TwinTurbo.NET: Nissan 300ZX forum - LED Turn Signal Timing Circuit Modification.
People Seeking Info
 
   


     
Subject LED Turn Signal Timing Circuit Modification.
     
Posted by Ash's Z on October 07, 2015 at 4:51 AM
  This message has been viewed 2431 times.
     
     
Message The Z32 is equipped with a current sensing relay for the turn signal markers that doubles the blinking rate if a bulb is burned out. Most all vehicles are equipped with this type of configuration to indicate the fault.

This fault indicating system is all packaged within the blinker relay itself. This relay carries the Nissan logo but appears to be manufactured by a company "Niles", with a part number "IF302". Because it has the ability to change the conduction rate based on circuit load, this relay isn't your average relay, or even just a timer relay. It has the ability to sense the circuit load and depending on said load, switch between pulse rate "X" or "X*2".
The problem with converting to LED-based turn signal lamps is that they draw so little current compared to the original filament-based bulbs that came in the vehicle. Because of this, when LEDs are installed, this relay goes into "fault mode" and blinks the lamps at the faster "X*2" rate. While they still work, any law enforcement officer will recognize this as your vehicle having a light out and provide them with cause for a traffic stop.... not to mention, most everyone else on the road would probably scoff at the sight and think things like "Heh, look at that moron in his fast, fancy car who can't afford to replace his turn signal bulbs!" Ya, none of us want to be that guy!
The relay circuit uses a proprietary IC to handle the load sensing and timing of the relay. It is marked with a "NILES 0011B" with a subscript of "+2K". After some exhaustive searching on the web I gave up trying to find a datasheet on this IC and then took steps of reverse-engineering into action.

After a few minutes of speculation as to how this IC may work, observing the other components within the circuit, and then composing a few theories as to how it can be configured, I went forward with the testing.
An analog timing circuit is rather crude when you break it all down into its constituent components. A 555 timer circuit is one of the most crude, simple, perfection of elegant and robust design that has ever been engineered. It is essentially a battery charger and battery discharger. We all know it takes a certain amount of time to charge a battery and depending on how much you put that battery to use after it is charged, it will take a certain amount of time to discharge. This is how an analog timer circuit works (specifically the 555 timer).
So lets build a charger that will charge a battery. Lets say this battery/charger combination will charge the battery in 1 second. Then, lets build a discharger that will fully discharge the battery, but this "discharger" has a control knob that allows us to control the amount of time it takes to discharge this battery.

Now lets set the discharger circuit with a resistance that discharges the battery in 1 second.

In this case, to fully charge and then fully discharge the battery it is going to take 2 seconds for this cycle to occur.

In this charging/discharging circuit, we configure it such that it continually charges and discharges the battery, over and over and over and over.
Now in this circuit design, we have a sensor that monitors the charge state of the battery. This sensor can tell if the battery is fully charged or fully discharged. When the battery is fully discharged, this circuit has a terminal that outputs zero volts. When the battery is fully charged, this same terminal has an output of 12 volts. In this case, the "sensing circuit" will switch "on-off" as the battery cycles through the charges/discharges.

It should be pretty obvious at this point how such a configuration is called a "timer".
Side note: I've been using the term "battery" to simplify the explanation but this isn't exactly correct. The correct term is a capacitor. Capacitors and batteries are similar in the fact that they will store a charge, however, capacitors are like really small batteries that can be quickly charged and quickly discharged. Capacitors can also be fully "charged" or fully "discharged". They can also be charged "in between". Meaning: You can charge a capacitor to any amount of voltage within its maximum rated limit. Capacitors are available in all sorts of ranges - anywhere from millivolts up to thousands of volts capacities.

Back to the turn signal timer relay:
While it should be clear how a timer circuit works at this point, the one thing I haven't addressed is how a circuit measures electrical load, or resistance.
There are many ways to detect if current is flowing through a circuit. Most of you are familiar with a timing light - where it has a clamp that you place over the ignition wire which provides the signal to the strobe light circuit to make it light off the xenon bulb as the ignition signal to the plug is delivered. The "clamp" is simply an inductor: a coil of wire that picks up the magnetic field produced when the current flows through the plug wire. That is one way to detect the flow of electrons through a wire.

Another way of doing it is to put a battery in series on one side of the circuit and monitor the charge state of that battery, or capacitor. When current flows through a circuit with a capacitor in series, the device in the circuit that is doing some kind of work, be it an electric motor, a solenoid, a coil, a light, etc, some of the power moving through that circuit will be absorbed by this capacitor. The result of this is that the capacitor will have slightly more voltage potential than it did before you passed current through the circuit. If you had a sensing circuit that monitored the voltage of this capacitor you would be able to detect if the circuit was open or closed. i.e. if a bulb is blown or not.
So I went in to speculation once again... hypothesizing.
The problem with LEDs is that they require VERY LITTLE current to operate.. They are VERY efficient at converting electrical power into light energy. The result of this is that they also require very little current to operate.

This is where the problem, two fold, comes into play.
We already know this current sensing relay will switch between normal flash rate to a rate of 2x if it senses a reduction in electrical load, but even if we were to modify the circuit in such a way that lowers the electrical load that causes it to switch from normal switch rate to the faster rate, LED's require such miniscule amounts of power that the circuit wouldn't be able to detect the small difference between the condition where all of the LEDs were working or if one were blown.
Adding to that, automotive electrical systems are incredibly noisy, electrically speaking: the alternator's output is based on an alternating-current mechanical system where the stator and rotor dipole interaction produces a highly rippled current flow... And the ignition system: producing anywhere from 15-30 thousand volts of discharge through the spark plugs, grounded through the block, connected to the main electrical "body" ground (and anywhere from 26 times a second for a 4-cylinder at idle all the way up to 466 times a second for a V8 at 7000RPM)... In short, there is a lot of voltage fluctuation in an automotive electrical system.

The result of this voltage fluctuation combined with the use of low-amperage LED's that are controlled by a load-sensing circuit is that, after a few hours of trying to stabilize the response of this timer relay in all conditions, it became clear that the only solution would require measures far beyond what any rational person would invest.
Simply put: making any adjustments to the charging capacitor in the load-sensing circuit was fruitless.. The LED's simply pulled far too little amount of current when they were on to be able to detect the difference of when one of them were "blown".

So, as with almost every benefit, also comes a sacrifice.

With that said, I propose to you a solution:
In order to run LED's in your Z, you are going to have to give up the blown bulb detection benefit. Fortunately, LED's are quite robust and the fact that these "bulbs" are made up of an array of ~21 different elements, even if one of them goes out, you still have many more to do the job.

The solution:

Once you have the relay out, you will need to trim off the molded edge of the body casing so you can remove it from the base. Once this is trimmed you can pull the guts out.

When looking at the assembly with the component side up, there is a white silkscreen print on the board labeled "VR2". There is a small 1/8W resistor in this position. The resistor has a value of 45KOhm. De-solder this resistor and replace it with a 100KOhm, 1/8W counterpart. This is "slightly" more than needed, but will result in a slightly slower switching rate... slower is cooler, anyhow.. wink emoticon
Put it all back together and seal it up with some superglue, silicone,... reinstall..

What you have done here is simply changed the "discharger" value such that when the relay is in "fault mode", it will actually blink at a "normal mode" rate... plus about ~10%... just to be a little different.

If you were to put normal bulbs back into the system after this mod, they will blink very, very slowly, BTW... (as to be expected)
If you want to be clever and go one step further with this, instead of installing the 100K resistor, put in a 200K potentiometer. This will allow you to adjust the blinking rate anywhere from 2X normal rate down to ~10Hz... (I did not test the absolute maximum switching rate of the relay..)

You are welcome. ;-)


Enthusiasts soon understand each other. --W. Irving.
Are you an enthusiast?

If you are out to describe the truth, leave elegance to the
tailor.
Albert Einstein

     
Follow Ups  
     
Post a
Followup

You cannot reply to this message because you are not logged in.