I've been looking at creating a simple circuit to slowly turn a halogen daylight bulb on over about 30 minutes as soon as power is supplied to simulate dawn.

I've got a 240VAC 42W halogen bulb that will be used in conjunction with a 7 day timer socket. I've been looking at various triac and triac opto driver (MOC3020) circuits to try and find one that will do this but they all want to be really fancy and use microcontrollers or other such things to do far more than I need. I just want a simple circuit that will brighten the bulb over a set time (30m) when it is turned on by the timer socket.

Help with values and component ratings would also be greatly appreciated.

Thanks for the responses. Further to this; size and heat dissapation are factors for this circuit as it will need to fit within some sort of lamp holder, also when I have looked at the triac solutions they all talk about chopping the ac waveform and doing from the zero crossing point to provide a smoother operation which made me think of the following:

Is it feasible to chop the ac waveform with a pwm at a higher frequency (5kHz)? Would this make it buzzier of lessen the buzzing or is it just totally unnecessary?

  • 4
    \$\begingroup\$ This sounds like something that might be doable with a NTC thermistor, even... \$\endgroup\$ Jan 19, 2015 at 23:32
  • \$\begingroup\$ Keep in mind that, in a true sunrise, the brightness changes in a non-linear fashion, more like an exponential curve. I have done a similar thing in a grow light for a house plant, but it is a complicated thing with microcontrollers, PWM and LEDs (see this if interested). \$\endgroup\$
    – anrieff
    Jan 20, 2015 at 17:09
  • \$\begingroup\$ I would consider starting with a commercial lamp dimmer. \$\endgroup\$
    – Hot Licks
    Jan 20, 2015 at 18:55
  • \$\begingroup\$ Thanks, I have indeed obtained a lamp dimmer and taken it apart. The circuit is nice and simple so hopefully I can work the thermistor into it and then play around with the placement of it to get 30 mins. \$\endgroup\$
    – James
    Jan 24, 2015 at 22:33
  • \$\begingroup\$ I wish people would quit saying crap like "I don't want a microcontroller, I just want a nice and simple circuit." THE MICROCONTROLLER IS USUALLY THE EASY WAY! \$\endgroup\$
    – Daniel
    Nov 12, 2015 at 3:53

2 Answers 2


The basic circuit is pretty simple. C1 charges through R1 and turns on the triac when the threshold voltage of the diac is reached.


simulate this circuit – Schematic created using CircuitLab

Now you've got the choice over how you vary that resistance with time. ThreePhaseEel mentioned the excellent idea of using a Negative Temperature Coefficient resistor (aka the common thermistor), tied somewhere near the lamp. Note though, that the lamp will not light up linearly with time as NTCs' resistanceVStemp is a decreasing exponential. This might be a problem.

The first thing that pops into my mind would be a 555 timer in astable configuration supplying a clock to a 555 in PWM configuration, with the modulation input being the charge of a capacitor (either through constant current or within one time constant to make it linear). The output of the PWM 555 could be either optocoupled to a MOSFET or directly tied to a solid state relay. However this sounds like overkill. If you prefer simplicity over cost effectiveness, you could replace the 555 with a microcontroller, which gets everything down to wall DC adapter-uC-optocoupler-gate resistor-MOSFET. Still, sounds like overkill.

I personally like ThreePhaseEel's solution, and I'd like to hear that someone else has an idea as simple that makes the lamp light up linearly.


I am artist If you aren't space constrained, grab a counter like this. Hook each bit to control some isolated switch, relay or otherwise. Run each switch over a high power resistor. Tweak resistors and clock speed according to preferences. You now have a digital dimmer.

  • \$\begingroup\$ A lot of heat and big components, especially if you meant mechanical relays. \$\endgroup\$
    – WalyKu
    Jan 20, 2015 at 16:57
  • \$\begingroup\$ I think in theory it should be no more heat than the lamp would generate itself. Correct me if I'm wrong but by adding a resistance in series with the lamp is just going to reduce the current the system consumes and actually reduce the total I^2R heat losses. This is afterall what an analog resistive dimmer light switch does using a potentiometer. \$\endgroup\$
    – lm317
    Jan 20, 2015 at 19:50
  • \$\begingroup\$ But the lamp is rated at 42W. Those are expensive resistors. \$\endgroup\$ Jan 20, 2015 at 21:15
  • \$\begingroup\$ Yeah they wouldn't be your normal thru holes that's for sure. One could in theory break open a bunch of old style dimmer switches and steal those resistors. I did mention that this solution requires space and high power resistors. Call it a junk yard worst case solution but it's guaranteed to work. OP didn't mention their application or such requirements. It certainly wouldn't fit into a wall but if it's for an art installation it might work. \$\endgroup\$
    – lm317
    Jan 21, 2015 at 23:36
  • \$\begingroup\$ This is not a good approach. It's wasteful and expensive! \$\endgroup\$
    – Daniel
    Nov 12, 2015 at 3:59

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.