Creating a circuit for turn signals on a vehicle

[iamsuperbleeder]

Hello everyone!

I'm just getting into circuit designing and PCB layout and construction. That said, I've been doing my homework, but still would like a little input on my project


I'm working on adding additional LEDs on the side mirrors of my truck, that will act as running lights and turn signals. Originally I was just going to use a very simple circuit with resisters and diodes and make the LED's light dimly with the running lights, and full brightness with the turn signals. However, I've been looking at another option which, if done right, will make them be lit at full brightness with the running lights on, and then when the turn signals flash when the running lights are on, the LEDs will flash in an alternating pattern that the turn signals on the truck do. When the running lights are NOT on, they will simply flash at full brightness along with the turn signals on the truck.

It's kind of hard to explain, so I made a simple animation that does exactly what I want to circuit to do:


**broken link removed**




Using a logic gate simulator, I was able to get this to work perfectly, using XOR gates. So upon research, I've found the 7486 IC is what I can use in my circuit for the XOR gates (only two gates are needed in my current design).

The logic part of the circuit and the 7486 IC isn't what's getting to me; it's the unput voltages that I'm concerned about.


Obviously a vehicle's voltage is suppose to be 12v, but while running, a vehicle can be at 14.5v normally, and the supply is very unstable as far as fluctuations go (just because of the nature of the alternators vehicles use to keep the battery charged).

I know I'm going to have to use some 5v regulators like some 7805's, as well as some caps to take the fluctuations out, before sending ANY power to the IC. What makes this a pain is that I'll have to have 4 separate 5v regulator circuits for just the inputs; 1) constant supply, 2) running lights, 3) left turn signal, 4) right turn signal. Is there an easier way to go about doing this aside from using 4 regulators and 8 caps? All of those components are going to take up quite a bit of space on a PCB, and I'm hoping to keep the package as small as possible so that I can wire it in and just velcro it up under the dash some where.

My second question would be, what would be the best way to, after I drop the voltage down to 5v for the circuit, bring the voltage back up to 12v for the two outputs? Just a couple of transistors to use as a switch, with a 12v feed in, the 5v output from the IC as the trigger, and the output as, well, the output? That's the only logical solution I can think of.



I wish it could only be as simple as being able to run the full 12v through the IC, but life's never that easy



I'm home-making this PCB, so it's just going to be a single layer PCB. I just placed an order with Jameco for a bunch of stuff to get me started (breadboard, organization box, grab bag of resistors and LEDs, a few 7486 IC's, a hand-full 7805's, and some other misc things), and I'm just picking up some standard pre-drilled PCB's and a project box from Radio Shack.


I'm using ExpressPCB to try and get a basic layout done right now, and can post a screen shot up of the layout once I think it's done. But I'm open to any suggestions or input as far as schematics or PCB layouts go. It is a pretty simple circuit, I'm just an electronics nuB for the most part as far as PCBs go anyhow...




Thanks in advance!

 

[iamsuperbleeder]

Here's the schematic that I had in mind.

Any input as to if this will work correctly? I'm pretty sure I covered all my loose ends, but... like I said, this is my first real circuit design



**broken link removed**

 

[colin55]

This is all you need:

 

[iamsuperbleeder]

but that wouldn't make them alternate when the running lights are on, would it? that looks like more or less just an OR switch, that will just light the LED when either input is energized

 

[colin55]

See the new circuit above

 

[iamsuperbleeder]

I'm still lost...

there's two inputs per side

so say we're working only one circuit

then we have two inputs; Running lights, and Turn Signal

basically I need the LED to be on if one or the other is on, BUT, if both inputs are on (or if both inputs are off), the LED should be off


I don't see how that circuit above will make the two inputs cancel each other out; it looks like all it does is make the LED light if just one or the other input is on, but with both of them on the LED would just stay on as well

 

[MikeMl]

Why don't you use the 18V version of CMOS logic? That way, you can do all of the logic without having to reduce the voltage, and then step it back up at the output.

Your transistor driver will dissipate a huge amount of power. The base is driven to only 5V, so the emitter will be at ~4.2V, so the Power will be (14-4.2)*(the LED current) = many W. You should use a grounded emitter transistor.

 

[crutschow]

With colin55's circuit, with both inputs high, the LED will be off since there is no return path for the current. Correct operation of his circuit assumes that there is a return path (apparently through the unlit filaments of any incandescent bulbs connected on either side) to ground when either side is not energized.

If you want to use your logic circuit, then to minimize the voltage problems use a CMOS gate such as a CD4030B which can operate up to 20V (you still should add a series resistor and zener or small capacitor to ground on all inputs to supress any high voltage transients above about 16V)

 

[iamsuperbleeder]

If you want to use your logic circuit, then to minimize the voltage problems use a CMOS gate such as a CD4030B which can operate up to 20V (you still should add a series resistor and zener or small capacitor to ground on all inputs to supress any high voltage transients above about 16V)

THAT is a perfect idea. That's the piece of information I was looking for; that CD4030B IC

Thank you so much!

I'm going to redesign my circuit real quick using that, and post it up. It should be a LOT more simple, compared to the PCB layout that I just created, lol


**broken link removed**




I'll redesign it with some small caps instead of zeners and resisters, and I'll post the new one up once I get it laid out with a 4030. I'd like to see if I get get the actual PCB laid out on my own

 

[colin55]

When the constant voltage is HIGH the LED flashes out of sync with the flasher. When the constant voltage is LOW the LED flashes in sync with the flasher.

 

[crutschow]

I'll redesign it with some small caps instead of zeners and resisters, and I'll post the new one up once I get it laid out with a 4030. I'd like to see if I get get the actual PCB laid out on my own

You want to use a resistor as well as a cap to minimize any transients. Try 10-50kΩ in series with 0.1µF to ground on each input. Use 100Ω and 10µF to decouple the power to the chip.

 

[iamsuperbleeder]

When the constant voltage is HIGH the LED flashes out of sync with the flasher. When the constant voltage is LOW the LED flashes in sync with the flasher.

so let me get this straight... 4 diodes, configured in that layout you posted above, will result in what I'm trying to do? I'm just having a hard time seeing that for some reason... I'm just going to have to construct the circuit and try it out myself

the LED assembly has a resister built into it already to be used on an automotive 12v circuit, so I wont have a need to add another resister

 

[colin55]

I can see that not only you, but other electronics persons, who have been in the business for 30 years, are also have trouble seeing how the circuit works.
That's why I say, the skill in electronics is to take a complex circuit and simplify it.
Your circuit was far too complex and the first circuit I designed in 30 seconds but did not satisfy all the requirements. The second circuit was designed as a result of the extra requirements you requested. Most of the time, if a circuit doesn't work, a few extra diodes can be added to get the gating needed. My circuit relies on the fact that the constant line has unlit globes to take the signal to ground and the flasher circuit has unlit gobles to take the signal to ground, so you have to put the circuit into the actual application for it to work.

 

[iamsuperbleeder]

Somehow, I think I know exactly what you're talking about now... assuming that by "globe", you are talking about a standard filament incandescent bulb; the type of bulbs that the original lights are on the vehicle.

I can try this out on a bread board, but I'll have to test it with some incandescent bulbs and not LED's, wired up how they are on a vehicle, which shouldn't be a problem since there's plenty of small 12v bulbs, like #194's, that I can use for the test.

Tell me if I'm getting this right now. Basically what's happening is, the filament in the incandescent bulb is acting as the ground wire for the LED when there is no + current applied to that bulb. So when a + feed is given to the LED and only ONE of the other inputs, it has a completed circuit threw the ground of the other bulb that is not lit. But, when BOTH sources are energized, they complete their own circuits, and take the ground away from the LED, thus causing it to not light.


Now, what if the original lights are LED's, and not incandescent bulbs? They are not at the moment, but I may change them in the future. If there WERE an LED bulb in place, could a simple resister at the original bulb location, between the + and - wires, make the circuit work again like if there was an incandescent bulb in the circuit, or would that cause my newly installed LED lights to be dimly lit, since there's a resistance in the ground source?



Thanks again for all you're help!

 

[colin55]

My circuit relies on the fact that unlit globes have a very small resistance (15%) and the voltage drop at 20mA will be very small. That's why this circuit works as an "add-on."

 

[crutschow]

Now, what if the original lights are LED's, and not incandescent bulbs? They are not at the moment, but I may change them in the future. If there WERE an LED bulb in place, could a simple resister at the original bulb location, between the + and - wires, make the circuit work again like if there was an incandescent bulb in the circuit, or would that cause my newly installed LED lights to be dimly lit, since there's a resistance in the ground source?

If you replace the bulbs with LEDs you could place a resistor (or a small incandescent bulb) in parallel with the LED to provide the additional load required for the circuit to work. Since it would be in parallel with the LED it would not affect its current or brightness.

 

[iamsuperbleeder]

I think I've got it now



**broken link removed**





and this would be the set if I replace the bulbs with LED's, correct?



**broken link removed**

 

[colin55]

If you replace the bulbs with LEDs you could place a resistor in parallel with the LED

It would have to be a low value.

It's better to use a very small globe with a 10R resistor to limit the turn-on current.

 

[iamsuperbleeder]

some body mentioned to me that I could just use a pair of bipolar transistors instead of 4 separate diodes?

does that sound right?



they also provided this:


**broken link removed**

 

[colin55]

I am bipolar. If the transistors are just as screwy as me, then you can use them.