I am building a large array of Light Switched relays, and due to the size of the array i want to keep the part count down.

Each 'Pixel module' has a 12V SPDT, bulb and an activate-on-dark circuit. There is a strong spotlight on the device and the viewers shadow is obscuring the light in their shadow so the swing is large.

I have seen the obvious LDR-and-Pot circuits with a junction out to a NPN Transistor, and the more stable comparator alternative

I was looking at "solar cell" types of Photodiodes (PiN Type), and I can't see why I cannot just lay one across a PNP transistor (Collector to Base?) or something so it applies a Base voltage when lit up.

Can someone briefly explain to me why I am wrong? I figure it would be all over the net if it was that easy, and i cant find any examples!

Is this possibly the cheapest way to make a light switched 12V relay driver?

Note: I prefer solutions using discrete circuitry, as I have a thing for human readability, visual components etc :P

  • \$\begingroup\$ Why not use phototransistors? \$\endgroup\$ May 5, 2014 at 3:01
  • \$\begingroup\$ I thought that was what a Photodiode was, but the "base" is the sensor... Will it carry 12V at 200ma for the coil, or do i cascade it to a NPN peraps? \$\endgroup\$ May 5, 2014 at 3:19

2 Answers 2



simulate this circuit – Schematic created using CircuitLab

If you used TO-220 MOSFETs to switch the relays, you could drive them with photodarlingtons such as the OP535A (to shunt away gate current).

You can experiment with the value of R1 to get the desired sensitivity. Try maybe 10K to start. At four components per relay driver, I don't think you'll get much simpler.


If you really want simple, small, and cheap, then the first thing to do is to ditch the large, klunky, and expensive relay. You have 12 V available, so find lights that work on that, then drive them directly with a transistor.

As for the light detector, I'd start by looking at CdS photoresistors. Two transistors should provide enough gain, and the gain can then be positive to allow for a little bit of positive feedback. This will provide hysteresis so that the light will snap on and off instead of fading on and off at the transistion. That will be useful to keep the switch transistor either full on or full off so that it will dissipate little power and you therefore won't need a heat sink.

Here is the topology I'm thinking of:

R1 and R2 form a voltage divider. As the ambient light gets dimmer, the resistance of R1 increases and the output of the voltage divider goes down. This turns on Q1, which turns on Q2, which turns on the light. R4 limits the current thru Q1 and the base of Q2 to a safe maximum, even if Q1 has infinite gain. R3 provide a little positive feedback as I mentioned above. When the light is switched on, the right wide of R3 goes low, turning Q1 and Q2 on a little bit more. The same works in reverse. When the ambient light goes higher, R1 goes lower, raising the output of the R1-R2 divider, which switches off Q1, which switches off Q2, which raises the right side of R3 to re-enforce the process.

I didn't put values on the parts because you haven't given us any indication how much current the lights require. However, R3 should be large relative to R1. You only want just enough positive feedback to get snap action, not to make a significantly different light to dark versus dark to light threshold.

Note that this same circuit can drive a 12 V relay instead of a 12 V lamp directly, although in that case you need to add a reverse diode accross the relay.


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.