Sequential 194 Reverse Polarity or 3157 Switchback Led Turn Signals

crutschow

Joined Mar 14, 2008
34,433
We don't have to program any type of pause and/or repeating cycle because it will be switched on and off with each flash, no?
No.
The cycle just has to be reset so that it always starts at the beginning of the cycle for each flash.

Here's a simulation of my preliminary circuit to show the sequence.
The top green trace is the simulated turn signal.
The red, blue, and violet traces are the LED sequence.
As you can see it sequences when the turn signal is high (12V on) and not when it is low.
Does that seem about like what you want?

upload_2017-7-8_22-48-8.png
 
Last edited:

Thread Starter

Ziggylittles

Joined Jul 2, 2017
36
Wow, what program are you using to create those? Those are fantastic diagrams.

However no, it appears that the above sequence is 1 on, 1 off, 2 on, 2 off, 3 on, 3 off.

My idea was that once a bulb was on it would stay on for the remainder of the sequence.

1 on, 2 on, 3 on.
 

crutschow

Joined Mar 14, 2008
34,433
It's a free analog simulator from Linear Technology called LTspice.

Okay, I'll look at keeping them on until the end of the sequence.

So the first bulb would go off and on with the turn signal, with the second and third in sequence.

Does this look like what you want?

upload_2017-7-9_8-42-16.png
 
Last edited:

crutschow

Joined Mar 14, 2008
34,433
Yes! That looks exactly right. I suppose at this point we need an exact figure for the length of the flash?
Approximately.
I'll add a trimpot to the circuit so the sequence timing can be adjusted to fit the flash time after it's built.

Below is the latest iteration which includes both the left and right sequence.
I believe the sequence is what you said you wanted for the 5 LEDs (L1 through L5).

The hazard sequence will also work as you wanted when both turn signals are flashing together.

upload_2017-7-9_12-37-59.png
 
Last edited:

Thread Starter

Ziggylittles

Joined Jul 2, 2017
36
Fantastic. Let's use the working figure of .24 seconds for the amber flash duration. Stopwatch trials resulted with a .24 reading a large majority of the time. Any other readings were mostly longer, I'd attribute to late reaction time.
 

crutschow

Joined Mar 14, 2008
34,433
Okay, I think we are getting closer.
Here's the latest simulation with an on-time of 0.24s and a period of 0.5s.
I also added MOSFET drivers for the LEDs (simulated with the added resistors that's inside the LED), so the LED currents (arbitrary value) are shown in the plot.

upload_2017-7-9_15-46-42.png
 

Thread Starter

Ziggylittles

Joined Jul 2, 2017
36
Perhaps I was reading the display wrong. That figure seems very wrong compared to other led bulbs with specifications available from the manufacture.
 

Thread Starter

Ziggylittles

Joined Jul 2, 2017
36
I did not examine for differences but it was certainly very bright. The prongs of the multimeter made full and clean contact on both ends.
 

Thread Starter

Ziggylittles

Joined Jul 2, 2017
36
I disconnected the circuit's wire near the bulb socket and put the meter inline there. I can check again in the morning this time putting the meter directly from battery to the socket.
 

Thread Starter

Ziggylittles

Joined Jul 2, 2017
36
The meter is digital and normally they have a delay in reading. It is possible the flashing did not allow for accurate reads at all. I can try taking a reading with the amber constant on instead of flashing.
 

Thread Starter

Ziggylittles

Joined Jul 2, 2017
36
I used the multimeter again but this time on one of the bulbs I have in the rear of the truck (amber for brake/turn signal and white as a reverse light). 2.8 amps was the result each time.
 
Top