Tailight sequencer not using PIC

[BiggyR]

Hi All,

I'm trying to figure out a retrofit taillight "chase" sequencer for an older car. Instead of one brake lamp on each side there will be three. I want to use the original wiring to the rear of the car and add a "black box" in the trunk. I intend to run a separate number 8 or so wire from the battery for power so the current doesn't have to go through the old wiring. Three lights on each side for turn signals that chase each other.

Logic is this:

1. If turn signal is on there is a pulse from the switch that triggers the chase circuit.This is the same wire that is on steady for brakes. This can be a 4017 driven by a 555.

2. If the brake pedal is pushed all three lamps come on, they do not sequence. Most of the circuits I have seen cause the lamps to sequence when the line goes high.

3. The turn signal overrides the brakes. In other words, if the brake pedal is pushed while signaling a turn one side sequences and the other has all three lamps on.

What should I use to drive the lamps? The 4017 has very little current capability. Do I drive a transistor that in turn drives a relay or can a mosfet be used for fewer parts count? Can the 4017 drive a power mosfet directly?

There are a number of circuits on the 'Net but I haven't seen one that uses only the two wires (left and right brake lights) and the original stock wiring. I don't mind running a power wire from the battery but don't want to get under the dash to change the car's stock setup.


Thanks,

BiggyR

 

[Norator]

If the falling edge of the turn signal triggers the chase sequence, how will your circuit distinguish between this falling edge and a brake pedal being released? I think the truth table for this needs more work.

 

[Diver300]

The 4017 can drive a power MOSFET directly. On a 4017 there is one output high at a time, so if you drive an N-channel MOSFET with the output, that will work. You will be switching the negative side of the lights, so the positive will be connected to a permanent power feed.

If you only have one wire for each lamp, then you may need to combine the signals from both sides of the car, and to have something that remembers what happened recently. For example, if the left wire is on, and the right wire is off, is the car turning left and not braking, or is it turning right and braking? You need to decide how to tell those conditions apart.

 

[3v0]

Is this what you are after ? Or do you really want one light on at a time ? I have to ask because the tbird style sequence is what most people are after.

 

[BiggyR]

No, I want the "chase" sequence, with only one on at a time. This is easy to do with the 4017. I have some IRLZ44 mosfets that should work. The ground side switching may present a problem as the bulb is a dual filament used for the lights, too. Maybe just make the socket "hot" and ground the filament. Then use a relay driven by the "hi" to the tailights (not the brakes) to ground the light filaments, which can all be wired in parallel. But then there would be no sequencing if the trun signal was supposed to be on also. Some logic truth table work needed here.

Yes, the logic is rather involved. Maybe I need to run a separate line from the brake switch? I want the lights to definitely NOT sequence when the brakes are pushed. I get enough attention from the cops because it is an old car.

 

[3v0]

It does not matter but now you have me curious about what the car is.

I won't be much help here as i would use a pic which is the better choice IFF you can program.

 

[crutschow]

Hang on, I'm working on a circuit using a couple 555's, a couple 4017's, and a few logic gates that I think will do what you want. Should have it by tomorrow morning (California time).

 

[BiggyR]

Car is a 1972 Comet, loke a Maverick. We already have it running good, even the A/C works with the old R12. We always thought the tailights were ugly and are replacing them with those from a 67 Impala.

 

[Norator]

Assuming the brake light is on for an indeterminate period t1 to t2, depending on your braking habits or necessity, and
the signal lights are on with a known pattern with ontime falling within t1 to t2,
your pattern detection circuit would need some time to discriminate, which means your turn signal light or brake lights coming on would be delayed.

The more delay, the more certainty the signal has been correctly identified.
This doesn't sound so desirable.

 

[3v0]

Car is a 1972 Comet, loke a Maverick. We already have it running good, even the A/C works with the old R12. We always thought the tailights were ugly and are replacing them with those from a 67 Impala.

Love it

 

[crutschow]

Below is a simulation of my circuit for doing the sequence:
All lights go on when the first pulse arrives.
If it's continuous, as when braking, then all the lights stay on.
If it is the turn-signal sequence, then all lights stay for the first pulse duration and the next pulse starts the sequencer routine.

LTspice doesn't show the pinouts for the devices, or the power and ground for the logic circuits on the schematic so you will need to get those from the data sheets.

The A, B, and C outputs can directly drive grounded source N-MOSFETS, or P-MOSFETS with their sources to +12V (directly to +12V, not V+ on the schematic). The MOSFETS should be rated for 100V or more.

All capacitors should be rated for 20V or more.

Edit: R6 and C7 provide a small delay to avoid a glitch in the sequencer first output pulse.

R1 and R2 should be 1MΩ trimpots so you can adjust the time for the two 555 periods. U1 must be adjusted to have a timeout slightly longer than period of the turn-signals (the schematic is set for a 1 second period). U2 determines the sequence rate of the lights.

Make sure you add 0.1μF caps from each IC power pin to ground for proper decoupling.

Of course you need two of these circuits, one for each set of tail-lights. The 4081 has left over gates which you can share between the two circuits as well as sharing V+. Connect any unused gate inputs to ground.

Since I haven't built the circuit, the actual circuit may operate slightly differently than the simulation shows, but I'm reasonably confident that it will work. If you built it and have any problems, post them and I'm sure we can resolve them.

Edited schematic:

View attachment Taillight Sequencer.asc

 

[Norator]

With vehicles there is a load dump transient which goes negative by several hundred volts which you may want to guard against.
The Special Functions (Automotive) section of National Semiconductor's Application Notes had some typical circuits.

This problem may be overconstrained in that there may be no acceptable solution given the constraints on the problem. I'm thinking now if the OP senses the brake switch closure he can greatly increase the reliability of this circuit at a small increase in complexity.

 

[crutschow]

With vehicles there is a load dump transient which goes negative by several hundred volts which you may want to guard against.
The Special Functions (Automotive) section of National Semiconductor's Application Notes had some typical circuits.

This problem may be overconstrained in that there may be no acceptable solution given the constraints on the problem. I'm thinking now if the OP senses the brake switch closure he can greatly increase the reliability of this circuit at a small increase in complexity.

You can help guard against the negative transient by placing a rectifier across C3 in my circuit (anode to ground).

I think my circuit pretty well meets the constraints of his requirements.

 

[MikeMl]

With vehicles there is a load dump transient which goes negative by several hundred volts which you may want to guard against...

Automotive load dump cause the vehicle bus voltage to spike positive. The problem happens when the alternator is outputting high charging current to the vehicle battery when something disconnects the battery (human or blown fuse/link).

The field of a typical alternator has a high inductance (several Henries). If the alternator is spooled up to ~50A, the field current is ~2 to 3A. If suddenly the battery is disconnected, the voltage regulator tries to shut down the field current, but due to the field inductance, it takes a good fraction of a second or more for the field current to decay down to the lower, correct value, during which the output voltage spikes to >50V. Normally the battery would be there to absorb the spike, but with the battery out of the circuit, bad things happen.

Moral of the story, never disconnect the battery while the engine is running.

 

[Norator]

This
https://www.google.com/patents/US5121066
goes both ways and the Nat. Semi. circuit I saw had a 1000 PIV diode such that negative transients were blocked, IIRC.
Dunno'. . .


Also, the OP can mount his own brake pedal travel detection Microswitch to avoid mucking with the OEM vehicle wiring.

 

[MikeMl]

Getting a reverse voltage in an automotive environment is extremely rare, and is usually caused by some Dumb Sh.. reverse jumping a dead battery (+ to -, - to +).

 

[BiggyR]

I think that for safety's sake and to be legal the sequence MUST be:

Left turn signal only- no brakes applied- sequence starts immediately as soon as there is power to the left side brake light wire from the turn signal switch. The right side should NOT get even a flicker of power.

Left turn and brake applied at the same time- Left side does the sequencing and right side all three lights come on at once and don't blink. No delay

Brakes appplied then left turn signal switched on- First all 6 lights on then the left side starts sequencing.

The simplest way to do this without any delay or false signals is to have a separate line from the brake switch to the back of the car. Cut the wire going from the brake switch to the turn signal switch and run it to the back. Now it is simple logic:

1. Signal from the turn signal- start the sequencer and use a pulse omission detector to keep the system cycling until the turn signal is switched off. Doesn't matter if the brakes or on or off.

2. Signal from brake switch- turn all lights on on both sides unless the turn signal pulse omission detector is running. If it is then the drive for the brake lights on that side is inhibited.

I have to study this some more.

original Comet tailights



What it will look like with "new" tailights



partial schematic of car- note wire from brake switch to turn signal switch.

 

[crutschow]

I think that for safety's sake and to be legal the sequence MUST be:

Left turn signal only- no brakes applied- sequence starts immediately as soon as there is power to the left side brake light wire from the turn signal switch. The right side should NOT get even a flicker of power.

Left turn and brake applied at the same time- Left side does the sequencing and right side all three lights come on at once and don't blink. No delay

Brakes appplied then left turn signal switched on- First all 6 lights on then the left side starts sequencing.

The simplest way to do this without any delay or false signals is to have a separate line from the brake switch to the back of the car. Cut the wire going from the brake switch to the turn signal switch and run it to the back. Now it is simple logic:

1. Signal from the turn signal- start the sequencer and use a pulse omission detector to keep the system cycling until the turn signal is switched off. Doesn't matter if the brakes or on or off.

2. Signal from brake switch- turn all lights on on both sides unless the turn signal pulse omission detector is running. If it is then the drive for the brake lights on that side is inhibited.

I have to study this some more.

.........................

If I understand you correctly, my circuit does that without any added brake signal needed, assuming that in your present circuit, the turn-signal on the single wire to the appropriate side keeps pulsing whether the brakes are applied or not.

 

[BiggyR]

I think somebody needs to breadboard a circuit to see how it all works. If I can find the parts in my messy garage.......

 

[crutschow]

I think somebody needs to breadboard a circuit to see how it all works. If I can find the parts in my messy garage.......

Sounds like the next logical step. Let us know how it goes.