I bought my oldest son a used 2007 Mazda 3 last week. It is in great condition and he loves it. The problem is that since it is an older base model it doesn't have any extra bells or whistles like automatic headlights. He was driving a Toyota Corolla that he passed down to his little brother but it had automatic headlights so he never had to worry about turning them off. I was worried he would forget to turn them off one day off at college and come back to a dead battery.
I racked my brain trying to come up with a simple solution that perhaps involved tapping a fuse that provided power only when the ignition was on and somehow using that to supply power to the headlights but the problem is that there are five different fuses for the headlights: Right-side low beam, left-side low beam, right-side high beam, left-side high beam and one 40 amp fuse under the hood that runs them all. Even if I were to somehow interrupt the power to the headlights through one or all of the fuses, the car still thinks the headlights are on because the switch is still in the on position and the car makes a loud warning sound while the drivers side door is open with the ignition off. That would become annoying.
I decided the best way was to go to the source and that was the headlight switch on the turn signal stalk. Luckily exposing that only involved loosening four screws. Two under the steering wheel shroud and two that hold the stalk to the steering wheel column.
I probed the stalk assembly using my multimeter and determined that interrupting the peach wire was the same as turning the headlight switch to the off position. How I determined this was because there was continuity between the peach and red wires when the switch was turned to headlights and continuity between the peach and orange wires when the switch was turned to parking lights but no continuity between any of them when the switch was turned off.
I decided to use a 12 volt automobile relay to interrupt the peach wire when the ignition was off to simulate the headlight switch being in the off position. I pushed the peach wire's pin back out of the wire connector and luckily it was almost identical to an RC servo pin of which I have plenty. I used some thin RC servo wire and peeled off the yellow wire so I was left with only a red and back wire. I soldered and crimped a female servo pin onto the black wire and pushed it into the wire harness and soldered and crimped a male servo pin onto the red wire and pushed that into the peach wire's female pin. Basically what I did was make the peach wire super long because shorting the ends of the red and black wires together completes the peach wire's original path.
This is the stalk assembly with the protective cover reinstalled. In the second picture you can see my red and black servo wire disappearing into the dash. I fished it over to the fuse panel located in the passenger seat footwell.
Here is the 12 volt automobile relay mounted behind the fuse tray. The white wire is ground or negative and the red wire goes to a positive terminal that is energized only when the ignition is on. When that happens it closes the relay and provides continuity between the yellow and blue wires that I have connected to the ends of my red and black servo wire using a JST connector. Remember, shorting the red and black wires together completes the peach wire's circuit.
I used a mini fuse tap to connect the positive terminal of the relay to a fuse position that is energized only when the car's ignition is turned on. In this case it was to a moonroof which this car definitely does not have.
Now when the car ignition is turned off the relay opens and interrupts the peach wire and the headlights turn off regardless of the switch position on the stalk. If the switch was left on, then when he starts the car the headlights come on automatically which my son said he might do since his driver's ed instructor told him he should drive with his lights on all the time anyway. At least now I won't have to worry about him running down his battery! Thank you for looking!