How does MLA work?


Introduction

Passive matrix supertwisted nematic (STN) LC displays are commonly used for medium information content displays such as in mobile phones and PDAs because good image quality and low power consumption can be obtained with the low manufacturing costs inherent in passive matrix displays. While contrast and viewing angle of STN displays have improved dramatically over the past few years, STN response times still remain sluggish, just fast enough to follow cursor movement, and much too slow for video.

Conventional line-at-a-time STN addressing is adequate for slow-responding LCDs used in word processors but causes reduced brightness and contrast ratio in STN panels fast enough for cursor motion in notebook applications.   Conventional addressing completely breaks down for STN panels fast enough to show video images.

The cause of the speed problem is not the STN display itself, but rather the conventional multiplexed addressing method used to drive passive matrix LCDs. By designing  STN displays for increased speed and using a new drive method called Active Addressing™, it is possible to obtain bright, high-contrast, full-color, full-motion video images. The image quality is comparable to active substrate, thin film transistor (TFT) displays which have much higher manufacturing costs.


The problem

With conventional LCD addressing the matrix rows are sequentially selected with a 20 V pulse, taking about 1/60th of a second (16.7 ms) to scan (select) all the rows, one at a time during a frame period. The column drivers supply the column signals in synchronization with data on the selected rows in order to obtain the desired information pattern on the display. Conventional LCD drive works well with the usual STN liquid crystal displays whose inherent response to a voltage change is several hundred milliseconds and thus many times longer than a typical frame period of 16.7 ms. Under these conditions the liquid crystal integrates over a frame period and the transmission is determined by the root- mean-square (rms) average voltage applied across the pixel; the result is a bright, high-contrast image.

 

The "on" pixel waveform of a standard addressed LCD drives the pixel to a constant "white" state (blue line) in a slow responding LCD, but in a fast LCD (red curve) the rapid decay of transmitted light intensity after the select pulse results in less perceived brightness. 

In order to show full-motion video images, the inherent STN response must be reduced below 50 ms. The STN response time can be readily brought into this range by decreasing the cell gap to about 4.5 microns and filling the display with low-viscosity liquid crystal mixtures. However, in this case the response time of the LCD is not much longer than the frame period and the current row-at-a-time addressing method breaks down; after each select pulse, the liquid crystal significantly decays before the next one refreshes it, causing the display to fade. The display no longer responds to the rms voltage averaged over a cycle time, but rather to voltage changes occurring within the frame time (red curve in figure). This phenomenon has been referred to as 'frame response'. The resulting decreased display brightness, poor contrast ratio and increased flicker have prevented the use of these fast cells.

Attempts have been made to decrease frame response by altering the device and drive parameters. Altering device parameters such as twist angle, pretilt angle, thickness/pitch ratio and gap thickness has been shown to be ineffective. Increasing the vertical refresh rate of the multiplex drive has the desired effect of decreasing the time between selection pulses, but has the drawback of shifting the energy of the drive signals to higher frequencies where it is strongly attenuated by the low- pass filtering action of the panel. The result is increased 'ghosting' manifested by vertical trails under written characters and increased power consumption. Another approach is to alter the relative gain of the row and column signals, the bias ratio, but this has the disadvantage of decreasing the selection ratio from its optimum value.


The solution

Because frame response is so clearly related to the pulsed nature of standard LCD addressing waveforms, making the pixel waveforms more uniform should result in more rms-like behavior for fast-responding LCDs. Active Addressing drive achieves this behavior by introducing multiple selection pulses that are distributed throughout the frame. With more selection pulses per frame, the amplitude of each selection pulse is lower, and the tendency for frame response is decreased without sacrificing the optimum selection ratio. Furthermore, by distributing the multiple selection pulses over the frame, the liquid crystal director does not have sufficient time to decay between the selection pulses, allowing it to properly react to the rms voltage of the drive waveform for optimum brightness and contrast. This scheme is generally referred to as Multiple Line Addressing (MLA).

 

wpe2.jpg (12176 bytes)

With MLA even a fast responding STN display responds to the rms value of the drive waveform.

 

Reduced power consumption, too

By virtue of the lower voltage selection pulses, MLA driven displays dissipate considerably less power than conventionally driven STN displays. The power consumption of a 4-inch diagonal CGA panel (640x200), for example, has been reduced by a factor of eight through a combination of improved power supply efficiency, lower operating voltage of the row and column drivers and the ability to drive at lower frame frequencies without observable flicker.   Reducing display power has recently become extremely important in digital mobile phone applications.  The example in the table below compares the power usage in a Personal Digital Assistant (PDA) having a 160x160 STN display addressed with standard drive and MLA drive. Driver chips for these applications are commercially available.

 
  Standard drive MLA drive
PDA without display 50 mW 50 mW
160x160 STN display 17 mW 2.3 mW
Total 67 mW 52.3 mW
Power budget used by display 25.4% 4.4%