# 12 Volts or Nothing

Hello so I am hooking up a LED strip to my car but I am having trouble with the dimmer on the dashboard. To make this very simple and easy what I am trying to achieve is to cut off the power from the LED circuit if the voltage falls under 11V. I was thinking of using a relay and have the power hooked up to the NO and positive side of the relay, therefor if the voltage falls under 11V, the relay would go to the "closed" position and the circuit would get no power, however there are no 11V relay. And I say 11V because I am getting 11.6V, not 12V.

The dimmer varies the voltage between 3-11.6V, and when the voltage falls under 11V, weird stuff happens with the IR controller box, I don't want to explain what happens cause it gets complicated.

So that's my problem/question. I am not easily able to get the a constant 12V supply, I can only tap into the 3-11.6V wire. So using that 1 wire, what simple circuit can I make?

• Relays don't turn on exactly at their specified voltage - a 12 volt relay will probably operate when its coil voltage rises above ~ 9 volts or so, and not release until the voltage drops below ~ 6 volts or so. You will need a voltage comparator to make the relay operate at a closely controlled voltage. Commented Mar 20, 2015 at 22:52
• What car is it that works on 11.6V and not 12-14V? Commented Mar 20, 2015 at 22:52
• The easiest solution is do a chain of LEDs in series that has a combined voltage drop 11 volts. I would still use resistor to limit current. The only problem is that the light intensity will drop down with voltage. Commented Mar 21, 2015 at 3:06

## 2 Answers

I think the reason that your LED strip controller box gets messed up when you dim the lights is that most modern automotive dashboard dimmer circuits use PWM to control the brightness. This PWM will royally mess up the LED controller.

I do have a suggestion for you to try.

Most of the automotive dashboard dimmers that I've seen have high-side switches for the dash lamps. Most, but certainly not all.

Most inexpensive LED controllers have low-side switches. That is: all of the LEDs are wired up as common-anode and the FETs inside the controller pull the cathode line of each LED string to ground to turn it ON.

IF your dashboard lamp dimmer does have a high-side switch, you can try the following: cut the common (+) line that goes from the LED controller to the LED strips. Feed 12V power directly to the LED controller. Connect the (+) line that goes to your LED strips to the dashboard dimmer output.

This should allow the LED controller to work correctly and the LED strips will dim just like the dashboard lights.

Do note that you may get beating effects if the LED strips are chasing at a fast rate or are dimming. The only way to know for sure is to try it.

How about this?

simulate this circuit – Schematic created using CircuitLab

Zener diode D1 is a fixed reference that is compared to a specific fraction of the supply that is set by VR1. As the supply changes, so does the voltage from VR1, but D1 doesn't change. Thus, you can sense your own supply.

R2 sets the current through the zener to roughly what the datasheet says it was tested with. Do some math to figure out the correct value.

R3 interacts with VR1 to create some hysteresis so it doesn't oscillate. If the output is high, then the threshold effectively goes down a bit; if the output is low, then the threshold effectively goes back up. Increase for more of a hair-trigger; decrease to make it more "latchy". Maybe even make it variable also.

Don't drive the relay directly from this circuit! Use a transistor and a flyback diode as described in the answers to many other questions on this site. If it's backwards, don't swap the + and - inputs of the comparator as can be done with other circuits. This one needs them like this for the hysteresis to work. Instead, put an inverter (logic, not power) between the comparator and the relay transistor.

• Rather then dragging an inverter in, do flip the + and - inputs, add some series resistance between the zener and the input, and keep the positive feedback. You can design it either way. Commented Mar 21, 2015 at 0:17
• Yeah, that'd work but it adds to the complexity of what I would consider to be a single block. The inverter seems separate to me and can be made with one transistor and one resistor if it must be minimal. Either your or my way would technically work though. Commented Mar 21, 2015 at 0:26