Turn signal controller

eyhonda

Coast to Coast
I just built a simple turn signal controller. Easy to build inside an hour. It is made up of a 555 timer (8 pin dual inline IC), two resistors, a cap and a small relay. I doubled this for one left and one right. It fits in a small box of 2" x 2.5" x 1.5" high. The timer is in astable mode. Basically, when I power it, it puts out a self-triggered square wave. A 10k and 20k resistor along with a 10uf cap produces about a 1 hertz signal. It can be changed with different resistor values. Now, I can use LED turn signals without load resistors. LED turn signals with load resistors defeat the advantage of low current draw.

turn signal controller.jpg


The 555 timer datasheet is here:
https://www.national.com/ds/LM/LM555.pdf
 
Wouldn't it be a lot easier and smaller to just buy an electronic flasher relay for 12 bucks?

Not to down you, I think it's awesome you did it yourself just seems like a whole lot of work for a generic 12 dollar part at auto zone. Or does it do something cool?
 
No, those relays still need a current draw to switch them. They won't switch LED's. Unless, there is a relay out there that is load-independent.

Building one is easy (and fun) for me since it is my day job (electrical engineer). Most of the parts are just extras lying around my office. Free parts, built in an hour. Also, I built an LED brake light. That was a little more involved (60 LEDs and 40 resistors). I like building small electronic parts anyway.

It's sort of the builder who is a machinist. Why would he buy clip-ons or brackets off the shelf if he can mill them himself?
 
They do have no load flashers. They usually say something about LEDs on them. I have one on my bike I bought from Advance Auto for about $12. You can also get them from the places that sell LED signals like custom dynamics and superbrightleds.com.

But I agree. Making your own is very nerdy and cool.

I wish people on this board would stop telling people they need resistors to get LED turn signals to work. It is just false.
 
For what it's worth, you can get away with only one flasher if you locate it prior to the indicator switch. Gray wire is the +12V to the switch, I think.
 
Thanks you two for thinking it was cool to build it. I'll post about the LED brake light and electronic ignition I made soon after I finish it.
 
As always, I recommend the Tridon EP34. Works amazingly in solely LED equipped systems.

That being said, I dig your approach to it! Always nice to make something with your own two hands!
 
Sonreir said:
For what it's worth, you can get away with only one flasher if you locate it prior to the indicator switch. Gray wire is the +12V to the switch, I think.

I thought that thru. That strategy won't work. The circuit must be either always on or seperately switched. In the stock scenario, the switched load of the bulb triggers the flashing relay. The turn signal switch is just a circuit path diverter. In my scenario, the only way to turn on the flasher was to put power to it. So, I have to put two circuits after the turn signal switch. It, in turn, becomes the power switcher to either the left or the right circuit. The common pin of the switch is always powered.
 
eyhonda said:
I thought that thru. That strategy won't work. The circuit must be either always on or seperately switched. In the stock scenario, the switched load of the bulb triggers the flashing relay. The turn signal switch is just a circuit path diverter. In my scenario, the only way to turn on the flasher was to put power to it. So, I have to put two circuits after the turn signal switch. It, in turn, becomes the power switcher to either the left or the right circuit. The common pin of the switch is always powered.

Would it not be possible to wire the 555 to +12V constant and then use the output to trigger a relay? The relay then controls the switching to the signal switch. It's always oscillating, but that shouldn't matter until you signal left or right, correct? Or is there problems with having the 555 always on?
 
Oh, yeah. Good idea. But the 555 timer can only source or sink 200ma. Hence the need for a relay. In your scenario, I would have only one 555 timer but I would still need two other circuits (mini relays or a PNP transistor) to handle the LED current after the switch. I'm not sure what the current is on a turn signal. But say it has 10 LEDs at 20ma each, that would bump the limit. Unless, you were thinking to put the single relay ahead of the switch. In that case, the relay would always be switching (although with no load). Datasheet says life span is about 5 million low load. I suppose it would be OK. I'll think about it for my next one. I need one for the Monster. Is it better to save half the current circuit and keep the relay running or just double it the way it is now?? In an OEM situation, it would likely be engineered as a double circuit (or use a micro). Thanks, Sonreir. You always keep me thinking. BTW, what happened to our embedded ignition idea?
 
Booligan said:
As always, I recommend the Tridon EP34. Works amazingly in solely LED equipped systems.

That being said, I dig your approach to it! Always nice to make something with your own two hands!

Thanks! I have that relay on my Monster to flash the brakes (also built LED brake light). At that time, I didn't think about this flasher. Otherwise, I would have built it then .
 
I scrapped my original software design and I'm in the process of reworking it.

I still can't seem to get my test case for the brake light flasher working though (the one I sent to you). Admittedly, I've not put very much time into it, but it sure is frustrating to have it working perfectly in my IDE but then not do a damned thing in the real world.

The ignition, my original design used a single hall sensor and then calculated the RPMs and timing based on the frequency. I had to use a 20Mhz microcontroller to get enough resolution on the resulting delay. My thought of using 8Mhz was putting the timing off as much as 10° after all the calculations were done and so I've had to up it. Even with that change it was proving to be a bit tricky and so I'm working on a two-sensor design that will only time the difference between TDC and 5°. This seems a bit counter intuitive, but by reducing the time between the reset of my timer, my calculations can all be done in smaller data widths and it's saving me a bunch of cycles that were previous spent on processing the calculations.
 
We need the same micro in order to coordinate programming efforts. You can migrate to a pic16f73 or I migrate to your pic16f84a. I have the pic16f73 micro (bin full of them). I need one of your micros or I have to modify the code to work with my micro. Do you have an extra one to send me? PM me for my address.

Also, the LED's you sent me worked awesome! Although, they had a big volt drop (~ 3+ volts) and I can only string 3 in series. I had 20 strings of 3 LED's! I have extra and will likely build some turn signals too. I'll put them in a some BMW side lenses I have as spares.
 
Yeah... I think they're rated at about 3.3V. I was going to try to build a headlight out of a couple hundred of them, all wired in parallel. Maybe I still will. :)
 
Back
Top Bottom