Simply put, what I am trying to do is modify my car's turn signals so that when the turn signal lever is activated for more than about 400ms it will act normally and cancel as soon as the lever is released, but if it is activated for less than 400ms or so, it will activate a time delay so it will stay on for 3-5 flashes.

I do a lot of highway driving (lots of drive right pass left) so I change lanes a lot. I've seen some newer cars with this feature where you just momentarily move the turn signal lever and it will flash several times, but if you hold it then it operates as normal; so I want to replicate this in my car.

I've looked into time delay relays, but as far as I can figure out, that solution would mean the turn signal would stay active for a given amount of time after the lever is released all the time, not just when momentarily activated.

I have basic electrical knowledge, and virtually no electronics knowledge, so I don't know if there is some sort of relay(s) to use, or if a microcontroller would be a better solution. I have some programming knowledge so figuring out that part shouldn't be too hard, I just need to know the best route to take on getting there.

My current flash-regulator rate fluctuates with voltage, and it's an older car, so the voltage goes around between 12-14.5V and the flash rate is all over the place, so maybe I could kill two birds with one stone and use a uC to also regulate the flash rate as well.

Reading up on some different uCs, it seems like it should be fairly easy to accomplish what I am trying to do, but I still can't figure out how to get the time delay to operate only when the turn lever is active for the given period of time. (with a 50-100ms lower cut-off to prevent the opposite side time delay from engaging when the lever returns).

I'm a poor college student, so cheaper solutions are better solutions even if it's more difficult; I like a challenge

 

basically form what it sounds, you will only need something to control the actual flasher relays, so you will have to put in some controlling circuit in between the flasher lever input and the flasher relays..... so if it "sees" an input at a given amount of time it would flash the flashers a set amount of times (or time period), the controller then continuously checks input from lever and basically repeats the same process if still "on", if not it just goes back into "standby mode"... so a dual 555 might work (556) for each flasher side, one timer in the dual will be set as a one shot, and the second timer can be set up as an astable to flash the lights at the desired speed..... you will have to have a dual timer for each side......

My .02

 

You can use a 555 timer as a one shot which is OR connected to the power line that goes to the turn signal flasher, so whichever power is on longer runs the turn signal duration. The problem with car turn signals is they also have the four way flashers wired into them and the wiring is kind of complicated.

 

You can use a 555 timer as a one shot which is OR connected to the power line that goes to the turn signal flasher, so whichever power is on longer runs the turn signal duration. The problem with car turn signals is they also have the four way flashers wired into them and the wiring is kind of complicated.

He could use some diodes on the output of the 555 circuit to buffer it it he ever had to turn the 4 way flashers on..... By just tapping onto the existing VCC wires off of the flashers, should not affect any other signals to them.....
Same concept as alarms systems that flash lights, etc. when alarm is set off... or flashing the lights with the car finder button......

My .02

 

Is this what a typical turn signal flasher schematic would look like?

-Adam-

 

A similar request came up in onther forum a while back, and I posted an ideas circuit using logic ICs and passive time constants, as an alternative to a uC based option.

Even then I made the comment about regulating the system using a simple 5V regulator -to help keep the flash rate constant!


Not what you are looking for I suspect -everything is PIC based these days -- but it might give you some ideas, if you decide not to go the software based route

 

You can use a 555 timer as a one shot which is OR connected to the power line that goes to the turn signal flasher, so whichever power is on longer runs the turn signal duration. The problem with car turn signals is they also have the four way flashers wired into them and the wiring is kind of complicated.

If I recall correctly, I think the four way flasher mechanism in my car bypassed the turn signal circuit.

Here is the schematic (or at least what it is supposed to be). My car has the digital cluster without the trailer adapter. I can only make some sense from this.

Thanks for the ideas!

 

basically form what it sounds, you will only need something to control the actual flasher relays, so you will have to put in some controlling circuit in between the flasher lever input and the flasher relays..... so if it "sees" an input at a given amount of time it would flash the flashers a set amount of times (or time period), the controller then continuously checks input from lever and basically repeats the same process if still "on", if not it just goes back into "standby mode"... so a dual 555 might work (556) for each flasher side, one timer in the dual will be set as a one shot, and the second timer can be set up as an astable to flash the lights at the desired speed..... you will have to have a dual timer for each side......

My .02

I see how the astable part would work, but I don't know how the one shot part would work. It would need to send the out signal to the astable for, let's say, 3 seconds after the trigger signal is gone but only if the trigger signal is present for more than about 50ms but less than about 400ms (just numbers that sound good, might want them to be a bit different in practice), if the trigger signal is present for less than 50ms or more than 400ms, then the one shot would stop sending the out signal to the astable immediately.

This is a learning process for me, so please bear with me

 

The lane change flash is a nice feature that many new cars have.
It used to be just the luxury cars but it's starting to show up on other cars too.
Are you sure you need 3 seconds of on time?
Most of the cars I have seen (used to be a Cadillac Technician) only flash like 3 times for a lane change.
A 555 would easily handle this.
And if you think about it, 2 seconds after you release the lever isn't that long even on a regular turn.

 

possibly using some R/C de-bouncing circuit to the input trigger of the one-shot (monostable) that way you can have some time constant in there for the cap to build up charge to trigger the first timer..... just trying to guess here really, just working things off the top of my head....... if you have a schematic of something we could work with..... let me see what I can dig up....

 

@9

Yeah, I think 3 times might be a bit short though. I'd want to play with 3-5 flashes to find which I like best. It would be an option to just have it continue to flash a few times after the lever is released all the time, but I would like to avoid that if possible. That's why I was thinking maybe the programable microcontroller could work, and then I wouldn't need to change parts to find the number of flashes I like. It seems like it would work well, but I still wouldn't know how to program the 50-400ms part with BASIC. Not that I know the language well, just from what I was reading I didn't see a possible way to do it.

@10
I have the schematic in post #7. Sorry for it being split.

 

When the schematic I posted with post #6 was actually built, I uncluded jumper links from the 4017 outputs. Allowed you to select from 1 to 9 flashes by just moving one jumper link.

Chips cost less than $5 -- built on a bit of stripboard. No programming required (other than the jumper link!)
Not sure that a microcontroller is always the easiest answer for a one off project!

 

So will this circuit flash 1-9 times all the time after the turn signal cancels? Or only when the lever is active momentarily?

 

1st momentary switch - preset number of flashes, then auto cancel.

'Hold' for longer than the C3, R3 time constant, - output continues to flash, until a further momentary switch press cancels immediately.

Any momentary switch activation, at any time the flasher is active, even during the fixed flash number period, will cancel immediately.

The circuit shows an output relay that will provide a volt free changeover contact that follows the logic at the flash rate, and was originally connected straight to the flasher bulb. A second channel is needed for the other side flasher of course.

How that logic is integrated into a specific existing flasher arrangement, particularly if there is any mechanical latch involved, will of course depend on how that system functions in the first place, and how much of it you are intending to replace?

 

I'm pretty adamant about this (I like my projects) so I'll do what it takes to get it done. I plan on picking up the parts today and figuring out how to build it. I see where I connect the turn lever, but I'm still unsure how to tie it into the system.

So once you hold it for longer than the C3 time, is the only way to cancel it to activate the switch again momentarily? Like if I'm sitting at the light with the signal on, once I complete the turn and the lever automatically returns will it cancel when the switch deactivates or will I need to activate it again momentarily to get it to stop flashing?

I'll need to bypass the regular flasher obviously, which has an electro-mechanical clicker. Where could I tie in another clicker? Just put it on the same system output that interfaces with the existing wires?

 

Yes - the original project was to replace a damaged mechnically latching lever control flasher, with simple momentary switches, which performed the descibed functions sequentially. No auto cancel of the 'latched' on condition.

If you're just intending to add the feature to an already functioning flasher, with mechanically latched levers, then you won't need most of that circuitry!
All you would need to do is to use a left and right 'on' contact that you can activate with a simple (non latching!) activation of the existing lever. Use that to activate a flip flop that will hold a relay contact on for a predetermined time, across the existing siwtch contact, and then reset.

Using the standard 'latch' would simply override the function, because it would be in parallel with the 'timed' output. You'll need one for each side of course. I could sketch something if you're interested?

 

I still want it to operate as normal where it will flash as long as the lever is latched, and stop flashing as soon as it becomes unlatched. If you know of a way to do that it would be great

I just want the auto-flash to happen when the lever is bumped for a moment. That's why I was thinking a uC might work, but then again I don't know much about it.

 

I still want it to operate as normal where it will flash as long as the lever is latched, and stop flashing as soon as it becomes unlatched. If you know of a way to do that it would be great

I just want the auto-flash to happen when the lever is bumped for a moment. That's why I was thinking a uC might work, but then again I don't know much about it.

I was driving my womans Charger this morning, and I noticed that it has this feature, but the lever has a momentary contact (blinker blinks 3 times then goes off), and a latched contact, so yours probably just has the latched position right? So you might have to add this momentary contact if possible.....

 

If you know of a way to do that it would be great

Something like the attached sketch should do what you need. You might need to select R2 to meet your time needs, and the C1/R1 debounce values may need to be modified to make sure you don't get false triggers.
The 4093 and LM393 are cheap, and should be easy to find.

The second half of the LM393 should be OK to drive the relays I've suggested, but if you shoose a relay that needs more than about 20mA to energise the coil, then you may need to add a separate drive transistor.

Just an few ideas to tryout if you want!