MOSFET. Such a dimmer contains also the electronics needed to modulate this device. These electronics can be simple or very complex.

The principle is simple. The mains AC is fed into a bridge rectifier with in the centre the IGBT or MOSFET. The IGBT or MOSFET can be switched on and off at any moment. The AC can be triggered on the leading edge, The trailing edge or somewhere in between and for the duration needed.

This makes such a dimmer suitable for: Halogen lamps Standard incandescent lamps Electronic mains LED lamps of different kind. etc.

If the dimmer needs with different kind of lamps then the control system has several settings to make that possible. In case of remote control the dimmer can become expensive.

Using the principle with the bridge rectifier and the IGBT or MOSFET it would be possible to make yourself but I did not work that out. Maybe the schematic included helps. The dimmers are very sensitive to overload but since only the IGBT or MOSFET is destroyed also simple to repair.

Remark: The previous picture had some wrong earth connections so I have changed it. The incomming PWM circuit is isolated from the mains side. Thanks to Elektor.

  • 1
    \$\begingroup\$ I think you need to remove the earth symbol on the right (bridge rectifier short-circuit) and remove the link between R3 and the earth on the left (you don't need it when you have the opto-isolator). See here for explanation of the rectifier problem. \$\endgroup\$
    – Transistor
    Oct 10, 2016 at 20:39
  • \$\begingroup\$ I've added a paragraph of some additional insights into my project. Based on your answer, I think the place where I need to focus my efforts is on the zero-cross detection. I believe that I can connect a PWM output on the PADI directly to the gate on a MOSFET and call the switch good. But making a zero-crossing circuit and connecting that to the PADI is where things get fuzzy. I found an article online that uses a 741 op amp for zero-cross detection. Maybe the PADI is sophisticated enough to take an unrectified v-in and I can do detection in firmware? At least this gives me a start. \$\endgroup\$
    – Thomas
    Oct 10, 2016 at 22:38
  • \$\begingroup\$ @Transistor: The drawing was not mine. I have changed it. \$\endgroup\$
    – Decapod
    Oct 11, 2016 at 1:37
  • \$\begingroup\$ @Decapod: That's better and safer. You should credit Elektor magazine with that schematic. It is immediately recognisable as their house style. \$\endgroup\$
    – Transistor
    Oct 11, 2016 at 1:40
  • 1
    \$\begingroup\$ @Thomas: "I believe that I can connect a PWM output on the PADI directly to the gate on a MOSFET ..." Be very careful with your beliefs. Isolate if at all possible as shown in the Elektor schematic. It prevents you getting high voltages on your micro and makes debugging much simpler. You should ensure good separation between the control and mains PCBs and keep wiring segregated. Observe minimum creepage distance rules around the opto-isolator. \$\endgroup\$
    – Transistor
    Oct 11, 2016 at 1:43

There is nothing about a trailing-edge dimmer that requires digital control any more than a leading-edge dimmer. Here are a couple of practical circuit descriptions and circuit diagrams to eperiment with the scheme.



  • \$\begingroup\$ Spookey Circuits?! what happens when there is a fault in the load like a short which can occur when some filament lamps blow? \$\endgroup\$
    – Autistic
    Oct 11, 2016 at 4:19
  • \$\begingroup\$ @Autistic, The better stuff uses IGBTs with desat sensing to provide a very fast turn off on over current (Also, often a current sense resistor or CT). \$\endgroup\$
    – Dan Mills
    Oct 11, 2016 at 7:11

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.