A technique for multi-line addressing in OLED displays5 October 2009Organic light emitting diodes have unique drive challenges; this circuit uses the TFT at each active matrix display or the OLED diode in a passive OLED display as a demodulator to detect OFDM carriers.
Multi-line addressing is a method of driving one or more lines simultaneously in a display to increase frame rate without increasing line rate and in the case of OLED displays, multi-line addressing can reduce power consumption, improve lifetime and generally give active-matrix capabilities to passive OLED displays (Reference 1).
Because passive OLED displays have a truly active device (an Organic Light-Emitting Diode) at each pixel, this diode can act as a demodulator for amplitude-modulated orthogonal frequency-division multiplexing (OFDM) carriers on the rows and columns of the display. Although this may seem at first like an unnecessarily complicated approach to addressing pixels in a display (after all, we just turn rows and columns high or low for most displays), Figure 1 shows that any use of binary (digital) signals cannot simultaneously address pixels on more than one line without inadvertently addressing pixels on other lines. As shown in the figure, an attempt to digitally control two pixels in different lines (pixel 1 and pixel 8 in this case) results in turning on two more unintended pixels, pixels 1 and 7 which are the mirror of pixels 2 and 8.
Figure 1: Problems with digital multi-line addressing
(Click on image to enlarge)
Because of the digital control problems, methods of multi-line addressing are inherently analog at the pixel level. Image data is still manipulated digitally in processors where methods of image decomposition are used to break an image into row and column data, which are then converted to analog signals by digital/analog converters (DACs). The analog row and column signals are basically OFDM carriers, where each frequency component in the row and column signals represents the control of a single pixel in the display.
The current POLED displays that implement multi-line addressing (and work in any active-matrix display without using Walsh functions such as in active addressing used for only passive LCDs) was initially described in Patent 5644340 filed in 1995 (Reference 2). In this method, each column signal in the display is a separate reference frequency (the same as a local oscillator) and each row is a linear combination of all the column reference frequencies with a given amplitude.
The intersection of each row and column signal then maps the frequency control of each pixel (the same frequency exists on each column, but is different on each row). Each pixel contains a simple demodulator circuit, which demodulates the incoming row and column signals to produce a signal-amplitude that controls the brightness of the pixel (Figure 2). In this way, all pixels can be controlled simultaneously with varying brightness.
Figure 2: Pixel cell architecture
(Click on image to enlarge)
Each pixel has exactly the same circuit: a demodulator for frequency discrimination of the row and column frequencies and a low-pass filter for producing a DC-amplitude control of the pixel. The frequency discrimination and low-pass filter characteristics in Figure 2 determine how close row and column frequencies can be spaced and what the highest frequency is required for a given display resolution.
As seen in Figure 3, a 1920×1080 HDTV display can be realized with a maximum line frequency of 385 kHz, assuming 200 Hz frequency discrimination. The frequency discrimination and frame rate of the display is controlled by the cut-off frequency of the low-pass filter at each pixel in Figure 2. The same maximum frequency of 385 kHz drives each line at the same time, reducing the need for a much faster line-by-line clock. With the low-frequency requirements of the display in Figure 3, power consumption is reduced for the same pixel brightness when compared to a display using a single, high frequency dot-clock.
Figure 3: Maximum frequency for HDTV
(Click on image to enlarge)
In a flashback to the days of crystal radio, it has been found that the OLED diode in the passive OLED display can act as both a demodulator and low-pass filter of row and column signals (Reference 3) [Editor's note: if you are unfamiliar with the diode and the basic, passive crystal radio, which was the first mass-market "electronic" circuit, you need to do some basic research--and even build one!] With the anode connected to the row and cathode to the column (or reversed if the polarity of the signals is taken into account), the OLED demodulator produces the characteristic sum and difference frequencies which when appropriately filtered with the LPF generated the intended DC control of the pixel. A thin-film transistor in an AMOLED display works just as well if not better as a demodulator when properly biased (with the source connected to a column signal and the gate to a row signal, for instance).
With the price of active matrix OLED displays (AMOLED) dropping rapidly, the advantage of multi-line addressing in OLED displays may seemed short-lived, but even AMOLEDs may be able to benefit from the reduced frequency and power requirements of multi-line addressing. The bigger advantages of multi-line addressing may come in the bandwidth savings in driving data to a display, as the lower pixel frequencies allow more bandwidth for increased frame rate based on the fastest OLED response time. Also, larger display resolutions such as UXGA can be developed that will run at high frame rates without taxing the OLED pixel response. With high-resolution and high-bandwidth display applications on the horizon, architectures that utilize multi-line addressing are likely to be considered.
References- Cambridge Display Technology Press Release
- "Frequency Mixing for Controlling Individual Pixels in a Display,"
- "Transient response of passive matrix polymer LED displays,"
About the author
Michael Harney is an Electrical Engineer working in industrial and vehicle electronics. He is the holder of four patents and has a Bachelor of Science in Electrical Engineering from Utah State University
---------------------------------------------------------------------------------
OLED Display Technology and Capabilities6 October 2009
The organic light emitting diode (OLED) display is becoming more and more popular, especially for mobile phones, media player and small entry level TVs. Contrary to a standard liquid crystal display, the OLED pixel is driven by a current source. To understand how and why the OLED power supply impacts the display picture quality, it is key to understand the OLED display technology and power supply requirements. This article explains the latest OLED display technology and discusses the main power supply requirements and solutions. A novel power supply architecture tailored to the OLED power supply requirements is also presented here.
Market environment
All major mobile phone companies by now offer one or more models featuring an OLED display. Sony has the first OLED TV in mass production and many other companies show first prototypes. The OLED display offers wide color gamut, contrast ratio, viewing angle and fast response time. This makes the display ideal for multimedia applications. The self-emitting OLED technology doesn’t require a backlight and the power consumption depends on the display content. Power consumption can be much lower compared to a LCD using backlight. With a larger panel size the superior image quality of an OLED becomes more noticeable. Therefore, more and more OLED panels being used have a display size >3” and the ultimate application in the future still might be the TV panel. Another market for the OLED display is certainly the flexible display. Currently, the OLED and electrophoretic display technology look most promising. The electrophoretic or bi-stable display being used for electronic reader applications needs to be improved in color quality. On the other hand, currently OLED display is not ready for mass production when using fully-flexible materials. This depends mainly on the backplane technology.
Backplane technology enables flexible displays
High-resolution color active matrix organic light emitting diode (AMOLED) displays require an active matrix backplane using an active switch to turn each pixel on and off. The liquid crystal (LC) display amorphous silicon process is mature and provides a low-cost active matrix backplane, and used for OLEDs as well. For flexible displays companies are working with an organic thin film transistor (OTFT) backplane process. This process also can be used for an OLED display to realize flexible, full color displays. Whether a standard or flexible OLED is being used the same power supply and driving mythology needs to be applied. To understand the OLED technology, capabilities and its interaction with the power supply, a closer look into this technology is given. The OLED display itself is a self-emitting display technology and doesn’t require any backlight. The material for the OLED belongs to the category of organic materials due to its chemical structure.
OLED technology requires a current control driving method
A simplified circuit, representing one pixel, is shown in Figure 1. The OLED has electrical characteristics very similar to a standard light emitting diode (LED) where brightness depends on the LED current. To turn the OLED on and off and to control the OLED current a control circuit, thin film transistors (TFTs) are being used.

In Figure 1, transistor T2 is the pixel control transistor turning each pixel on and off. This is similar to any other active matrix liquid crystal display topology. A T1 is used as a current source, and the current is given by its gate source voltage. The storage capacitor is Cs, which holds the gate voltage of T1 stable and clamps the current until the pixel is addressed again. The simple single transistor current source in Figure 1 has a major cost advantage since only two transistors are required. The disadvantage of the simple circuit is a variation in current depending on process variations and voltage variation of Vdd. The OLED power supply circuit usually provides two voltage rails: Vdd and Vss. The voltage rail, Vdd, needs to have very tight regulation to achieve best picture quality and to avoid image flicker. The voltage regulation accuracy of Vss, which usually is a negative voltage, can be less accurate since it has a minor effect on the LED current. The effect of voltage fluctuations on Vdd to the OLED display is shown in Figure 2.

As the voltage supply Vdd changes, OLED brightness changes as well. Any superimposed voltage ripple on Vdd, can cause horizontal bars on the image due to different brightness levels. Depending on the display, a voltage ripple larger than 20mV already can cause such a phenomena. The visibility of the horizontal bars depends on amplitude and frequency of the superimposed voltage ripple. As soon as the frequency interferes with the frame frequency the bars appear. Under a normal laboratory environment the superimposed voltage ripple on Vdd is usually smaller than 20mV. The problem appears as the display and power supply are integrated into a system. As soon as any sub-circuit in the system draws pulsating current from the system power supply a voltage ripple appears, common to all circuits connected to the system power supply. Typical sub-circuits drawing pulsating current are the GSM power amplifier in a mobile phone, motor driver, audio power amplifier or similar. In such systems, the system supply rail has a superimposed voltage ripple. If the AMOLED power supply doesn’t reject this ripple, it will appear on its output as well causing the discussed visible image distortion. To avoid this, the AMOLED power supply needs to have a very high-power supply rejection ration and line transient response.
For the AMOLED power supply a boost converter is required for the positive voltage rail, Vdd, and a buck-boost or inverter for the negative voltage rail, Vss. This puts the challenge to the IC manufacturer providing a suitable power supply IC providing a very accurate positive voltage rail, Vdd, and negative voltage rail, Vss, achieving minimum component height and smallest solution size.
To meet all these requirements a novel power supply topology is chosen to provide both positive and negative output voltage rails from a Lithium-Ion (Li-Ion) battery using just a single inductor.
SIMO regulator technology enables best-in-class picture quality
Figure 3 shows the typical application circuit using the TPS65136, a device with single-inductor multiple-output (SIMO) regulator technology. The device operates with a four-switch buck-boost converter topology. SIMO technology features best-in-class line transient regulation, buck-boost mode for both outputs and highest efficiency over the entire load current range.
Advanced power save mode enables highest efficiency
As with any battery-powered equipment, long battery standby time is only achieved when the converter operates at highest efficiency over the entire load current range. This is especially important for an OLED display. The OLED display consumes its maximum power when the display is fully white, and much lower current for any other display color. This is because only the white color requires all the sub-pixels red, green and blue to be fully turned on. For example, a 2.7 inch display requires 80mA current for a fully white picture and only 5mA current when icons or graphics are displayed. Therefore, the OLED power supply needs to provide high converter efficiency at all load currents. This is achieved by using an advanced power save mode technology reducing the converter switching frequency as the load current decreases. Since this is done using a voltage controlled oscillator (VCO), possible EMI problems are minimized and the minimum switching frequency is controlled to be outside the audio range of typically at 40kHz. This avoids possible audible noise caused by ceramic input or output capacitors. This is especially important when using the device in a mobile phone application and simplifies the design process.
Conclusion
Since OLED display technology is just emerging, there is still a lot of room to conserve power, increase OLED efficiency and minimize the total solution size. As OLED becomes more mature, it is also possible to use OLED for architectural lighting or as backlight for LC Displays. Both opportunities allow lower power consumption and higher design flexibility compared to traditional lighting solutions. For OLED technology, the future seems to be very bright.
References
To download a datasheet on the TPS65136, visit:
www.ti.com/tps65136-ca.
To learn more about this and other power solutions from TI, visit:
www.ti.com/power-ca.
Author
Oliver Nachbaur is a member of the Technical Staff at Texas Instrument in Germany where he is a System Engineering Manager for the Display Power Converter group. Oliver has over a decade of experience in the semiconductor industry working as an Applications Engineer and System Engineer on Power Management Products. Oliver received a degree in Electrical Engineering in Ravensburg, Germany. He can be reached at:
ti_onachbaur@list.ti.com.