Power Management For Automotive LCDs

Oct. 1, 2004
Automobile manufacturers are increasingly using liquid crystal display (LCD) technology to differentiate their products, particularly at the luxury end

For the PDF version of this article, click here.

Automobile manufacturers are increasingly using liquid crystal display (LCD) technology to differentiate their products, particularly at the luxury end of their product lines. This trend began with the introduction of global positioning system (GPS) navigation displays centered on the dashboard for viewing by both driver and passengers. Initially offered as options, these displays are now a standard feature in most luxury models.

More recently, automakers have begun adding entertainment systems to their high-end vans and SUVs, providing full-motion video for rear-seat passengers. Now, some premium vehicles even offer full-motion video for rear-looking CCD cameras on their GPS displays, to provide added safety while moving in reverse.

On the drawing board, automakers are designing cluster displays that will allow drivers to customize their speedometers, tachometers and engine monitoring gauges in formats tailored precisely to their preferences.


As with all flat panel displays, side or backlighting is the essential element in providing readable data presentation. Automotive displays also face a critical need for effective brightness control. Display brightness is a serious safety concern because levels required for daytime viewing can blind drivers at night. Display intensity that is more than 20% above optimal levels can cause momentary blindness and threaten driver safety.

To address this significant liability issue, designers have focused on incorporating sophisticated power management functionality that provides true lamp dimming across the entire ambient condition spectrum. This approach is suited for full-motion entertainment and back-view video displays as well.


Performance requirements for automotive LCDs differ significantly from flat panel units used in other applications. To operate in bright, sunlit ambient environments, automotive display backlighting systems must provide outputs of as much as 640 candelas/m2 (nits), about five times that provided in a typical laptop computer. This is the primary reason LEDs are not yet an acceptable alternative to cold cathode fluorescent lamps (CCFLs) in automotive display backlighting. Current LED offerings provide light output an order of magnitude below what is needed to meet dashboard readability requirements in full daylight conditions. Development is under way to address the cost and heat obstacles in providing these hi-bright LED-powered displays.

At the opposite end of the brightness scale, minimum backlight intensities to provide flicker-free readability in nighttime driving conditions are as low as three to five nits. To provide a consistent, safe, visual environment for both driver and passengers, CCFL automotive displays must be able to handle dynamic dimming ranges of 225:1 — about 10 times that typically provided in the best computer displays. Advanced digital dimming strategies can attain this dynamic range.


The temperature extremes from -30°C to +85°C encountered within an automobile further complicate display ballast design. CCFLs used for backlighting must provide the same high outputs on crisp February mornings in Minnesota as they do on balmy summer days in Florida to meet readability standards. In the cold, failure to do so promptly could lead a driver to believe that the display is not functioning.

CCFL lamps are normally rated for peak illumination at 30°C to 40°C above a 20°C ambient, and as temperatures decrease, light output falls off rapidly. In fact, with standard CCFLs, output is virtually unusable below zero degrees, because the small amount of light emitted is often pink or purple and far from that required for good data representation on an LCD. As a result, self-heating lamps have become a virtual requirement for automotive displays and can provide enough white light at cold temperatures to backlight a display.

Self-heating lamps — particularly L- or U-shaped devices now employed in automotive displays — can require run voltages as high as 1,500 V, about twice the operating voltage of CCFLs used in notebooks. They also require the use of boost or current overdrive techniques to achieve driver-suitable light levels quickly when displays are turned on in cold weather. For example, increasing the drive current in the lamps after ignition by 50% over normal operating levels results in three times the light output at low temperatures and also accelerates lamp heat-up to optimum operating temperatures in one-third the time.

While lamp manufacturers felt at first that overdriving their products would significantly reduce lamp life, extensive laboratory studies have shown just the reverse: Overdriving actually prolongs lamp life in cold ambient temperature environments. This is because operating at cold temperatures is much more detrimental to CCFLs than driving the bulbs for short periods at elevated currents, as long as maximum bulb temperatures are not exceeded.

Current overdrive of as much as 50% and for as long as 30 seconds is now typical. This essentially operates the CCFLs at voltages approaching lamp strike levels and elevated currents until warm-up is well under way. Consequently, step-up transformers used in non-automotive PWM-driven ballasts require significant re-engineering before use in automobiles — both to operate at required overdrive voltage and current levels and to meet target cost and form factors.


Bulb fault detection is particularly critical in automotive back-lighting systems because any arcing caused by a burned out or accidentally damaged CCFL is simply not allowable in the possible presence of vaporized or liquid gasoline. In virtually all inverter topologies, the sensing of bulb current provides fault detection — a lack of current denotes an open circuit and a failed or damaged CCFL. At low brightness levels and low duty cycles, differentiating lamp current from parasitic leakage can be difficult. However clever circuit design can meet this challenge.


The ballasts or inverters that supply power to turn-on (strike) and run CCFLs are technically challenging circuits. These inverters must accept a wide range of dc input voltages — from 8 V to 16 V in current automotive systems — and provide ac outputs of up to 3000 V to run the lamps when at low brightness levels. Making matters even more challenging, these same circuits must also provide a momentary strike voltage twice that of the typical run-voltage to ignite the lamp. In automotive applications, the circuits must be able to supply the current overdrive to rapidly heat the lamps during cold start conditions. In addition, all automotive applications require effective wide-range dimming capabilities to allow lamp output to match ambient light conditions and to prolong lamp life.

Moreover, CCFL inverters face the ongoing demands for ever-increasing efficiency to reduce heat and minimize power consumption, while simultaneously meeting the ever-lower cost models required in automotive electronic systems.


For many years, display manufacturers employed a Buck/Royer inverter topology (Figure 1) to strike and power CCFLs. This analog power topology is essentially a combination of a step-down buck voltage regulator and a self-resonant Royer oscillator with an integral step-up transformer.

The buck regulator consists of two power semiconductors, one power inductor, a PWM regulator IC and a high current capacitor. It provides a stable power source from the battery or line potential that can be varied to provide lamp dimming. The Royer oscillator consists of two transistors, a high-current resonance capacitor, a high-voltage transformer and a ballast capacitor in series with the lamp. It provides an alternating current source to drive the lamp.

The ballast capacitor controls current amplitude through the negative impedance of the lamp by dropping an approximately equal voltage across its positive impedance. Lamp current will be the total transformer output voltage minus lamp voltage (lamp voltage is inversely proportional to current) divided by the capacitor's impedance.

The minimum size of this transformer is limited by total power and strike voltage requirements. Because it must operate at twice the normal lamp running voltage all the time, its size must support full strike power and voltage, even after the lamp is ignited. This increases its size over new technology inverter transformers because larger dimensions are needed to protect against failure from corona discharge and high-voltage breakdown.

The Buck regulator allows a stable input voltage to the Royer oscillator, providing good line and load regulation across the entire input range. Lamp brightness is, therefore, insensitive to both static and dynamic variations in input voltage, and fairly constant conversion efficiency is achieved over the entire input voltage range. The Royer oscillator delivers a low crest factor sine wave current output signal to the CCFL, which is ideal for efficiency.

Dimming is straightforward, with amplitude modulation schemes providing a typical dynamic range of up to 5:1 — an obvious deficiency in most automotive applications. A simpler but significantly limiting dimming solution used in many automotive aftermarket products reverses display contrast as ambient light decreases below preset levels. This typically involves the simultaneous change in background color from light to dark and on-screen information from dark to light. The effectiveness of this approach is rudimentary and has questionable ergonomic characteristics and no application for video displays.

There are further shortcomings. The extra inductor, power transistors and the larger transformer of Buck/Royer inverters tend to keep costs and the physical size of the circuits high. Also, the relatively high currents flowing through the power semiconductors and large capacitors account for high losses, limiting overall efficiency to about 75% or less. Premium components can rectify this, but they raise costs further while doing little to reduce footprints.

More important, multiple resonances within the circuit, plus the lack of a voltage feedback mechanism and frequency control from the controller IC, make it difficult to limit the open-circuit voltage generated by Buck/Royer inverters. An unplugged or broken lamp can lead to arcing and self-destruction of the entire inverter assembly. In addition, with the Buck regulator and the Royer oscillator operating asynchronously at different frequencies, EMI and RFI are difficult to control — a real negative for displays that will be placed close to radios in the center console of a vehicle.


A number of innovative inverter topologies have been developed recently that combine more capable controller silicon with improved transformer designs that optimize performance, reduce component count and lower costs. One of the simplest of these advanced inverter topologies employs an LX1686 controller to implement a direct drive architecture that eliminates the need for the inductor and resonant capacitors found in a conventional Royer oscillator.

Instead, direct drive topology (Figure 2) uses a fixed-frequency pulse width modulation (PWM) control circuit connected directly to a high-voltage transformer primary through a pair of N-channel MOSFETs to provide a fixed-frequency regulation method for running CCFLs. The LX1686 uses a novel resonant striking technique that permits removing the inductor and resonant capacitor as well as replacing the high-voltage ballast capacitor in series with the lamp with a low-voltage dc blocking capacitor. Removing these costly and power-hungry components significantly improves inverter module cost, efficiency and size.

In addition, the high-voltage transformer used in the direct drive implementation has one less winding than that used in a Buck/Royer design, which leads to smaller form factors and lower costs.

By using a low-impedance dc blocking capacitor in place of a high-impedance ballast capacitor in the lamp circuit, the transformer output voltage required to operate the lamp after ignition is reduced by 50%, adding to efficiency and reduced size. Finally, a non-dissipative capacitor feedback signal can easily be provided from the transformer to provide the active open-circuit output voltage regulation needed to eliminate open-circuit hazards.

The 2-transistor N-channel drive scheme provides several advantages over similar configurations that use popular bipolar or P/N channel MOSFET drives:

  • Using ground-referenced transistors in conjunction with push-pull transformer operation allows the controller IC to be implemented in a low-cost, low-voltage CMOS process providing small die size and high performance while guaranteeing compatibility with common system bus voltages.
  • The controller IC can be interfaced through the N-channel MOSFETs to any desired system voltage simply by changing the turns ratio of the high-voltage transformer, providing an input voltage range of 3 V to 50 or more volts for the inverter module.
  • N-channel MOSFETs are significantly more efficient than bipolar transistors or P-channel devices of equal size and cost, and dual-channel N-FETs are readily available in small surface-mount packages to help reduce parts counts and module size over bipolar transistor implementations.

There is a limitation to the direct drive topology. PWM regulation with wide-ranging input voltages can cause high lamp current crest factors that limit direct drive inverter efficiency when at high line to approximately the same as that achieved using a Buck/Royer design. However, with regulated input voltages (8 V and 16 V are typically encountered in devices by automotive batteries), these crest factors are well within desired limits. So in most automotive applications, direct drive inverter efficiency is 10% to 30% higher than Buck/Royer implementations.


While the direct drive inverter topology is the lowest component count and lowest cost solution available for implementing a PWM-driven inverter, other PWM topologies are available that offer advantages in specific applications with minor increases in parts count, physical size and cost. For example, for input voltages above 40 Vdc that are proposed for advanced automotive power systems, configuring the circuit (Figure 3) as a half bridge will provide significant improvements in output current crest factor. Thanks to moderate MOSFET losses, this leads to improved efficiency and EMI performance.

Adding another pair of MOSFETs and a resonating secondary side capacitor in a full-bridge configuration (Figure 4) yields an even better output current waveform appropriate for today's and tomorrow's automotive displays, again raising efficiency with only minor increases in component count and BOM cost. When used in resonance, this topology can support 3:1 wide-ranging input voltages, and it develops little EMI.


Using PWM techniques to drive the high-voltage transformer provides opportunities for implementing new strategies within the controller for striking CCFL lamps and improving dynamic dimming ranges and efficiencies. For example, the high-voltage transformer and the output capacitance in the direct drive topology have an unloaded self-resonant frequency that is higher than the normal operating frequency. Sweeping the PWM oscillator through a 3:1 frequency range creates a rise in lamp voltage independent of the transformer turns ratio as the lamp circuit approaches resonance. This voltage rise can be up to 5 kV or even more in high Q circuits like this and can easily generate enough voltage to ignite a CCFL while maintaining transformer flux density at low levels to prevent magnetic saturation.

Because direct drive uses a very low impedance blocking capacitor in place of the Royer ballast capacitor, a voltage divider with the inherent parasitic capacitance of the lamp and that would reduce available strike voltage during ignition is not created. In general, using this strike technique allows CCFL ballast controllers to strike lamps far below the rated maximum output voltage specified for a particular CCFL (Figure 5), and at far lower voltage levels than other commercially available controllers.

By monitoring the lamp current to detect ignition, the direct drive controller can provide the necessary strike voltage, as well as the current overdrive needed for fast warm-up of self-heating lamps, then return the PWM stream to normal operating frequency to match the most efficient run conditions for a particular CCFL. In addition, adding an active clamp circuit to the controller provides the ability to set a maximum voltage set point to protect lamps and backlight assemblies against overvoltage conditions.


PWM digital dimming is rapidly becoming the dimming technique of choice because it is less display-sensitive and offers more flexibility for choosing brightness levels. Digital dimming can achieve a dynamic range of 200:1 and more — far beyond any analog dimming strategy.

Unlike analog dimming — which relies on modulating the amplitude of lamp current — fixing the output lamp current amplitude and modulating its on-time duty cycle can be used to implement digital dimming. This turns the lamp supply voltage on and off at a rate between 100 Hz and 300 Hz, at a duty cycle the controller directs. Because lamp gases remain ionized during turn-off intervals, the inverter does not have to re-strike the lamp after each turn-off event. With the PWM dimming input in the 100 Hz to 300 Hz region, the human eye detects no flicker.

Dimming at low-light levels required for automotive displays creates additional challenges, even for the advanced digital techniques employed in today's PWM-driven inverters. For example, at the low duty cycles encountered when providing light levels below 1% of rated maximum, only a few lamp current cycles per burst are firing the CCFLs each second. Here, any variation in the number of PWM pulses per burst would appear as display shimmer, creating potential driver distractions. At the same time, lamp current shaping becomes important. It is essential to provide a 20% to 40% overshoot in current value at low duty cycles to ensure that the CCFLs light over their entire length, avoiding light gradients on the display screen.

In addition, when dimming the lamp digitally instead of with the conventional analog method, lamp current is kept at normal amplitude and bursted on and off to reduce the average current that corresponds to its light output. As a result, actual lamp current amplitude is kept higher than the leakage current even at high dimming ratios, so detection by electronic comparators is possible to give a reliable indication of lamp failure and provide the necessary fault protection.


With ergonomic factors clearly dictating that automotive display brightness be automatically tailored to the ambient conditions, the selection of a light sensor to accurately detect ambient conditions becomes the critical factor in implementing the digital dimming control system. The most common light sensors in use today are phototransistors and PIN diodes. Both generate signals that vary with incident ambient light intensity, but each has significant drawbacks in automotive display brightness control applications.

The photodiode or PIN diode offers a linear response to incident light intensity, producing a current proportional to the ambient lighting. PIN diode output current is small (typically nanoamps), so some sort of signal amplification is needed. The spectral response of a photodiode is most sensitive in the infrared area so it typically requires the additional expense of an infrared filter to ignore light outside the visible spectrum. In fact, contributions from both the infrared and ultraviolet regions of the spectrum tend to increase the detector signal far beyond the ambient level sensed by the user's eye.

Phototransistors have a response similar to a typical bipolar transistor with a photodiode driving the base. The transistor beta provides the necessary current gain, which amplifies the weak signal of the photodiode input. Unfortunately, the transistor beta is not a particularly stable parameter so the gain of the phototransistor circuit varies with supply voltage, with temperature and with manufacturing process tolerance; this means the phototransistor has a limited usefulness as a calibrated linear light sensor and is not recommended for automatic ambient tracking systems.

In many automotive brightness control applications, including in-car displays, mirrors and headlamp dimming, response to light outside the visible spectrum can cause severe performance and ergonomic problems. In particular, the sensitivity of commonly used sensors to the red area of the spectrum like that emitted by today's LED tail lights can produce faulty control signals, dimming the mirror reflectivity or display brightness unpredictably and often at inconvenient times.

A new class of integrated-circuit visible light detectors is just emerging on the market. These devices typically consist of an array of PIN diodes on a single substrate, with individual diode characteristics within the array controlled in such a way as to match its overall spectral response closely to that of the human eye.

With PIN diode arrays, it's possible to filter out the response to visible light. When the response of an infrared sensitive PIN is subtracted from an otherwise matched full spectrum PIN, the result is a diode that is sensitive only to visible light. Using current mirrors, it is possible to accurately amplify the PIN diode current by adjusting the physical size characteristics of the current mirror transistors (which is straightforward on an integrated circuit). This way, the good temperature coefficient and linearity of the PIN diode is reflected in the sensor output.

As an added benefit, the quiescent current of these sensors is typically negligible and the devices are useful at low light levels encountered during nighttime driving. When such a device is used with an appropriate controller for CCFL, the signal from the sensor can be used to directly control display brightness for ambient variations perceived by the human eye.


With the proliferation of LCDs in automotive applications continuing apace, and the demands for ever smaller size and lower cost for all components employed in automotive display products, the benefits from advanced PWM topologies in CCFL inverters will become the driving factor for their acceptance in the market. The range of available solutions, and their companion visible light sensors for implementing automatic brightness control, is already sparking the transition away from analog solutions like Buck/Royer inverters and analog dimming techniques. Look for new innovative PWM topologies to emerge and move quickly into a wide range of automotive display-based products and systems.


Henry has been involved in the CCFL inverter and controller IC design since 1990, as a consulting engineer and, since 1994, as a senior engineer for Linfinity Microelectronics, which was acquired by Microsemi in 1999. He has system engineering responsibility for the lighting, audio, and power supply integrated circuits produced by Microsemi and has authored several patents for CCFL inverter innovations.

Sponsored Recommendations


To join the conversation, and become an exclusive member of Electronic Design, create an account today!