SCR Phase Control Dimmer Circuit

This SCR phase control works much like the common TRIAC dimmer, but has numerous advantages including increased current capability, robustness and absence of minimum voltage “snap-on.” A complementary, symmetrical trigger circuit consisting of two PUTs (programmable unijunction transistors) enables firing of two anti-parallel THYRISTORs (SCRs). The circuit makes up a two terminal power device that is simply inserted between the AC power source and load. Besides controlling the intensity of incandescent lighting, it is useful in controlling the speed of universal (commutator brush type) AC motors.

SCR Phase Control Dimmer Schematic

SCR Phase Control Dimmer Schematic Update

Bill of Materials
SCR Phase Control Dimmer BOM

Bill of Materials File

The ideal triggering device

The DIAC is a 28V bidirectional (bilateral) trigger device that is used on virtually all inexpensive phase controls. The trigger voltage is somewhat high for phase control of 115VAC. Years ago, there was a similar low voltage (6 to 8V) trigger device called a Shockley diode. Unfortunately, these never caught on and today are EXTINCT.

The 2N6027 Programmable Unijunction Transistor (PUT) can perform a similar function, but is polarity sensitive so it does not lend itself to TRIAC control. However, if two such PUT trigger circuits are employed for anti-parallel SCRs, some interesting things are apparent. Most important is the ability to control both trigger circuits with a single potentiometer so that both half-cycles are controlled identically. To obtain best balance, the zeners and capacitors must be matched. Both zeners and capacitors are specified for a tolerance of 5%, but I selected mine for better than 1% with my DMM. Purchase a few additional components so that you may obtain a good match.

PUT operation is simple. Its threshold voltage is programmable so that it can trigger at low voltages. In this circuit it is set via the 12V zeners. When the PUT anode voltage exceeds the gate voltage by one junction drop, the PUT fires and dumps the timing capacitor into the SCR gate circuit. It resets when the AC line voltage reverses.

I made a PUT relaxation oscillator flasher circuit that you may wish to experiment with:


The typical DIAC controlled TRIAC dimmer tends to “snap-on” at the minimum voltage when the adjustment is slowly increased. After snap-on, the voltage may be reduced if desired. The “snap-on” phenomenon is caused by the lagging phase relationship of the AC voltage signal that appears across the timing capacitor before the DIAC threshold is reached. After it initially fires, the timing starts at line voltage zero crossing.

The timing circuit in this control completely avoids this problem by resetting the capacitor voltage each half-cycle via reverse diodes across the two timing capacitors.

Advantages of SCRs over TRIACs

Thermally, there are numerous advantages. Splitting the output current in half reduces the device current significantly and provides a much lower thermal resistance to the ambient. Also, SCRs are rated for a maximum Tj of 125°, while TRIACs may be limited to only 110°C. Then, of course, SCRs are available in current ratings extending to hundreds of amps — well over ten times that of the largest TRIACs. SCRs are simply more rugged.

Its only disadvantage is obvious: it simply takes more hardware and circuitry.

The snubber

R1 & C3 form a snubber that is connected across the power devices. It performs two important functions as follows:

First it provides a circuit to absorb reactive energy when either SCR recovers (stops conducting current). As you may know, a rectifier conducts current in the reverse direction for a short period of time (e.g. 5µS) when it becomes reversed biased. The flow of this current stops abruptly and any series circuit inductance then causes the generation of a transient voltage (spike). The snubber is a place for this current to go so it cannot develop a high voltage.

Second, it provides a resistive load for high frequency line noise and transients. Such noise or transients can potentially cause an SCR to turn on for a half-cycle. One common source of transients is simply the power switch when it closes.

Film resistors are not robust in handling voltage transients so I specified a ceramic composition resistor. Polyester capacitors have poor reliability at 230VAC, so I specified a polypropylene capacitor that is AC rated. In my circuit, I used a Quencharc device that contains both resistor and capacitor — it is AC rated for 125VAC.


The load voltage is easy to see here as it increases in phase. The base line is rather indistinct due to the low power load used (7.5W lamp) — the snubber reactance causes significant voltage drop across the lamp.

The gate current indicates a peak current of about 200mA and a time constant of about 32µS — essentially 47Ω * 0.68µf. The leading edge spike is caused by the inductance of the 1Ω shunt resistor used. This can easily drive a much larger SCR — the first one I made used a 90A thyristor doubler. The minimum gate current pulse width rule of thumb is 5µS in which the gate current exceeds Igt (gate current threshold) — this allows sufficient time for the load current to ramp up to the SCR holding current.

The gate voltage has a different appearance than the output of the programmable unijunction for two reasons: First, the gate is a non-linear resistance that tends to act like 2 or 3 series diodes. Secondly, when the SCR is conducting, junction voltage appears on this terminal as well.

SCR Phase Control Project Photos

Note the absence of the isolation transformer — it is located under the bench.

For the future

AC Switch Circuit — turn this into an AC switch via a photomod


Join the conversation!

Error! Please fill all fields.
  • ajithsmngmail-com

    helo, can i use this circuit for high power scr switching. eg- for ac motor speed control

    • Jim Keith

      This control can possibly be used for motor speed control. It would be best running a universal type motor (brush type) because such does not have a ‘negative’ slope speed /voltage response. This type of motor is often used in AC operated electric drills and vacuum cleaner motors. It is also called a series-connected motor (field winding connected in series with the armature).

      Controlling an induction motor is generally disappointing as the speed tends to be unstable, and the active voltage control range is perhaps only 10% of the total phase control range. In other words, the speed of an induction motor tends to be unstable (either increasing or decreasing speed) at a given applied voltage. A shaded pole induction motor may be somewhat better due to is lossy nature, but such is generally a low torque machine.

      One interesting addition is tachometer feedback (pulse tachometer or DC tachometer). This at least would provide a linear speed response. However, the available torque is low at low speeds. A tachometer feedback application would make a really great project.

      One interesting possibility is the permanent split capacitor motor that is often used for air conditioner blowers. I have never played with one of these using a phase control.

      Avoid motors that have starting windings as in capacitor start motors etc. These can actually start in the reverse direction due to the capacitor phase shift in conduction with the phase control phase delay.

  • sam-rotaconiburst-co-za

    Would ith dimmer work on 400 volt if i change the Cap’s to 400 volt + and the Zener diodes to 24 volt ± or must i change the resistors too

  • Jim Keith

    Place an X rated capacitor across the mains and add a 50 to 100uH inductor in series with the load. Perhaps add a soft ferrite like the ones used on old monitor video cables.

  • Msolar

    Hi Jim,

    I made your circuit and it is working but it produces a lot of RFI. Only thing that I changed was the SCR(I use BT151). I use this circuit for charging capacitor to desired level. Do you have any suggestion about this problem?

  • ranjith

    Hi Jim,
    Thank u vy much for your quick response and the advices. Last time I have forgot to tell you the SCR I have tried was a big one not the same as yours SKKT42/12 it has vy low internal resistance between gate and the cathode so I think the drives must be not enough for the gats (I am not sure) what do u think about this situation and what do u suggest, please help me to overcome this problem. Tks.

    • Jim Keith

      Reduce R5 & R6 to 10Ω or so. The Igt of your SCR is about 150mA vs about 25mA for the smaller discrete devices.

  • ranjith

    Hi jim
    Just now I tried your circuit but unfortunately it doesn’t work, I made a pcb for that and checked several times nothing wrong with the components and the PCB circuit, only thing happened 250K v/control and the 100 ohms resistor burnt several times.finaly I removed .68 Caps but the result is same.please tell me what could be the fault. Tks.

    • Jim Keith

      I just retrieved my breadboard from my junk box and retested it to prove again to myself that it really works. The SCR and 2N6027 PUT pinouts are also indicated correctly on the schematic.

      If something is amiss in your circuit and the PUTs and/or the SCRs do not fire properly, line voltage remains across the pot & 100Ω resistor regardless of resistance setting –so these components will indeed burn up as you say. However, in the process it will likely destroy the PUTs and/or the SCR gate circuit. Note that you have only one chance of getting it right, so if it does not function initially, it will destroy itself. In other words, the voltage across the anti-parallel SCRs collapses as they turn on thus greatly reducing the voltage across the pot.

      My recommendation is that you replace the PUTs and SCRs, check the circuit and start again. Only this time do not advance the pot beyond 50% rotation until you observe that turning it slightly changes the light bulb intensity (use an incandescent light bulb for testing). Use a scope to prove it is working on both half cycles –this could also be proven via the use of two light bulbs with series rectifiers –both should have the same intensity. Also, measure resistor values –if R5 & R6 have changed value significantly, it may indicate that the SCR gates have been damaged.

      Also, check out my Tips on Troubleshooting:
      The big issue that stumps experimenters is a “false assumption.”

      Good luck!

  • prashant

    please send me a report of this project.

  • Ranjith

    Hi my friend the problem solved. Thank you for the quick response and wish you all the best.

  • ranjith

    This scr control circuit seems to be very good but the original circuit and the updated one is very similar,I can’t understand which one is correc. What’s wrong with the diode

    • Jim Keith

      Actually, both circuits are now the same–the one in the comments came first–later, the top schematic was updated. Originally, one of the zeners was connected reverse.

  • Ken Yanczer

    Hello, Nice circuit. Is there a way to add an adjustment to cause the lamp filaments to preheat at a very low level when the main pot is at zero? They would “idle” and perform faster than if they were cold.
    Ken Yanczer

    • Jim Keith

      No easy way to do this, but this is what I would try:
      First connect one filament to a DC supply so that you can measure the voltage and current required to minimally heat it.
      For the filament that is tied to the return line, connect a current limiting capacitor between one side of the filament and the supply line –calculate the reactance per the line voltage and required current.
      For the filament tied to the SCR switches, connect a low voltage transformer secondary with a limiting capacitor in the transformer primary–this will also require some experimentation.
      Then if you want to get really fancy, interrupt both circuits with a relay that picks up when the bulbs light –this could be sensed via an LDR or phototransistor.
      Good luck and have some spare bulbs available –then tell us how you made out…

Looking for the latest from TI?