Dimmer circuit for Omori gauge

As mentioned in the intro the Omori gauge uses LEDs for backlighting, white for the face and red for the needle.  When connected to +12V the gauge is VERY bright - enough to see the illumination even on a sunny day.  This is clearly way too bright at night, and in fact proved to be somewhat of a distraction to night driving.  The next problem to be tackled was to bring the gauge illumination down to the level of the rest of the cluster.

My first attempt to reduce the brightness was only moderately successful, and was simply a resistor in series with the illumination wire.  This served to reduce the current flow and therefore brightness of the backlight LEDs.  Before I go further, a little LED theory.

LEDs, unlike incandescent globes, maintain a relatively constant forward voltage irrespective of the current flow through them.  Brightness is controlled by varying the current, and is usually set by a series resistor.  The value of this series resistor is calculated using Ohm's law, with the voltage derived from the difference between the total supply voltage and the LED's forward voltage.  When two or more LEDs are connected in series, the forward voltages of the LEDs are added together for the calculation.

LED circuitIn this example (at left), two LEDs with a forward voltage of 3.5V each are connected in series to a 12V power supply, via a resistor to limit the current.  The voltage across the resistor will be 5V as shown.  If we require a forward current of 20mA, the resistor value can be calculated using Ohm's law, as follows:

R=V/I  (R is resistance in ohms, V is volts and I is current in amps)
=5/0.02 (20mA  = 0.02A)
=250 Ohms

I did not trace the backlight circuit in the Omori gauge, but from observation of the effect of fitting a resistor in series with the illumination wire, it uses two or more circuits similar to the example above.  The resistor effectively reduces the total voltage available to the backlight circuits, reducing the voltage across the internal series resistors, and hence the current through the LEDs.

Gauge with red castThe problem with this arrangement is that the Omori gauge uses red and white LEDs, which typically have different forward voltages.  Red LEDs are usually around 2V but white are 3V or more.  When dropping the total supply voltage to the backlight circuits, a point is reached where the red LED circuit's resistor still has voltage across it and the LEDs illuminate, but the higher voltage requirement of the white LEDs is no longer met, and the resistor voltage drops to zero.  Just above this point, the white LEDs dim to a significant degree but the red LEDs are still bright resulting in a red cast across the entire gauge face and red halos around the needle boss and surround (see photo).  The backlight also becomes very sensitive to voltage fluctuations which frequently occur, for example each time the brake lights operate the white backlight dims momentarily.

In order to dim the white and red LEDs evenly and maintain the proper look of the gauge, a technique different than simple voltage dropping was needed, and that technique is pulse width modulation or PWM.  Due to its efficiency, PWM is in very common use in power supplies for all kinds of applications, from mobile phone chargers to air conditioners to PCs.  Rather than reducing voltage, in PWM systems the full supply voltage is applied for a controlled period, then removed for a further period.  The duty cycle is the ratio of "on" time to the total "on" plus "off" time or cycle period, and is proportional to the actual power applied to the connected electrical load.  The cycle period is very short, usually running at several thousand cycles per second depending on the application.

Backlight using PWMUsing PWM to power the Omori backlight means that rather than dimming the LEDs, they are operated alternately at full brightness and off.  LEDs are able to switch on and off very quickly, and at a sufficient cycle rate our eyes perceive this as a reduction in brightness rather than flashing.  Because the red and white LEDs switch on and off simultaneously, they appear to dim evenly.  The circuit shown here runs at around 140 cycles per second - more than enough for a flicker free result.  Voltage fluctuations have virtually no effect, because there is still a significant voltage across the current limiting resistors in the gauge.  As can be seen in the photo, the red cast and halos are not evident.

Circuit details

It may sound complex, but the PWM circuit is actually very simple - only eight components are required, and all are cheap and readily available.  Below is a circuit diagram and brief description of the circuit operation.

PWM circuit for Omori LED backlightThe LM555 is a timer which derives its period by charging and discharging a capacitor via resistors between 1/3 and 2/3 of supply voltage.  The capacitor and resistor values determine the actual timing periods.  In this circuit, the output (pin 3) is initially high (+12V), powering the LED backlights and charging the 0.1uF capacitor via diode 1N4001 and 5K trimpot.  When the capacitor reaches 2/3 (around 8V) supply voltage, the LM555 output (pin 3) switches low (0V) and pin 7 goes low.  The capacitor then discharges via the 100K resistor and 5K trimpot.  Once the capacitor discharges to 1/3 (around 4V) supply voltage the output switches high again, pin 7 goes open and the cycle repeats. 

Adjusting the 5K trimpot allows variation of almost 0% to around 5% duty cycle.  In practice the Omori gauge closely matches the brightness of the rest of the cluster at around 2% duty cycle, so the adjustment range is ample.

None of the components in this design are critical and substitutes may be used.  For example the 1N4001 diodes could be replaced with 1N4004, I used greencaps for the two capacitors on the right side of the diagram, but ceramics could be used.  The 100uF capacitor is only for filtering and anything from 47uF to 470uF could be used.

Building it

There is a small PCB below the tacho in the MR2 cluster which is actually a signal conditioner for the factory boost gauge and may be removed.  I used this space to advantage and built a small PCB with the dimmer circuit to take its place.  Pictured below are the original PCB (left) and the new dimmer (right).
Old and new PCBs
And here is a solder side view:
Solder sideThe threaded lugs on the bottom of the board connect to the same positions on the cluster as the original board, and I have used these to carry the electrical connections to the new Omori gauge.  The unsoldered pads on the left side connect to the Omori gauge loom to complete the connections.  The full details of these connections are explained in the Electrical Wiring In Cluster page.
Here is the PCB fitted to the cluster and wired to the gauge:
Dimmer fitted to cluster
There is no reason for this circuit to be built exactly the way I have, but if you'd like to build the PCB to fit as above, the solder side photo should show enough detail to replicate my design.  The PCB is 27 x 50 mm, and just copy the spacing of the threaded lugs from the original board.  I used some screw terminals off old speakers to make the threaded lugs - just had to reshape them a bit and cut off some tabs.  You could desolder the lugs from the original PCB and use them - I didn't because I wanted to keep the original board intact.  If making a PCB is not practical, the circuit could quite easily be built on veroboard or the like, and it certainly doesn't have to be fitted inside the cluster.

Back to intro