BACKGROUND OF THE INVENTION A continuing objective in the field of electronics is the miniaturization of electronic devices. Most electronic devices include an electronic display. As a result, the miniaturization of electronic displays is critical to the production of a wide variety of compact electronic devices. For example, as electronic devices such as personal digital assistants, cell phones, digital still cameras, DVD players and internet appliances become ever smaller and more portable, the demands on the electronic displays for these products must meet difficult and seemingly contradictory requirements. On the one hand, the displays must provide increasing amounts of high quality visual information, sometimes approaching that of a desktop monitor.

Yet these displays must still be very compact and lightweight, consume little power, and be produced at low cost. Until recently, displays were not able to meet all of these requirements.

The purpose of an electronic display is to provide the eye with a visual image of certain information. This image may be provided by constructing an image plane composed of an array of picture elements (or pixels) which are independently controlled as to the color and intensity of the light emanating from each pixel. The electronic display is generally distinguished by the characteristic that an electronic

signal is transmitted to each pixel to control the light characteristics which determine the pattern of light from the pixel array which forms the image.

Two examples of electronic displays are the cathode ray tube (CRT) and the active- matrix liquid crystal display (AMLCD). There are other electronic displays, but none are so well developed as the CRT and AMLCD which are used extensively in computer monitors, televisions, and electronic instrument panels. The CRT is an emissive display in which light is created through an electron beam exciting a phosphor which in turn emits light visible to the eye. Electric fields are used to scan the electron beam in a raster fashion over the array of pixels formed by the phosphors on the face plate of the electron tube. The intensity of the electron beam is varied in an analog (continuous) fashion as the beam is swept across the image plane, thus creating the pattern of light intensity which forms the visible image. In a color CRT, three electron beams are simultaneously scanned to independently excite three different color phosphors respectively which are grouped into a triad at each pixel location. However, the CRT is impractical for use in a microdisplay.

In contrast to the emissive type displays such as the CRT, an AMLCD display utilizes a lamp to uniformly illuminate the image plane which is formed by a thin layer of liquid crystal material laminated between two transparent conductive surfaces which are comprised of a pattern of individual capacitors to create the pixel array. The intensity of the illumination light transmitted through each pixel is controlled by the voltage across the capacitor, which is in turn controlled by an active transistor circuit connected to each pixel. This matrix of transistors (the active matrix) distinguish the AMLCD from the passive matrix liquid crystal devices which are strictly an array of conductors controlled by transistors external to the image area usually in the periphery of the matrix. The ability of each transistor to control the characteristics of just one pixel allows for the higher performance found in AMLCD displays in contrast to the passive arrays. However, a drawback of the AMLCD display is the high power consumption incident to the illumination.

While some electronic products which contain an electronic display have memory for storing the data which is to be displayed, some do not. For instance, a television must activate the CRT display in real time as the broadcast signal is received unless a VCR or similar storage medium is employed. In computers, data is transmitted and stored digitally. Moreover, in portable electronics devices, size and power constraints require the use of semiconductor memory which stores data only in digital format. In digital electronic products, it is typical that a display controller is incorporated to receive and store the bit mapped image to be displayed and then to transfer that data to the display in a series of image frames at a rate high enough to look smooth to the eye. The semiconductor memory storing the image bits is called the frame buffer, and the rate at which the data is refreshed on the display is called the frame rate.

It is an advantage in many applications to display large amounts of information requiring more and more resolution in the display. High resolution displays may contain hundreds of thousands of pixels. As an example, the Super VGA (SVGA) display resolution consists of 480,000 pixels. With a simple monochrome image and no grayscale, the frame storage is only equal to the approximately one-half megabit frame size. However, were the image to be full 24 bit depth color (i. e., 3 colors and 8 bits of grayscale per color), the frame storage would approach 12 megabits. At the frame rates which are common today for high performance displays, at least 60 frames per second and up to 85 frames per second, as many as one gigabits per second must be transferred from the frame buffer to the display.

The state of semiconductor technology at present limits clock speeds to a level well below such transfer rates and parallel interfaces of 16 to 32 bit widths are typical in high performance displays.

It is a characteristic of analog displays that when the image data is stored in semiconductors, the digital information is converted to analog in a digital-to-analog converter (DAC) at the interface of the display. The digital representation of a pixel at the high standard of 8 bits of grayscale allows the creation of 256 separate shades

per color (16 million distinct colors). In high performance displays, multiple DAC channels are required to provide the bandwidth of data transfer required.

In the particular case of miniaturization of high resolution electronic displays, there is an advantage to reducing the size of the pixels which comprise the display. The need for such small devices has led to the development of a category of miniature displays often described as microdisplays with pixel sizes as small as 10 microns or less. In order to achieve this pixel resolution, active matrix devices have been developed utilizing silicon wafer fabrication of CMOS devices as opposed to thin- film transistors fabricated on a glass or quartz substrate. Single crystal silicon design rules are many times smaller than poly-silicon resulting in transistor sizes to easily fit microdisplay geometries. With the exception of techniques to separate the single crystal transistors from the silicon substrate utilizing lift-off technology, CMOS based active matrix displays are inherently opaque, and therefore must be reflective rather than transmissive like the poly-silicon devices. Thin film transistor (TFT) based transmissive devices are also opaque as transistors and interconnection lines, and optical efficiencies are very low for high resolution TFT displays.

The pixel sizes of microdisplays are too small for the resulting image to be directly viewed by the unaided eye, but can be magnified through projection optics to create a real image on a screen or wall or through a magnifier to create a virtual image in space. In practice, pixel sizes are limited today by magnifier and illumination considerations to geometries which are larger than single crystal silicon transistors, and in particular, useful pixels are even larger than multi-transistor SRAM cells.

This method can produce extremely compact, power efficient, and low cost displays that present high levels of information to the viewer, comparable to that of desktop computer monitors. Directly viewed displays cannot meet all of these requirements since a display that is viewed directly and has information content similar to a desktop (at least 640x480 viewable picture elements), must be as large as the displays typically found on ultra compact notebook computers (over 8"diagonal).

However, near-to-the-eye displays can be produced with very small overall

dimensions by using a magnifying optical system to create a virtual image at some distance in front of the viewer. An illumination system may be provided as part of the optical system for devices such as liquid crystal displays where light is not generated by the material. These magnified images can appear to be as large as a desktop monitor even though the display dimensions are one or two inches across.

Such small dimensions require that only a single electro-optic device be employed in the display. All of the image colors must be provided by this single device. (Larger systems, such as front projectors, can use multiple devices, one for each color.) There are two methods commonly used to generate color using a single electro-optic device. In the first, each picture element (pixel) is divided into three or more sub- pixels and a color filter, typically red, green and blue, is placed in the light path from each sub-pixel. The eye merges these sub-pixels to create a color image. This method suffers from significant light loss in the color filters, requiring up to four times as much power to be supplied to the illumination system. The color filters also add significant additional cost to the display. The second method avoids the high power requirement and added cost of the sub-pixel/color filter method. Instead, a single pixel is used for red green and blue images in a sequential manner.

The pixel sizes are also small relative to the size of color filters used in TFT AMLCD displays to create color triads for each pixel. There is a significant advantage to creating color through the sequential use of the entire array to create an image specific to each of the three prime color components. Through the utilization of separate light emitting diodes (LEDs) of each prime color to illuminate the display, the diodes can be turned rapidly on and off to correspond to the particular color component being displayed by the array at that moment. This method of color creation is called field sequential color wherein each color field is sequentially illuminated by the appropriate diode. Because at least three different color field images need to be displayed at a rate faster than can be resolved by the eye, the field sequential color method at least triples the data transfer rate required as compared to a monochrome display.

A need exists for a microdisplay system which can overcome the various above- described limitations of prior art display systems and be able to produce a high resolution color image while having a low vertical height for non-immersive viewing.

Microdisplays are commonly worn on, or positioned close to, the eyes of the user. If the user requires optical correction (i. e., wears corrective lenses), the close proximity of the display may make it impossible or uncomfortable for the wearer to wear glasses while using the microdisplay.

A need exists for a way to correct a user's vision when the user is utilizing a microdisplay.

These and other advantages are provided by the display system of the present invention.

SUMMARY OF THE INVENTION A microdisplay system is provided according to one embodiment of the present invention. The system includes headwear that is adapted for wearing on a head of a user. Types of headwear includes eyeglass-like devices, goggles, helmets, visors, etc. A display is detachably or permanently coupled to the headwear. One or more corrective lenses are detachably or permanently coupled to the headwear and positioned between the display panel and the head of the user.

Preferably, the corrective lens carries an optical corrective prescription of the user.

In one aspect of the present invention, a surrounding visual environment is visible to the user. Here, the corrective lens provides simultaneous refractive correction for the display and the surrounding visual environment. In another aspect of the present invention, the display is imaged at a distance from the eyes for enabling use of a refractive correction power of the user for a distance greater than the actual distance between the user and the display. In yet another aspect, the corrective lens provides different refractive corrections for viewing the display and for viewing the surrounding visual environment. One implementation of this uses a bifocal lens. In a further aspect of the present invention, two corrective lenses provide disparity- driven depth perception In one embodiment of the present invention, the corrective lens is detachably coupled to the headwear. Preferably, two corrective lenses are provided and are separated such that the lenses substantially match the individual separation of the eyes of the user.

Preferably, the corrective lens corrects myopia, hyperopia, astigmatism, presbyopia, accommodative disfunction, and/or oculomotor imbalances. To do so, the corrective lens has a prescribed optical property such as spherical refractive power, cylindrical refractive power, near addition power, and/or prism refractive power.

In an embodiment of the present invention, the display has a vertical extent of less than about 40 mm which provides"look over"and"look under"capabilities as well as allows for integration of the display panel into more versatile and aesthetic headwear. Vertical extent as used here is taken with respect to the normal viewing angle of a user standing upright looking straight ahead.

According to an embodiment of the present invention, a corrective lens device is provided for coupling to a microdisplay adapted for wearing near eyes of a user.

The device includes a pair of corrective lenses that are spaced laterally and that each have an optical corrective prescription of the user. A mounting portion is operably coupled to the lenses for detachably or permanently attaching the lenses to the microdisplay. Preferably, the lateral spacing of the corrective lenses substantially matches the individual separation of the eyes of the user.

In an embodiment of the present invention, the surrounding visual environment is visible to the user, and the corrective lenses provide simultaneous refractive correction for the display and the surrounding visual environment. Alternatively, the corrective lenses can provide different refractive corrections for viewing the display and for viewing the surrounding visual environment.

Preferably, the corrective lenses correct myopia, hyperopia, astigmatism, presbyopia, accommodative disfunction, and/or oculomotor imbalances. To do so, the corrective lenses have a prescribed optical property such as spherical refractive power, cylindrical refractive power, near addition power, and/or prism refractive power.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 depicts an illustrative microdisplay system according to a preferred embodiment of the present invention; Figure 2 is a top view of the illustrative embodiment of Figure 1; Figure 3 is a perspective view of a corrective lens device according to an embodiment of the present invention; Figure 4 is a block diagram of a display system according to a preferred embodiment of the present invention; Figure 5 is a side cross sectional view of a Liquid Crystal Module (LCM); Figure 6 is a chart illustrating a timing of displaying color fields to a viewer; Figure 7 depicts four ways the orientation of a pixel array can be configured; Figure 8 is a grid illustrating an address relative to pixel position for a backplane; Figure 9 illustrates a configuration write and read transaction waveform ; Figure 10 is a timing diagram depicting an exemplary waveform of a block transfer of two rows of six words each; Figure 11 is a timing diagram showing a demonstration of a waveform of a block read of 6 words;

Figure 12 is a block diagram of a backplane integrated circuit according to one embodiment of the present invention; Figure 13 illustrates how patterns are loaded into an array with and without a rotate pattern bit set; Figure 14 shows pixel arrays that demonstrate how data is moved relative to scroll direction; Figure 15 is a block diagram that shows the system components of an embodiment of a microdisplay according to an embodiment of the present invention; Figure 16 is a flow chart showing how each byte of image data is processed through the palette, adjusted by the"grid", and separated into individual bit planes; Figure 17 is a flow diagram depicting a process to convert 8-bit pixel data into pixels in a format amenable to the display system; Figure 18 is an illustration of an Analog Controller Chip (AIC); Figure 19 illustrates a transaction waveform during parallel write timing; Figure 20 illustrates a transaction waveform during parallel read timing; Figure 21 depicts an ITO voltage generation waveform; Figure 22 depicts an LED current generation waveform ; Figure 23 illustrates an LED timing waveform where itorst = 0; Figure 24 depicts a waveform for ITO and LED Timing with ito-rst = 1; and Figure 25 is a block diagram of an analog controller according to an embodiment of the present invention

DESCRIPTION OF THE PREFERRED EMBODIMENTS Microdisplay System Figure 1 depicts an illustrative microdisplay system 100 according to a preferred embodiment of the present invention. In the embodiment shown in Figure 1, the headwear resembles a pair of glasses. Similar to a pair of eyeglasses, the device comprises a mounting portion that incorporates displays, electronics and optics, and two temple pieces (shown in Figure 2) to help support the device over the ears. In other embodiments of the present invention, such headwear can include other eyeglass-like devices, goggles, helmets, visors, or any other item amenable to wearing on the head of the wearer.

In more detail, the microdisplay system includes a pair of displays 102,104 detachably or permanently mounted to the display mounting portion 106 of the head-borne device, i. e., headwear 108. The preferred type of displays are LCD displays.

The virtual computer display occupies only a portion of the total visual space of the wearer. The remainder of the wearer's visual space is not occupied by the device and enables the wearer to see their surrounding visual environment. This system is "non-immersive"because a portion of the real visual environment is visible to the wearer.

In more detail, each of the displays has opposite top and bottom edges 110,112 which define a vertical extent of each panel. The vertical extent is preferably less than about 40 mm, and ideally less than about 37 mm. These dimensions provide "look over"and"look under"capabilities as well as allow for integration of the display into more versatile and aesthetic headwear. The display is capable of

displaying an image at a resolution of at least 640x480 pixels to create desktop-like viewing.

The present invention preferably enables the user to view a virtual computer display that appears to be imaged at a designed distance from the user (e. g. 6 feet from the face of the user).

Figure 2 is a top view of the embodiment of the present invention of Figure 1. As shown, the headwear includes a pair of temple pieces 202 extending therefrom. As an option, a pair of spring hinges 204 can be used to couple the temple pieces to the display mounting portion of the headwear. The spring hinges cause the temple pieces to exert a constant clamping force on the head of the user to assist in securing the headwear to the head of the user. Preferably, soft conforming pads 206 are attached to the ends of the temple pieces to help grip the head.

One or more ear buds 208 may be attached to the headwear for producing audio.

Preferably, the ear buds snap into the end tips of the temple pieces from below. The audio wires (not shown) that carry the audio signal to the ear buds can be routed directly into recesses in the temple pieces.

Wiring 210 that is coupled to the display can be routed over the temple pieces so that it acts as a cantilever for reducing an effective weight of the headwear on a nose of the user, making the system feel lighter.

An adjustable nosepiece 212 can also be coupled to the headwear to assist in supporting the headwear and/or to provide greater comfort. Such adjustment can be vertical. Another such adjustment can be the width of the nosepiece. Preferably, the portion of the nose piece that contacts the skin is constructed of a soft, slip resistant material and has a large surface area to distribute the weight of the display system across a larger surface area of the nose.

As an option, an outer shield 214 can be positioned on an opposite side of the display with respect to the user. The outer shield is opaque with a partially reflective coating for producing an appearance of depth, thereby disguising the display system as a pair of sunglasses.

The head borne device that displays the virtual images fits close to the face of the user and may be able to be worn over the person's eyeglasses. Referring again to Figure 1, an optical lens device 114 that has two optical lenses (or, one lens for a monocular device) can be coupled to the headwear so that the user doesn't need to wear glasses when using the device.

Figure 3 is a perspective view of a corrective lens device according to an embodiment of the present invention. It should be noted that this device is presented for purposes of illustration only and should not in any way limit the scope of the invention. Further, the present invention will discuss the corrective lens device with reference to the illustrative display system of Figures 1 and 2, but it should be understood that the device can be utilized with display types other than those presented here, including non-head borne display devices.

Optical Corrective Device Referring again to Figure 3, the device includes a pair of corrective lenses 302,304 that are spaced laterally and that each have an individual refractive correction based on the optical corrective prescription of the user. A mounting portion 306 is operably coupled to the lenses for detachably or permanently attaching the lenses to the display system. As shown here, the mounting portion includes two flexible members that are inserted into mounting holes 116 of the headwear, where they are held in place by friction and, preferably, serrations. (See Figure 1.) One skilled in the art will understand the mechanics of this and other types of mountings that may be used. Guide members 308 can be used to stabilize the device. If the optical

corrective device is removable, multiple users are able to share a single electronic display.

In an embodiment of the present invention, the surrounding visual environment is visible to the user, and the corrective lenses provide simultaneous refractive correction for the display and the surrounding visual environment.

The visual display can be imaged at a long distance from the eyes, enabling a person's usual refractive correction power for long distances to be used. No focus of the instrument for nearer image distances is required or possible.

Different refractive corrections for the visual display and for surrounding visual space can be provided when desired to meet the refractive needs of the user. (The specific implementation is that a bifocal lens can be placed in the device, allowing distance visual correction for the virtual image and near visual correction for the surrounding visual space.) Other embodiments of the present invention include correction for one eye (monocular), correction for 2 eyes (binocular) or correction for 2 eyes with disparity-driven depth perception (stereo).

Various embodiments of the present invention allow for all common optical refractive corrections. This includes correction for conditions such as myopia, hyperopia, astigmatism, presbyopia, accommodative dysfunction and oculomotor imbalances. The lenses in the corrective device provide correction for these conditions by having prescribed optical properties of spherical refractive power, cylindrical refractive power, near addition power, and/or prism refractive power.

In one embodiment of the present invention, the lenses are located with respect to one another so they are of appropriate lateral separation to match the measured individual separation of the eyes (inter-pupillary distance or IPD) of the user. This

avoids"prism error"and the associated discomfort from conflicting visual stimuli.

(Eye lens focus distance is different from eye rotational convergence distance.) The optical correction system of the present invention, which comprises the lenses and their holder, are preferably designed to utilize the standard operating procedures of the eye care community and the ophthalmic correction industry. This includes the following standard operating procedures and products: the optical prescription normally written by an optometrist or ophthalmologist, the IPD measurement and specification, commonly used lens materials, commonly used lens fabrication procedures. The user can obtain lenses of appropriate power for the device from their usual and customary source.

Technotosv Overview The preferred microdisplay system of the present invention is a compact, low-power, high-resolution display system designed for mobile applications such as cameras, head-mounted displays, and portable Internet devices. Unlike traditional liquid crystal display panels, it is viewed near to the eye, like the viewfinder of a camera.

This near-to-eye viewing mode allows for the small size and power efficiency of the design.

Microdisplay Module and Support Components The microdisplay is designed to operate in two basic modes distinguished by the number of distinct colors required.

The most power efficient is an eight-color mode which is appropriate for viewing email messages and simple graphics such as topographic charts. This mode offers the benefit of low power consumption and minimum total component count. It is referenced below as Power Miser Mode with a total power requirement under 100mW.

Also supported is a high color mode that provides the equivalent color of an 18-bit LCD panel. In addition to the higher color depth, this mode offers the benefit of being easier to design into a system, and to program. This mode is referred to as Color Rich Mode throughout the remainder of the discussion.

A Color Rich Mode implementation according to the present invention provides the most functionality. This section will introduce the technology by describing a typical Color Rich implementation.

Figure 4 is a block diagram of a display system 400 according to a preferred embodiment of the present invention. A typical Microdisplay Color Rich implementation consists of the following.

A self-contained display module that contains Liquid Crystal Module Illumination 402 Optics Two support integrated circuits Color Rich Display Controller ASIC (CRASIC) 404 Auxiliary Integrated Circuit (AIC) A frame buffer SDRAM 406 Miscellaneous passive components Liquid Crystal Module Figure 5 is a side cross sectional view of a Liquid Crystal Module 500. The Liquid Crystal Module (LCM) is the primary image producer of the system. It is an 1 lmm diagonal, 800 column by 600 row, black and white LCD. The LCM is produced by covering an integrated circuit Backplane die 502 with a thin layer of Liquid Crystal material 504 and a cover glass 506 coated with Indium Tin Oxide (ITO) to form a common electrode. This type of display is called Liquid Crystal on Silicon (LCOS).

The IC Backplane is a standard 3.3V CMOS device using 0.35 micron design rules.

In essence it is an 800x600x3 bit Static RAM device (SRAM), with proprietary embedded timing and control logic. The top metal layer contains an array of 800 by 600 squares, each 11 microns on a side. These aluminum squares are highly reflective and act as mirrors. The liquid crystal material directly above each mirror will allow light to pass through depending on the electric field between the metal mirror and the ITO electrode coating on the cover glass. This effect enables the 480,000 pixels on the backplane to act as individual light valves. The LCM is a postage stamp size liquid crystal panel capable of displaying 2300 dpi resolution images.

Illumination and Field Sequential Color The LCM does not produce light: A separate light source must be provided. If a white light source were used the LCM would provide a black and white or gray scale display. The Microdisplay uses a triad of red, green, and blue Light Emitting Diodes (LED) to illuminate the LCM and a process called Field Sequential Color to display full color images.

A field sequential color device presents the image to the viewer as separate fields of Red, Green and Blue in rapid succession. Figure 6 illustrates how the fields are presented to the user. When this is done at a high repetition rate, the viewer's brain merges the fields to form a single full color image. This is the same phenomenon that causes 23 frames of still photographs to appear as 1 second of continuous motion when shown through a movie projector.

The amount of light produced by the LED triad is very little compared to the lamps used in projectors. It is, however, more than sufficient to produce a bright, clear image for the viewer because the display is held close to the eye, and ambient light does not interfere with the display.

Optics

As mentioned above, the LCM is an 1 lmm diagonal display with 11 micron pixels.

When the user holds the display near the eye however, the image appears to be a 110 cm diagonal picture located 2 meters from the viewer. This effect is achieved through optics which act like a compound microscope to magnify the image 13.5 times. The LCM is precisely attached to one face of the optics module and is held in place by means of a cradle. The Illumination triad is attached to a separate face of the optics, and is held in place in the same way. The LCM is attached to a flexible printed circuit, which provides the electrical interface to the display module.

Color Rich Display Controller The Color Rich Display Controller ASIC (CRASIC) is an IC, which controls the timing of the Backplane, and illumination to produce rich color images. More details of the CRASIC are provided below.

The CRASIC is designed to interface easily to 8,16, or 32-bit RISC Microprocessors, and hide the complexity generating images with the Backplane.

The chip uses a directly attached SDRAM to store a linear frame buffer representation of the screen, and an additional copy of the same information, separated into bit planes. The CRASIC creates these separate bit planes automatically as the CPU writes the linear frame data.

Additionally, the CRASIC provides an internal palette RAM which enables 8-bit color values to be expanded into 24-bit colors before being dithered and converted into the proper bit frame format for the Backplane.

The CRASIC includes an embedded RISC CPU that feeds data to the Backplane. An instruction set enables the system designer to precisely control the transfers to the Backplane. This instruction set also supports functions such as overlays for cursors and generic BITBLT operations. See the sections below on the CRASIC and CRISP instruction set for more detail.

Auxiliary IC The AIC chip is the third IC. It acts as a companion device for the Backplane by providing all the analog functions required to produce images.

The first major function it provides is current drive control for the illumination LEDs. The current level for each LED can be varied independently, allowing the color balance of the display subsystem to be software controlled. This is advantageous since the electro-optical characteristics of individual LEDs vary over the operating temperature range of the display system.

The AIC also controls the common ITO voltage. The Backplane is a 3.3v digital device. The pixels on the top metal layer are either 3.3v or ground. The ITO voltage is driven to a magnitude and offset which optimizes the E-field between the pixels and the ITO layer on the cover glass. The precise voltages are also temperature dependent, and may be controlled as appropriate through software.

Due to the tendency of liquid crystal ions to migrate in a static E-field, the field polarity must be frequently inverted so the LC material will not lose its optical effect. The AIC chip enables polarity reversal of ITO voltage supply to accomplish this important function.

The AIC chip also provides an internal temperature sensing function that enables software to determine the temperature of the AIC and the Backplane.

Summary The Backplane embodied within the Liquid Crystal Module provides the primary image for the display system. The AIC chip provides all of the analog voltages needed to drive the illumination LEDs and the common electrode ITO cover glass.

The Color Rich ASIC provides the primary system interface, and precisely controls the timing of the other two chips.

Backplane Technology A preferred backplane of the present invention is a high-speed, low-power integrated SVGA digital CMOS Backplane for use in a reflective silicon micro display such as the one described above. The backplane interfaces either with a microprocessor directly, or an external frame, transforming image data into a matrix of pixel electrodes (or pixels). This then, in conjunction with a common counter electrode, drives individual voltages across a liquid crystal material. When illuminated, light is reflected or absorbed at each pixel, which doubles as a mirror, according to those voltages. An optical image is observed when all of the pixels are viewed together.

Features 66 Mhz operating frequency sub lms Backplane refresh times 8 bit address bus Configurable registers 32 bit data bus DMA capabilities Block transfer capability Entire row transfers with zero wait states Interrupt Generation RGB and Duelchrome modes Automatic data inversion for LCM Pin Assignment Tables 1-4 set forth pin assignments and pin descriptions.

Table 1: Display Module Connector Pinout Pin Signal Pin Signal Pin Signal Pin Signal Name Name Name Name 1 Vddl 21 A6 41 D15 LED1 Red 2 Vdd2 22 D1 42 D3 LED2 COM 3 readyN 23 wrN 43 D31 LED3 Blue 4 clk 24 D18 44 D20 LED4 Green 5 rdceN 25 csN 45 D7 6 Gnd 26 D10 46 Vdd2 7 irqN 27 rstN 47 iotos 8 D16 28 Gnd 48 D12 9 AO 29 A7 49 led 10 D8 30 Vddl 50 D28 11 A1 31 D22 51 ito_glass 12 D24 32 D26 52 D4 13 A2 33 D14 53 Vddl 14 DO 34 D2 54 D21 15 A3 35 D30 55 blue 16 D17 36 D19 56 D13 17 A4 37 D6 57 green 18 D9 38 D11 58 D29 19 A5 39 D22 59 red 20 D25 40 D27 60 D5 Table 2: Backplane Pin Out by Pad Number Pin # Name Pin # Name Pin # Name Pin # Name 1 gnd0 20 D[16] 39 vdd3 58 red 2 vddO 21 D [8] 40 D [12] 59 green 3 readyN 22 D [24] 41 D [28] 60 blue 4 rdceN 23 D[0] 42 D [4] 61 vdd4 5 irqN 24 D [17] 43 D [21] 62 gnd4 6 A[0] 25 D [9] 44 D [13] 7 A[1] 26 D [25] 45 D [29] 8 A[2] 27 D[1] 46 D[5] 9 A [3] 28 D [18] 47 D [22] 10 A [4] 29 D[10] 48 D [14] 11 A [5] 30 gnd2 49 gnd3 12 A [6] 31 vdd2 50 D [30] 13 wrN 32 D [26] 51 D [6] 14 csN 33 D [2] 52 D [23] 15 rstN 34 D [19] 53 D [15] 16 A [7] 35 D [11] 54 D [31] 17 vddl 36 D [27] 55 D [7] 18 clk 37 D [3] 56 ito 19 gndl 38 D [20] 57 led Pin Description Table 3: Pin Description 1

Name Direction Description A [7: 0] Input System address. clk Input System clock. rstN Input System reset. wrN Input Access direction:"high"for reads,"low"for writes. csN Input Chip select. Active low. D [31 : 0] Bi-directional System data. irqN Output Interrupt request. Active low. rdceN Output Read data clock enable. Active low. readyN Output System ready. Active low.

Table 4: Pin Description 2 Name Direction Description A [7: 0] Input System address. blue Output Blue indicator for both ITO and LED control. Active high. clk Input System clock. csN Input System chip select. Active low. D [31: 0] Bidir System data bus. Valid bus widths are D [7: 0] (8 bit), D [15 : 0] (16 bit), and D [31: 0] (32 bit). Direction is controlled by wrN pin. When wrN is low, these pins are configured as inputs. green Output Green indicator for both ITO and LED control. Active high. gnd System ground. In connecting to the system, . gndO, gnd2 and gnd4 may be shorted together, as may gnd I and gnd4. Preferably, the system ground for gnd0 and gnd4 should be kept separate from the ground for gnd2 and gnd3. irqN Output Interrupt request. Active low. ito Output ITO voltage indicator. led Output LED On indicator. Active high. red Output Red indicator for both ITO and LED control. Active high. rdceN Output Read data clock enable. Active low. readyN Output System ready. Active low. rstN Input System reset. Active low. vdd System power. In connecting to the system, vddO, vdd2 and vdd4 may be shorted together, as may vddl and vdd3. Preferably, the system power for vddO, vdd2and vdd4 should be kept separate from the power for vddl and vdd3. wrN Input System write. Active low.

Data Orientation and Format Orientation The orientation of the pixel array can be configured one of four ways, as shown in Figure 7. Figure 7 shows the convention of Normal 700, Horizontal 702, Vertical 704, and Horizontal/Vertical 706.

Data Format Pixel data transferred to and from the backplane can be formatted in two ways: RGB and monochrome. RGB Data is formatted 4 bits per pixel, or 2 pixels per data byte.

Monochrome data is formatted 1 bit per pixel, or 8 pixels per data byte.

RGB In an RGB data byte, two bits in each byte are unused and are denoted'X'. The bits marked'R'are always written to, or read from, bit plane 0,'G'to plane 1, and'B'to plane 2. In RGB mode, data for 8 pixels can be packed into one 32-bit data word.

The backplane also supports double and single byte word lengths. For example, if the host system decides to write RGB data for only two pixels in one write cycle, then the backplane can be configured to look only at the first eight bits of the data bus for pixel data. The following table shows the relationship between the data bus, relative pixel number, and the different size transfers. Table 5: RGB Data for 8,16, and 32 bit Transfers Data 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 09 08 07 06 05 04 03 02 01 00 Bit 8 Bit------xBGR acBGR 16 Bit xBGR xBGR xBGR xBGR 32 Bit XBGR xBGR xBGR xBGR xBGR xBGR xBGR xBGR Pixel # 7 6 5 4 3 2 1 0

Monochrome In monochrome format, each bit of data gets mapped into one pixel. Each bit on the data bus gets mapped inside the Backplane into one of two programmable pixel colors, according to the value of the bit. With the bus configured to 32 bits, 32 pixels of data are present in one data transfer. This is four times the compaction of RGB format. The following table shows how each bit is mapped to a pixel, and the relative position on the data bus for all three sizes of transfers. Bit 0 of the first word is mapped to PO, or pixel 0, for example. Aside from supporting an extremely thin monochrome client, the monochrome format can be used on data reads to filter a color pixel array for a particular color pixel.

Table 6: Monochrome Data on a 16-Bit Data Bus Data D3 D3 D2 D2 D2 D2 D2 D2 D2 D2 D2 D2 D1 D1 D1 D1 Bus 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 8 - - - - - - - - - - - - - - - - 16 - - - - - - - - - - - - - - - - 32 P31 P30 P29 P28 P27 P26 P25 P24 P23 P22 P21 P20 P19 P18 P17 P16 Data D1 D1 D1 D1 D1 D1 D9 D8 D7 D6 D5 D4 D3 D2 D1 DO Bus 5 4 3 2 1 0 8--------P7 P6 P5 P4 P3 P2 PI PO 16 P15 P14 P13 P12 P11 P10 P9 P8 P7 P6 P5 P4 P3 P2 PI PO 32 P15 P14 P13 P12 P11 P10 P9 P8 P7 P6 P5 P4 P3 P2 PI PO

Backplane Addressing The byte address of a particular pixel on the Backplane IC is a concatenation of its row and column numbers, with its least significant bit (LSB) truncated. The LSB is removed because there are two pixels at each byte address. The column number ranges from 0 to 799, requiring 10 bits. The row number runs from 0 to 599, also requiring 10 bits. The following example illustrates how to calculate the byte address for a given pixel position. The pixel in row 234 and column 567 is given in the upper nibble of the byte as follows.

{234, 567} {10'b0011101010, 10'bl000110111} {10'b0011101010, 9'bl00011011}<BR> 19'b001 1101 0101 0001 1011<BR> OxlD51B Figure 8 illustrates the address 800 relative to pixel position for the Backplane.

System Bus Address Bus The Backplane is indirectly addressable from the external system. This means that all data transfer to the Backplane is accomplished through block moves, dma, or other register controlled operations. The Display System has 8 address bits. The internal configuration registers are accessed whenever the most significant address bit A [7] is driven high. The IC resumes block transfer mode when A [7] is low. The

remaining address bits, A [6: 0], are used for register addresses when A [7] is high, and ignored otherwise.

Data Bus The Backplane can be configured to look for 8,16, or 32 bit transfers. Writing to the data bus width register of the Display System configures valid data widths. After reset, the Display System's data bus width register is set to byte mode. This means that only D [7: 0] are valid.

Bus Protocol Bus Handshake The host system (CRASIC) initiates all transactions between itself and the Display System. To begin a cycle, the host issues a chip select (csN) and a write enable (weN) to the Display System. At the end of the cycle the host samples the readyN. If the readyN signal has not been selected by the Display System, the transaction must be restarted.

Address data only needs to be driven by the host system for a Display System register transaction. For Display System register accesses, address data is driven at the same time the csN and weN are driven. Otherwise, address data is ignored by the Display System and does not need to be generated by the host.

When weN is driven low, the Display System considers the transaction to be a write transaction. After driving csN and weN, the host checks for readyN assertion. If the readyN signal has been asserted, the host sends data. The readyN signal must remain asserted for each data word placed on the bus. If the readyN signal goes invalid during a-multi-word write cycle, the entire transaction must be restarted.

A valid read transaction occurs when the weN signal is left high following a csN assertion. If the readyN signal is selected immediately after the csN is selected, and

weN is high, then a valid read cycle has been started. The readyN signal will de- select, and some indeterminate amount of time later the Display System will put data on the bus. Data is valid with rdeN selected by the Display System.

Block Transfers Block transfers are a very important function of the Display System. They are important because they are one of only two ways (DMA is the other) that data can be written to, or read from, the Backplane. The block transfers enable all or part of a row of pixel data to be written to the Display System without wait states. A rectangular region of arbitrary shape can be transferred as a sequence of these rows, with only a 3-wait-state delay between rows.

Figure 9 illustrates a configuration write and read transaction waveform 900. Figure 10 is a timing diagram depicting an exemplary waveform 1000 of a block transfer of two rows of six words each. The address boundaries of the data are configured prior to the transfer. The values on the address bus are ignored during the transfer.

Block reads are accomplished in a similar manner. Figure 11 is a timing diagram showing a demonstration of a waveform 1100 of a block read of 6 words. Again, the address range for pixel data is set up prior to starting the block read cycle. Once the cycle begins, the Display System places data words on the bus an indeterminate time later. This fact should be of no great consequence. The Display System is capable of automatically reading data, inverting it and writing it back to the array very quickly.

It is only when the host system wants to verify data for system test purposes that a block read will be used. Therefore, the latency of a block read, as seen by the host system, is not important.

A final benefit to the Display System's block transfer capability is that block transfers can work in conjunction with the interrupt mechanism. The Display System can generate interrupts that indicate the end of a field or frame. These interrupts can then be used by the host system to start a new block transfer.

Even with block transfers, the time for loading the whole Backplane is considerable.

The following table shows the update duration as a function of the data bus widths possible on the Display System, and a sampling of clock frequencies.

Table 7: Full Backplane Update Times Bus Width # Clock Cycles Duration Duration @30MHz@60MHz 8 Bits 240K 8 ms 4 ms 16 Bits 120K 4 ms 2 ms 32 Bits 60K 2nE Backplane IC Configuration Figure 12 is a block diagram of the Backplane Integrated Circuit 1200. The components include a pixel and SRAM array 1202, a system interface 1204, a register data store 1206, and timers and counters.

Configuration Registers The configuration registers control the operation of the chip. They control everything from basic parameters like data bus width, to complex timing, to special operations such as scrolling. Accessing the configuration registers themselves, however, is simple and fixed, with no special control lines to drive. The configuration registers are grouped in 5 areas, System Interface, General Timing, Strobe Control, LCD Control, and DMA Control. Each of these groups will be discussed in the sections to follow.

The configuration registers are accessed whenever A [19] = 1. The address for a specific configuration is given by A [10: 4]. The values of A [18: 11] and A [3: 0] are ignored on a configuration access, so that a large number of aliases exist.

In the tables below, A [19] = 1 is assumed, and the registers are mapped according to the value of A [10: 4] only. Since A [3: 0] is ignored, new register addresses occur every sixteen bytes. By convention, the register addresses below occur in multiples of 10 (hex): 0, 10,20,30,40, etc. Because of aliasing, addresses OxOOO and OxOO1 point to the same register, although only address OxOOO is given in the tables.

Not always are all 8 bits defined for a particular configuration register in the tables below. Where a bit is undefined, it should be considered reserved, and a 0'written to it for compatibility with future chips. Reading an undefined bit always returns a 'O'.

Not all possible values of A [10: 4] map into a configuration register. Such values should be considered reserved for future configuration registers. Only values of'00' should be written to reserved register addresses, to maintain compatibility with future chips. Reading a reserved register always returns'00'.

All registers below are read/write accessible, except where noted. All address values are given in hexadecimal. The Init column indicates the values of the respective variables at reset.

Configuration Register Address Map Table 8: Configuration Register Summary Address Range Register Group Description Chip ID, data format, bus width, burst OxOOO-Ox170 System Interface control, block transfer setup, and interrupt handling. Time unit, frame, field, and time slot Ox180-Ox2FO General Timing length definitions. Ox300-Ox3F Strobe Control Strobe specification by time slot. LED timing, ITO refresh timing, data Ox400-Ox5fU LCD Control inversion, and ring electrode control. Ox600-Ox7fD DMA Control Self test, scroll, and pattern fill.

System Interface Register (0x0) The Display System revision number is contained in the 8 bit Chip Id register. This register is read only.

Table 9: System Interface Register Addr Bits Init Name Description Ox000 Chip identification (read only). Always 7: 0 1 7D returns OxO1 on a configuration read. OxO10 Pixel format. Enter 1'for monochrome 7 0 monochrome data packed 8 pixels to a byte. Enter'0' for RGB data packed 2 pixels to a byte. Dummy bit. Value of bits used to fill out 6 0 dumbit data buses containing RGB pixels on reads. RGB Color value written to the pixel for 5: 3 7 background_color an arriving'1'bit in monochrome mode. RGB Color value written to the pixel for 2: 0 0 foreground_color - an arriving'0'bit in monochrome mode. 0x020 Pin readyN behavior on reads. Enter'1'to keep readyN de-selected on the return of the last read data, and during the bus 2 0 ready_off turnaround cycle. Enter '0' to assert readyN on the last data and bus turnaround cycles. Data bus width. Enter: 00: 8-bit bus 1: 0 0 width 01 : 16-bit bus lx : 32-bit bus. 0x030 Vertical orientation. Enter 0'to position 1 0 vertu flip row 0 at the far top, and row 799 at the far bottom. Enter 1'for the reverse. Horizontal orientation. Enter 0'to position column 0 at the far left, and 0 0 horiz_flip column 799 at the far right. Enter'1'for the reverse. 0x040 Burst size. Enter: 00: 1-word burst (no burst) 2: 1 0 burst siz 01 : 2-word burst 10: 4-word burst 11: 8-word burst. Burst address ordering. Enter: 0 0 subblock 0: linear ordering 1: sub block ordering 0x050 Row auto increment. Enter'1'to xfer_row_incremen 1 0 automatically increment xfer row at the t- end of each row in a block transfer. Block transfer enable. Enter'1'to enable block transfer operations. Enter 0'to use 0 0 Mocker the address in A [18: 0] as the address for a data access. 0x080 Lowest order 8 bits of address of left column for block transfer operation. 7:0 00 xfer_left_col[7:0] Lowest bit, xfer_left_col[0] is read only and always returns'0'. 0x090 Highest order 2 bits of address of left 1: 0 0 xfer_left_col[9:8] column for block transfer operation. OxOAO Lowest order 8 bits of address of right column for block transfer operation. 7:0 00 xfer_right_col[7:0] Lowest bit, xfer_right_col[0], is read only and always returns'1'. OxOBO Highest order 2 bits of address of right 1: 0 0 xfer_right_col[9:8] column for block transfer operation. OxOCO xfer_start_row[7: 0 Lowest order 8 bits of starting row in 7:0 00 block transfer operation. OxODO xfer_start_row[9:8 Highest order 2 bits of address of starting 1:0 0 row in block transfer operation. OxOEO Lowest order 8 bits of current row in block transfer operation. Initialized to 7: 0 00 xfer_cur_row[7:0] xfer_start_row[7:0] whenever block_xfer =0. Read only. OxOFO Highest order 2 bits of address of current row in block transfer operation. Initialized 1: 0 0 xfer_cur_row[9:8] to to xfer_start_row[9:8] whenever block_xfer = 0. Read only. Interrupt enable register #0. Writing a'1' to a bit enables the detection of the Ou100 corresponding condition to cause an interrupt request on pin irqN. 7 irq enable [77 Pattern fill done enable. 6 irq_enable [67 Data inversion done enable. 5 irq_enable [5 Pending data inversion enable. 4 0 irq_enable[4] Frame done enable. 3 0 irq_enable [37 Field done enable. 2 0 irq_enable[2] Region done enable. 1 0 irq_enable[1] Slot done enable. 0 irq_enable [0 Tick done enable. Ou 110 Interrupt enable register #1. 1 0 irq_enable[9] Self-test done enable. 0 0 irq_enable[8] Scroll done enable. Interrupt status register #0. On a read, a '1'in a bit indicates the detection of the corresponding condition since the last ou120 time the bit was cleared. Writing a 0'to a bit clears the bit. Writing a'1'has no effect. 7 0 irg-status [77 Pattern fill done status. 6 0 irq_ status [6] Data inversion done status. 5 0 irq_ status [5] Pending data inversion status. 4 irq_ status [47 Frame done status. 3 0 irg-status [3J Field done status. 2 0 irq_ status [2J Region done status. 1 0 irq_ status [1] Slot done status. 0 0 irq_ status [0 Tick done status. 0x130 Interrupt status register #1. 1 0 irq_ status [9] Self-test done status. 0 irq_ status [87 Scroll done status.

System Registers (OxlO-OxFO) Pixel Formats (Ox 10) There are 3 bits of data associated with every pixel in the Backplane. In RGB mode, when the MSB of the low and high order 4 bits in an 8 bit data word is thrown out, the remaining 6 bits are stored at each pixel pair location. In Monochrome mode, the value of each bit in the data word tells the Display System which value to store at each pixel location. For example, if the 8 bit data word is 11110000, then each of the first 4 pixel locations would be filled with the 3 bit foreground color value, and the next four pixel locations would be filled with the 3 bit background color. Refer to previous table for the description of each bit in the pixel format configuration register, Ox 10. In this register the user can select between Monochrome and RGB mode, set the background and foreground color, and set the value of the dumbit. The dumbit is the value of the extra two bits in an 8 bit word when reading from RGB pixel values.

Bus Width and Ready Control (Ox20) The width of the data bus can be configured in the system register Ox20. The width can be 8,16, or 32 bits in length. This register also contains a bit called ready off which controls the behavior of the readyN signal on reads. When this bit is set, the readyN signal will be de-selected on the return of the last read of data, and during the bus turnaround cycle. The reverse is true when the ready off is set to 0.

Orientation (Ox30) The orientation register allows the pixel orientation to be changed in one of four ways. The orientation can be flipped vertically, horizontally, both vertically and horizontally, or unmodified. A vertical flip puts pixels from the right side over to the left. A horizontal flip puts pixels from the top to the bottom.

Block Move Control (0x50-OxFO) There are nine registers that control block moves. They are block control register, left column, right column, start row, and current row. Several of these registers are split into two parts for high and low order bits. Register 0x50 controls starting and stopping the block moves. The other registers set the row and column position for the block move.

Interrupt Control Registers (0x100-0x130) Interrupts can be sent to the host system from the Display System following several events. The criteria for controlling the interrupt select is setup in the Interrupt Configuration registers that range from address 0x100 to Ox130. Registers 0x100 and Ox 110 are the interrupt enable signals that specify which Display System events can cause an interrupt to occur. Registers Ox 120 and Ox 130 are the interrupt status registers that indicate which Display System events have triggered an interrupt.

When registers Ox120 and 0x130 are read, a 1 in a particular bit position indicates that the corresponding Display System event has caused an interrupt. The host system can write a 0 to appropriate interrupt status register bit position, to clear the interrupt event and thereby end the interrupt cycle.

The Boolean expression for the output interrupt request pin irqN is irqN =- ! (irqstatus & irq_enable), Where the AND (&) operation is bit-wise between the irq_status and irq_enable arrays, and the bits of the resultant array are ORed (|) together, and then inverted General Timing Registers (0x180-Ox2FO) Tick (Ox 180-Ox 190) The Tick Configuration register is used for setting up the tick length of the in system timers. The tick length can be 32,64, or 96 clock cycles. This register also contains the tick-enable bit. The tick-enable bit enables or disables all timers.

Time Slot (OxIEO-OxIFO) The Time Slot register contains the timer overflow values for the transition and flash regions of fields 0 to 3. This register also contains the remaining count for the current time slot.

Field (OxlCO-lD0) The Field register is used for setting the number of time slots in the transition and flash regions.

Frame Configuration (OxAO) A frame can be defined as having one, two, three, or four fields. While each field has the same number of time slots, the time slots for each are individually programmable. This way the lengths of each field are individually programmable.

The timing of interrupts, ITO refreshes, LED flashes, and bit plane strobing are all entered relative to the definition of a frame. The Frame Configuration register holds the number of fields in a frame, and the field division. The field division bit chooses between having only one flash region per frame, or a flash region following each transition region.

Table 10: General Timing Registers Addr Bits Init Name Description Ox180 Tick length in clock cycles. Enter 0'for 7: 0 FF tick cycles 32 cycles,'1'for 64 cycles,'2'for 96 cycles, etc. Ox190 Master enable for all timers. Enter 0'to turn off all timers. A rising edge on this bit serves as the master trigger for all 0 1 tick enable timers. Note that the ITO refresh timer is enabled directly out of reset, prior to a rising edge on tick enable. OxlAO Number of fields in a frame. Enter 0'for 2: 1 0 nfields 1 field,' !'for 2 fields, etc. Field division. Enter'1'to define each field as containing only a flash region. 0 0 flash only Enter 0'to divide each field into a transition region, followed by a flash region. OxlCO Number of time slots in the transition 4: 0 0 ntrans slots region of a field. Enter 0'for 1 time slot, '1'for 2 time slots, etc. OxlDO Number of time slots in the flash region of 4: 0 0 nflash_slots a field. Analogous to ntrans slots. OxIEO Highest order 8 bits of current count of 7: 0 00 cur_slot_lng [7 : 0] remaining ticks in a slot. Read only. Interpret 0'as 1 tick, 1'for 2 ticks, etc. OxlFO Highest order 8 bits of current count of 7: 0 04 cur_slot_lng[15:8] remaining ticks in a slot. Read only. Ou200 Lowest order 8 bits of length of a time slot trans_slot_lng[0][ 7: 0 00 7 : 0] in the transition region for field 0. Enter 7 : OJ '0' for 1 tick,'1'for 2 ticks, etc. 0x210 trans_slot_lng[0][ Highest order 8 bits of length of a time 7:0 04 15:8] slot in the transition region for field 0. Ox220 trans_slot_lng[1][ Lowest order 8 bits of length of a time slot 7:0 00 7:0] in the transition region for field 1. Ox230 trans_slot_lng[1][ Highest order 8 bits of length of a time 7:0 04 15:8] slot in the transition region for field 1. Ox240 trans_slot_lng[2][ Lowest order 8 bits of length of a time slot 7:0 00 7 : 0] in the transition region for field 2. 0x250 7 0 04 trans_slot_lng [2] [Highest order 8 bits of length of a time 7:0 04 15 : 8] slot in the transition region for field 2. Ox260 7 : 0 00 trans_slot_lng [3] [Lowest order 8 bits of length of a time slot 7: 0 00 7 : 0] in the transition region for field 3. Ox270 trans_slot_lng[3][ Highest order 8 bits of length of a time 7:0 04 15. 87 slot in the transition region for field 3. Ox280 flash_slot_lng[0][7 Lowest order 8 bits of length of a time slot 7:0 00 :0] in the flash region for field 0. 0x290 flash_slot_lng[0][1 Highest order 8 bits of length of a time 7:0 04 5 : 8] slot in the flash region for field 0. Ox2AO 7 0 00 flash_slot_lng [l] [7 Lowest order 8 bits of length of a time slot 7:0 00 01 in the flash region for field 1. Ox2BO flash_slot_lng[1][1 Highest order 8 bits of length of a time 7:0 04 5:8] slot in the flash region for field 1. Ox2CO flash_slot_lng[2][7 Lowest order 8 bits of length of a time slot 7:0 00 :0] in the flash region for field 2. Ox2DO 7 : 0 04 flash_slot_lng [2] [1 Highest order 8 bits of length of a time 7: 0 04 5 : 8J slot in the flash region for field 2. Ox2EO flash_slot_lng[3][7 Lowest order 8 bits of length of a time slot 7:0 00 zain the flash region for field 3. Ox2FO flash_slot_lng[3][1 Highest order 8 bits of length of a time 7:0 04 5:8] slot in the flash region for field 3.

Color Strobing Registers (0x300 - 0x3F0) Separate gray scale algorithms can be specified for the flash and transition regions of each field. Each algorithm consists of an assignment of bit planes to time slots for up to 32 slots. The gray scale algorithm for the flash region is repeated for every field.

The gray scale algorithm for the flash region is shared among all fields.

Alternately, each flash or transition region can be treated as having only a single time slot, and can be assigned different bit planes from those of other flash and transition regions.

Table 11: Strobe Control Registers Addr Bits Init Name Description Ou300 Color of time slots 3-1 in the transition 7: 2 00 trans_color[3:1] region. Analogous to trans color [O]. Color of time slot 0 in the transition region. Enter 0'for red strobe (plane 0), '1'for green strobe (plane 1),'2'for blue strobe (plane 2), or'3'for no strobe at all. 1: 0 00 trans_color[0] If ntrans_slots = 1, then trans_color [0] is the color of the transition region of field 0 only. Else trans color (O] is the color of time slot 0 for the transition regions of all fields. 0x310 Color of time slots 7-4 in the transition region of all fields. Enter 0'for red strobe (plane 0),'1'for green strobe (plane 1), '2' for blue strobe (plane 2), or'3'for no 7: 0 00 trans_color[7 : 47 strobe at all. The color of time slot n is ignored unless ntrans_slots >= n. Ox320 Color of time slots 11-8 in the transition 7: 0 00 trans_color[11:8] region of each field. Analogous to trans_color [7 : 4]. Ox330 Color of time slots 15-12 in the 7: 0 00 trans_color[15:12] transition region of each field. Analogous to trans_color[7:4]. Ox340 Color of time slots 19-16 in the 7: 0 00 trans_color [19. 16] transition region of each field. Analogous to trans color [7 : 4]. 0x350 Color of time slots 23-20 in the transition 7: 0 00 trans color (23 : 20J region of each field. Analogous to trans_color[7:4]. Ox360 Color of time slots 27-24 in the transition 7: 0 00 trans_color[27:24] region of each field. Analogous to trans_color [7 : 4]. 0x370 Color of time slots 31-28 in the transition 7: 0 00 trans color (31 : 28J region of each field. Analogous to trans-Colorcolor[7:4]. Ox380 Color of time slots 3-1 in the flash. 7: 2 00 flash_color[3:1] Analogous to flast_color[0]. Color of time slot 0 in the flash region. Enter 0'for red strobe (plane 0),'1'for green strobe (plane 1),'2'for blue strobe (plane 2), or'3'for no strobe at all. 1: 0 00 flash_color[0] If nflash slots = 1, thenflash_color [O] is the color of the flash region of field 0 only. Elseflash color of time slot 0 for the flash regions of all fields. 0x390 Color of time slots 7-4 in the flash region of all fields. Enter 0'for red strobe (plane 0),'1'for green strobe (plane 1), '2'for blue strobe (plane 2), or'3'for no 7:0 00 flash_color[7:4] strobe at all. The color of time slot n is ignored unless nflash slots >= n. Ox3AO Color of time slots 11-8 in the flash 7: 0 00 flash_color [I 1 : 8] region of each field. Analogous to flash_color [7 : 4]. Ox3BO Color of time slots 15-12 in the flash 7: 0 00 flash_color[15:12] region of each field. Analogous to flash_color[7:4]. Ox3CO Color of time slots 19-16 in the flash 7: 0 00 flash_color [19 : 16] region of each field. Analogous to flash color (7 : 4J. Ox3DO Color of time slots 23-20 in the flash 7: 0 00 flash_color [23 : 20] region of each field. Analogous to flash_color [7 : 4]. Ox3EO Color of time slots 27-24 in the flash 7: 0 00 flash_color [27 : 24] region of each field. Analogous to flash color 7.: 4j. Ox3FO Color of time slots 31-28 in the flash 7: 0 00 flash_color [31. 287 region of each field. Analogous to flash_color [7 : 4].

ITO Refresh Registers (Ox490-Ox4AO) The ITO Refresh register allows the host system to setup Display System ITO inversion automatically, or manually. This register also controls the relationship between the ring polarity and the ITO voltage, the refresh interval, and provides the status of ITO and ring levels. Polarity switching of the ring electrode can be synchronized with ITO refreshes, or put in a manual control state. When ITO inversion is set to automatic control, the frequency can be set in units of frames.

Internal data inversion can be set for all at once at the time of ITO refresh, or broken into two stages. In the first stage, the field preceding an ITO refresh is used to invert the data strobed in the field concurrent with the ITO refresh. The two-stage format doubles the power consumption incident in internal data inversion.

LED Control Registers (Ox4EO-Ox5BO) The LED Control registers setup and control the behavior of the LED's. In these registers, the delay length after the flash can be set for each led. These registers also provide the ability to configure the led manually using the led level field.

Table 12 : LCD Control Addr Bits Init Name Description 0x400 Lowest order 8 bits of address of left column for data inversion operation. 7: 0 00 inv_left_col[7 : 0/ Lowest 6 bits, inv_left_col [5 : 0], are read only and always return'OxOO'. 0x410 Highest order 2 bits of address of left 1: 0 0 inv_left_col[9:8] column for data inversion operation 0x420 Lowest order 8 bits of address of right column for data inversion operation. 7:0 3F inv_right_col[7:0] Lowest 6 bits, inv_right_col[5:0], are read only and always return'Ox3F'. Ox430 Highest order 2 bits of address of right 1: 0 3 inv_right_col[9:8] column for data inversion operation 0x440 Lowest order 8 bits of address of top row 7: 0 00 inv_top_row[7:0] for data inversion operation. Ox450 Highest order bit of address of top row for 0 0 inv_top_row[8] data inversion operation. Ox460 inv bottom row [7. Lowest order 8 bits of address of bottom 7: 0 2B 0/row for data inversion operation. Ox470 Highest order bit of address of bottom row 0 1 inv_bottom_row[8] for data inversion operation. Ox480 Automatic data inversion. Enter '1' to 2 1 auto invert invert data on ITO refresh. Ignored if auto_refresh = 0. Staged inversion. Enter 1'to invert data over two consecutive fields. Enter 0'to 1 0 staged_invert invert data in one stage. Ignored if auto_refresh = 0. Manual data inversion. Initiates data inversion on a rising edge of this bit. 0 0 man invert Ignored if auto_refresh = 1 and auto_invert = 1. 0x490 Automatic synchronization of ring 4 1 ring_follow electrode to ITO refresh. Enter 1'to enable. Polarity of ring electrode relative to ITO 3 1 ring-polarity value. Enter 0'to echo ITO value,'1'to invert. Valid only if ring_follow = 1. Ring electrode voltage. Read only if 2 1 ring_level ring_ follow = 1. ITO Voltage. Read only if auto refresh = 1 0 ito level 1. 0 1 auto refresh Automatic ITO refresh enable. Ox4AO Interval between refreshes. Enter 0'for 1 7: 0 0 refresh_interval frame, '1' for 2 frames, etc. 0x4E0 Highest order 8 bits of current count of remaining ticks in the flash delay. Read 7: 0 00 cur_delay[7:0] only. Interpret'0'as 1 tick,'1'for 2 ticks, etc. Ox4FO Highest order 8 bits of current count of 7: 0 04 cur_delay [l 5 : 8] remaining ticks in the flash delay. Read only. 0x500 Lowest order 8 bits of illumination delay flash_delay[0][7: 0 7: 0 00 after the start of the flash region for field 0. Enter '0' for 0 ticks, '1' for 1 tick, etc. 0x510 Highest order 8 bits of illumination delay flash-delay[0][15: 7: 0 00 after the start of the flash region for field 0. Ox520 Lowest order 8 bits of illumination delay flash-delay[1][7 : 0 7: 0 00 after the start of the flash region for field Ox530 Highest order 8 bits of illumination delay flash-delay[1][15: 7: 0 00 after the start of the flash region for field Ox540 Lowest order 8 bits of illumination delay flash-delay[2][7 : 0 7:0 00 after the start of the flash region for field 2. 0x550 Highest order 8 bits of illumination delay flash_delay[2][15: 7:0 00 after the start of the flash region for field 8] 2. 0x560 Lowest order 8 bits of illumination delay flash_delay[3][7 : 0 7: 0 00 after the start of the flash region for field 3. Ox570 Highest order 8 bits of illumination delay flash_delay[3][15: 7:0 00 after the start of the flash region for field 8] 3. 0x580 Field 1 illumination. Analogous to field 0 5:3 0 led_color[1] illumination. 2 0 led-color [OJ [2] Blue illumination in field 0. led_color [0] [1] Green illumination in field 0. Red illumination in field 0. Enter a'1'to 0 O led_color [0] [0] turn the red LED on during the flash region of field 0. 0x590 Field 3 illumination. Analogous to field 0 5: 3 0 led_color[3] illumination. Field 2 illumination. Analogous to field 0 2: 0 0 led_color[2] illumination. Ox5AO Manual LED control. Enter'1'to control the led pin manually using the led level 3 0 man-led field (5BO : 3). Enter 0'to put the led pin under automatic control of the chip timing system. Manual blue control. Enter'1'to control the blue pin manually using the blue-level 2 0 man blue field (5BO : 2). Enter 0'to put the blue pin under automatic control of the chip timing system. Manual green control. Enter'1'to control the green pin manually using the 1 0 man_green green_level field (5BO : 1). Enter 0'to put the green pin under automatic control of the chip timing system. Manual red control. Enter'1'to control the red pin manually using the red level 0 0 man_red field (5B0 : 0). Enter 0'to put the red pin under automatic control of the chip timing system. 0x5B0 3 0 led level LED Voltage. Read only if man-led 0. 2 0 blue level Blue voltage. Read only if man_blue = 0. Green voltage. Read only if man_green 1 0 green level 0. 0 0 Red level Red voltage. Read only if man red = 0.

DMA Control Registers (0x400 - 7F0) The DMA Control registers can be divided into several groups. The groups are Data Inversion, Pattern Fill, Scrolling, and Self Test.

The Data inversion group is located from Ox400-Ox480. These registers contain the row and column pixel array positions of the region to be inverted. Inversion of this region can happen automatically or manually, depending on the value placed in the man-invert and auto invert register bits.

The Pattern Fill is achieved by writing to several registers. These include, DMA Region Registers (Ox600-Ox670), Pattern Configuration Registers (Ox680- Ox6BO), and Fill Configuration Register (Ox6CO). The DMA Region Register set is used to set up the pixel array area to be filled with a pattern. It contains right and left column pixel positions, and top and bottom pixel positions. The Pattern Configuration Registers are where the pattern to be loaded into the region is set up.

Finally, the Fill Configuration Register turns on or off the pattern fill. Figure 13 illustrates how patterns are loaded into the array 1302 with and without the rotate pattern bit set.

Scrolling can be accomplished by writing to the DMA Region Registers and the Scroll Configuration Register (Ox6DO). Once the DMA region is specified, the Scroll Configuration Register is used to specify the direction of scrolling and enable scrolling. Figure 14 demonstrates how data is moved relative to the scroll direction.

Self-test is a feature that allows the host system to check the integrity of the Display System pixel array automatically. The registers associated with self-test are located between addresses Ox6EO to Ox7FO. The self-test is started, stopped, and paused by writing to the register bits at Ox6EO. If the Display System has any defective pixels in the array, the register at Ox6FO will contain the number of failed pixels. The rest of the registers from Ox700-Ox7FO contain details about failed pixels, such as column and row position.

Table 13: DMA Control Registers Addr Bits Init Name Description Ou600 Lowest order 8 bits of address of left column for scroll, pattern fill, and self-test 7: 0 00 dma le operations. Lowest bit, dma_left_col[0], is read only and always returns'0'. Ox610 Highest order 2 bits of address of left 1 : 0 0 dma_left_col[9:8] column for scroll, pattern fill, and self-test operations. Ox620 Lowest order 8 bits of address of right dma_right_col[7:0 column for scroll, pattern fill, and self-test 7:0 1F ] operations. Lowest bit, dma_right_col[0], is read only and always returns'1'. 0x630 Highest order 2 bits of address of right dma_right_col[9:8 1: 0 3 column for scroll, pattern fill, and self-test operations. Ox640 Lowest order 8 bits of address of top row 7: 0 00 dma_top_row[7:0] for scroll, pattern fill, and self-test operations. 0x650 Highest order 2 bits of address of top row 1: 0 0 dma_top_row[9:8] for scroll, pattern fill, and self-test operations. Ox660 Lowest order 8 bits of address of bottom dma_bottom_row[ 7: 0 57 row for scroll, pattern fill, and self-test 7:0] operations. Ox670 Highest order 2 bits of address of bottom dma_bottom_row[ 1: 0 2 row for scroll, pattern fill, and self-test 9 : 8J operations. Ox680 Pattern for red data (plane 0) during 7:0 55 red_pattern[7:0] pattern fill and self-test operations. 0x690 Pattern for green data (plane 1) during 7: 0 AA green_pattern[7:0] pattern fill and self-test operations. Ox6AO Pattern for blue data (plane 2) during 7: 0 55 blue_pattern[7 : 0/ pattern fill and self-test operations. Ox6BO Pattern rotate. Enter'1'to rotate the red, green, and blue patterns by a number of 0 1 pattern_rotate bits each equal to the row number modulo 8. Ox6CO Pattern fill enable. A rising edge on this 0 0 fill_enable bit initiates a pattern fill. Ox6DO Scroll enable. A rising edge on this bit 2 0 scroll enable initiates a scroll. Scroll direction. Enter 0'to scroll 1: 0 0 scroll direction upward, 1'to scroll downward,'2'to scroll left, and'3'to scroll right. Ox6EO Self-test enable. A rising edge on this bit 1 0 self-test_enable initiates a self-test. Self-test pause. A'1'on this bit pauses the self-test between the writing and 0 0 self-test_pause checking stages. Other than for diagnostics, this bit should in general be left at'0'. Ox6FO Total number of failing pixels, up to a 7: 0 0 nerr maximum of 255, detected by the self- test. Ox700 bad_word_col[0][ Bits [7: 6] of the starting column address 7:6] of bad_word [0]. Number of bad pixels, up to a maximum of 63, in bad_word[0], where 5 : 0 0 bad_word_nerr[0] bad_word[0] is the first pixel word found to have failing pixels. If there is no bad_word[0], then bad word_nerr [O3 = 0. 0x710 bad word col Highest order 2 bits of the starting column 1: 0 0 [0] [. : 8] address of bad word [0]. Ox720 bad_word_row[0][ Lowest order 8 bits of the row address of 7: 0 0 7 : 0 bad_word [O]. Ox730 bad_word_row[0][ Highest order 2 bits of the row address of 1: 0 0 9:8] bad_word[0]. Ox740 bad_word_col[1][ 7: 6 0 Analogous to bad_word_col[0][7:6]. 7:6 5: 0 0 bad word nerr [l] Analogous to bad_word_nerr[0]. Ox750 bad word col 1: 0 0 Analogous to bad_word_col [0][9:8]. [1][9:8] 0x760 bad_word_row[1][ 7: 0 0 Analogous to bad_word_row[0][7:0]. z Ox770 bad_word_row[1][ 1: 0 0 Analogous to bad-word-row [OJ [9.-8]. 9:8] Ox780 bad_word_col[2][ 7: 6 0 Analogous to bad word_col[0][7:6]. 7:6] 5: 0 0 bad_word_nerr[2] Analogous to bad_word_nerr [O]. Ox790 bad word col 1: 0 0 Analogous to bad_word_col [0][9:8]. [2][9:8] 0x7AO bad_word_row[2][ 7: 0 0 Analogous to bad_word_row[0][7:0]. 7:0] Ox7BO bad_word_row[2][ 1: 0 0 Analogous to bad_word_row[0][9:8]. 9: 8] Ox7CO bad_word_col[3][ 7: 6 0 Analogous to bad_word_col[0][7:6]. 7:z 5:0 0 bad_word_nerr[3] Analogus to bad_word_nerr[0]. 0x7D0 bad word col 1: 0 0 Analogous to bad_word_col [0][9:8]. [3][9:8] Ox7EO bad_word_row[3][ 7: 0 0 Analogous to bad_word_row[0][7:0]. 7:0] 0x7F0 bad_word_row[3][ 1 : 0 0 Analogous to bad_word_row[0][9:8]. 9:8]

Electrical Characteristics-Basic Chip Parameters The table below gives the basic chip parameters for the Backplane IC.

Table 14: Basic Chip Parameters Parameter Value Units Notes Technology 0.35 um Single-poly, quad-metal, salicide process. Maximum Internal operating clock identical to system Operating 66 MHz bus clock. Frequency Voltage on individual pixels alternates Maximum 3. 6 V between ground and operating voltage. Operating Voltage Voltage on ITO driven independently. Minimum 2. 7 V 3V Battery operation supported. Operating Voltage Minimum Operating 0 C Temperature Maximum Operating 70 C Temperature Dimensions 12.9 x 9.6 (mm) ^2 Resolution SVGA 800 x 600 Pixels. Pixel Type SRAM Pixel Depth 3 Bits Power miser mode/color rich mode. Color rich mode value does not include additional Maximum 30/100 mW power consumed by an external frame Operating Power buffer. Illumination power excluded in both cases.

Color Rich Display Controller (CRASIC) Description The Color Rich Display Controller serves as the external frame buffer controller and system interface for the Microdisplay. It receives image data from a microprocessor or other external host, reformats the data, and transmits the data to the Backplane IC of the Display Module. The Backplane IC maps the data onto an 800 by 600 (SVGA) Liquid Crystal on Silicon display. The Color Rich Display Controller controls the transfer of the frame buffer data to the display, providing an enhanced color rich image that is illuminated with the help of the Analog Controller, and the LEDs.

Features 66 MHZ synchronous interface to host processor.

. 3. 3V 1/Os 2. 5V Core power Implemented in. 25u cmos 304 pin BGA package Awake/Active/Dormant and Sleep power saving modes

Host MPU to MicroDisplay, Analog Controller and SDRAM System Interface and Timing Pin Description Tables 15-22 set forth pin assignments and pin descriptions.

Table 15: COLOR RICH CONTROLLER Pinout Pin Description Pin Description Pin Description Pin Description 1 CRDD [28] 26 CRMA [1] 51 D [3] 76 OS_irqBN 2 CRDD [27] 27 CRMA [5] 52 D [23] 77 OS_rstN 3 CRDD [22] 28 CR_casN 53 D [1] 78 SDI4 4 CRDD [14] 29 CRdqm [2] 54 D [14] 79 CRDA [7] 5 TCK 30 SDI7 55 OEN 80 OS_clkB 6 MST 31 ABRD-WRN 56 A [7] 81 OSclkA 7 MMS2 32 SDI5 57 A [2] 82 CRDD [24] 8 SD01 33 NANDOEN 58 A [21] 83 CRDD [29] 9 SDO6 34 TDO 59 sirqN 84 CR_DD [30] 10 SDO3 35 CODEC_DATO 60 A [11] 85 CRDD [31] 11 CRMD [24] 36 PCDBN 61 A [17] 86 CRDD [25] 12 CR_MD[25] 37 ACE2N 62 A [23] 87 CRDD [19] 13 CRMD [26] 38 PIORN 63 A [20] 88 CRDD [26] 14 CRMD [27] 39 PCDAN 64 TPD 89 CR_DD [21] 15 CR MD [28] 40 PWAITN 65 OS_powerN 90 CR DD [20] 16 CRMD [29] 41 IOIS16N 66 OS_power 91 CR_DD[1] 17 CRMD [30] 42 IGPIOJ) 67 SDI 92 spare3 18 CRMD [31] 43 D [9] 68 XTST 93 TMS 19 CRMD [19] 44 D [12] 69 OS_irqAN 94 MMS0 20 CRMD [15] 45 D [7] 70 XTCK 95 MMS3 21 CRMD [16] 46 I2C_data 71 AIC_csBN 96 SDO7 22 CRMD [17] 47 D [5] 72 OScsBN 97 SD04 23 spare4 48 D [10] 73 AIC_csAN 98 CR_MD[20] 24 OGPIO_0 49 D [4] 74 CRDA [5] 99 CRMD [21] 25 CR_MA[7] 50 D [24] 75 OS_rdceAN 100 CR_MD [22] Pin Description Pin Description Pin Description Pin Description 101 CRMD [23] 126 ABCE1N 151 spare1 176 SDO5 102 CR_MD[10] 127 D [8] 152 OGPIO_1 177 CR_MD [4] 103 CR_MD[11] 128 D [21] 153 FORCE1 178 CR_MD[5] 104 CRMD [12] 129 D [6] 154 PLL_LOCK 179 CR_MD[6] 105 CRMD [13] 130 D [11] 155 CRDA [4] 180 CRMD [7] 106 CRMD [14] 131 D [25] 156 OS_rdceBN 181 CR_MD[8] Pin Description Pin Description Pin Description Pin Description 107 CRMD [18] 132 D [17] 157 OS_rdyBN 182 CR_MD[9] 108 CRMD [2] 133 D [31] 158 OS_wrN 183 CR_MA[12] 109 CRMD [3] 134 D [2] 159 SDI3 184 CR_mcke 110 CRMA [6] 135 D [29] 160 AIC_clkA 185 CR_MD[0] 111 CR_MA[11] 136 D [0] 161 AIC_clkB 186 CR_MD[1] 112 CRMA [4] 137 D [22] 162 CRDD [23] 187 spare7 113 CRMA [0] 138 A [19] 163 CR_DD [15] 188 CRMA [10] 114 CR_mwrN 139 RDY 164 CR_DD [9] 189 CRMA [3] 115 CR_mclk 140 vss 165 CR_DD [16] 190 CRrasN 116 SDI6 141 A [9] 166 CRDD [17] 191 CRdqm [3] 117 IGPIO_1 142 A [22] 167 CRDD [18] 192 CR_dqm[0] 118 CS3N 143 A [4] 168 CR DD [13] 193 ACD2N 119 AWP_IO16N 144 A [12] 169 CR DD [6] 194 BCD1N 120 CODEC_WS 145 A [18] 170 CR_DD[0] 195 BCD2N 121 BCE1N 146 I2C_clk 171 CR_mcs0N 196 CODEC_BCLK 122 AWAITN 147 TPC 172 spare2 197 NANDWEN 123 POEN 148 BCE2N 173 TRST 198 BWP_IO16N 124 CFPWRONN 149 ABCE2N 174 MMS1 199 SCLK_C 125 ACE1N 150 XSM 175 RESETN 200 PSKTSEL Pin Description Pin Description Pin Description Pin Description 201 PCE2N 226 FORCEO 251 SD02 276 VDDL 202 ACD1N227 PLLOUT252 VSS277 GPIO2 203 BWAITN 228 CR DA [2] 253 VDDH 278 VSS 204 D [13] 229 OSrdyAN254 CRMA [8] 279 VDDH 205 D [20] 230 CR_DA[0] 255 VSS 280 CS5N 206 D [18] 231 CR DA [3] 256 VDDL 281 VSS 207 OGPIO_3 232 SDI2 257 CR_MA [9] 282 VDDL 208 WEN233 CRDA [6] 258 CRMA [2] 283 A [8] 209 D [16] 234 SDI1 259 VSS 284 VSS 210 D [15] 235 CRDD [2] 260 VDDL 284 VSS 211 D [30] 236 CR DD [8] 261 CR_dqm[1] 285 - 212 D [28] 237 CR DD [3] 262 VSS 286 A[10] 213 D [27] 238 CR DD [10] 263 VDDH 287 VSS 214 A [14] 239 CRDD [11] 264 CODECDATI 288 VDDL 215 SDCLK240 CRDD [12] 265 VSS 289 TDI 216 spare8241 CRDD [5] 266 VDDL290 PLLCLKIN 217 A [15] 242 DE 267 SFRM-C 291 VSS 218 A [3] 243 VSS 268 VSS 292 VDDL 219 A [16] 244 VDDL269 VDDH293 CRDA [1] 220 A [5] 245 SMCK 270 PCE1N 295 VDDH 221 A [13] 246 VSS 271 VSS 296 CR DD [7] 222 A [6] 247 VDDH 272 VDDL 297 VSS 223 SDO248 errcrasic273 D [19] 298 VDDL 224 FORCEZ 249 VSS 274 D [26] 299 OScsAN 225 CLKCTL 250 VDDL 275 VSS 300 VSS Pin Description Pin Description Pin Description Pin Description 301 VDDH 302 CRDD [4] 303 VSS 304 VDDL Table 16: System Interface Pins Name Direction Description RESETN Input System reset. SIRQN Output System interrupt request OEN Input System read. Active low WEN Input System write. Active low. CS4N Input System chip select. Active low CS5N Input System chip select. Active low RDY Input System ready A [23: 0] Input System address. D [31 : 0] Bi-System data. directional SDCLK Input System clock.

Table 17: SDRAM Interface Pins Name Direction Description CRMA [10 : Output Memory address bus. Multiplexed row and column 0] address. Row addresses use MA [10: 0]. Column addresses use MA [7: 0]. Note that MA [10] is a control signal during column address strobing. CRMD [31: Bi-Memory data bus 0] directional CRMCLK Output Memory clock. CRRASN Output Row address strobe. Latches row addresses on the positive edge of mclk with mrasN low. CRCASN Output Column address strobe. Latches column addresses on the positive edge of mclk when low. CRMWRN Output Memory write enable. Enables write operation and row precharge. Latches data in starting from casN and mwrN low. CRDQM [3 Output Data input/output masks by byte. Serves as the byte : 0] enables on writes CRMCKE Output Memory clock enable. Used to enable auto refresh during sleep mode. mcsN Output Chip select. Always active following initial 100 us wait on power on. Active low.

Table 18: Interface Pins Name Direction Description Output Display address. CR_DA[7:0 ] CR_DD[31: Bi- Display data. 0] directional AICCSAN Output Chip select for first Analog Controller. Active low. AICCSBN Output Chip select for second Analog Controller. Active low. AICCLKA OutputClock for first Analog Controller. AICCLKB Output Clock for second Analog Controller. OSCSAN Output Chip select for first Backplane. Active low. OSCSBN Output Chip select for second Backplane. Active low. OSCLKA Output Display clock for first Backplane. OSCLKB Output Display clock for second Backplane. OSIRQAN Input Display interrupt request for first Backplane. Active low. OSIRQBN Input Display interrupt request for second Backplane. Active low OSRESET Output Backplane reset. Active low. N OSRDCE Input First Backplane read data clock enable. Active low. AN OSRDCE Input Second Backplane read data clock enable. Active low. BN OS RDYA Input First Backplane ready. Active low. N OSRDYB Input Second Backplane ready. Active low. N OSWRN Output Shared write for both Analog Controllers and Backplanes Active low. OSPOWE Output R OSPOWE Output RN Table 19: Serial Interface Pins Name Direction Description SB_CLK Output Serial clock. Should be connected to a pull-up resistor. Bi- SBDATA Serial data. Must be connected to a pull-up resistor. directional Table 20: Tap Control Pins

Name Direction Description TPC Input Tap clock. TPD Input Tap data.

Table 21 : Test Pins Name Direction Description TCK Input Boundary Scan TAP controller clock TMS Input Boundary Scan Setting of condition for TAP controller TRST Input Boundary Scan Reset for Tap controller TDI Input Input of scan data of BSR and IR TDO Output Output of scan data of BSR and IR XTST Input Setting of LSI testing condition XSM Input Setting of scan shift condition in LSI test mode XTCK Input Test clock input SDI Input Data input for scan chain SDI [1: 7] Input Additional inputs for scan chain SDO Output Data output for scan chain SDO [1: 7] Output Additional outputs for scan chain MST Input Setting of memory unit testing condition SMACK Input Mode clock for RAM unit testing MMSO-4 Input Selection of memory macro for testing. FORCE0 Input Force all outputs to 0 for testing FORCE1 Input Force all outputs to 1 for testing FORCEZ Input Force all outputs to Z for testing.

Table 22 : PLL Pins Name Direction Description PLLCLKI Input 3.6MHZ input to multiply to 68.4 for internal core N frequency PLLRST Input PLL reset PLLOUTOutputPLLCLOCK output for debug. PLLLOCK Output Signal indicating PLL is locked. High indicates Locked PLL CLKCTL Input Clock input bypass for PLL. PLLSEL Input Low will select the PLL BYPASS as input clock. PLLVDD Input Analog VDD for PLL PLLGND Input Analog GND for PLL

Color Rich Display Controller Configuration Figure 15 is a block diagram that shows the system components of an embodiment of the Microdisplay 1500. The Color Rich Display Controller can support up to two Display Modules 1502.

The Color Rich Display Controller [CRASIC] is comprised of three major subsystems and peripheral GLU (primarily intended for an SA1110 host processor).

1. SDRAM Controller 1504. The SDRAM section arbitrates access to the SDRAM for DAPPER, CRISP, and a host CPU. It also takes care of SDRAM refresh.

2. DAPPER 1506. The"Dithering and Planarization Process Engine & Router," converts incoming 8-bit/pixel data into the 24-bit color space, then dithers down to 9,12, or 15"Bit Planes". The dithering process attempts to preserve general 24-bit color depth by sacrificing absolute spatial resolution, e. g., adjusting the color of adjacent bits to give an overall illusion that color has been preserved.

3. CRISP 1508 is a very limited, but highly programmable, processor that manages timing and data transfers to the MicroDisplay (s) to produce images. The CRISP may, in its full implementation, also take care of tasks such as Cursor management, LC temperature compensation, and stereo audio. See the section on the CRISP, below.

The peripheral GLU includes a PS/2 port, supplemental logic for Compact Flash slots, and Serial device bus master.

Configuration Registers

The external interface for the Color Rich Display Controller provides the means for a host processor to access the entire contents of the SDRAM, MicroDisplay (s) registers and memory, and Analog Controller (s) registers.

These devices, along with Color Rich Display Controller own registers, are arranged into a unified memory map.

The host processor may be interfaced using the full 22-bit address bus, allowing direct access to the entire map. For applications where fewer address lines are desirable, just 13 address lines may be used. In this case, device registers are fully addressable and SDRAM is accessed indirectly through the use of an auto- increment pointer register (see Addressing Control register definition).

Table 23 : CRASIC Address Map Address Device Name Description Range 0x000000 SDRAM Must be at least 512K to support"Color through Rich", but can be as large as 8Mb. Used Ox7FFFFF for display buffers, cursors, palettes, etc. 0x800000 MicroDisplay memory Organization of memory depends on through [Right Device] MicroDisplay mode. Ox8FFFFF 0x900000 MicroDisplay display [Left Organization of memory depends on through Device] MicroDisplay mode. Ox9FFFFF OxAOOOOO MicroDisplay display [Both Organization of memory depends on through Devices-Write Only] MicroDisplay mode. OxAFFFFF OxBOO000 Auxiliary device This is undefined at this time. It might be through used for audio, or as a high-speed data OxB7FFFF transport device, such as Fire-Wire, that could benefit from CRISP's DMA capabilities. Note : at the very least, we ought to have an external chip enable and ready line associated with this"device". It could be serviced by MicroDisplay memory FIFO, using all the same signals. OxB80000 DAPPER Color Palette (8-bit All 256 locations are used to hold the through to 24-bit Lookup Table) palette. These registers are write-only. OxFFFFFF OxCOO000 Analog Controller registers Up to 256 8-bit registers. Registers through [Left Device] appear in least significant byte of 32-bit OxC7FFFF word access. OxC80000 MicroDisplay registers Up to 256 8-bit registers. Registers through [Left Device] appear in least significant byte of 32-bit OxCFFFFF word access. OxDOOOOO Analog Controller registers Up to 256 8-bit registers. Registers through [Right Device] appear in least significant byte of 32-bit OxD7FFFF word access. OxD80000 MicroDisplay registers Up to 256 8-bit registers. Registers through [Right Device] appear in least significant byte of 32-bit OxDFFFFF word access. OxEOOOOO Analog Controller registers Up to 256 8-bit registers. Registers through [Both Devices-Write Only] appear in least significant byte of 32-bit OxE7FFFF word access. OxE80000 MicroDisplay registers Up to 256 8-bit registers. Registers through [Both Devices-Write Only] appear in least significant byte of 32-bit OxEFFFFF word access. OxF00000 CRISP and DAPPER Up to 256 registers. Registers may be up through registers to 32-bits. While primarily used by OxF7FFFF CRISP, these registers are fully visible to the Host processor. 0xF80000 CRASIC External interface These are always readily accessible to the through and GLU peripheral registers host processor. They are only accessible 0xBFFFFF to CRISP through Data Transfer operations such as the MOV instruction.

OxB80xxx DAPPER Registers Table 24: (24-bit)-Palette (Color Lookup Table) Addr Bits Init Name Description 0x000 Palette 00 23: 1 Blue Lookup 00 Blue value for 8-bit pixel equal to 0x00 6 (write only) 15: 8 Green Lookup 00 Green value for 8-bit pixel equal to 0x00 (write only) 7: 0 Red Lookup 00 Red value for 8-bit pixel equal to 0x00 (write only) 0x004 Palette[01 23: 0 RGB_Lookup_01 Same as Palette [0J except for pixel equal to 0x01 0x008 ... Palette 02 23: 0 RGB_Lookup_02 Same as Palette[00] except for pixel equal to 0x02 0x00C ... Palette 03", 0x010 ... Palette[04] ... 0x3F0 ... PaletterFC". 0x3F4 ... Palette[FD 23: 0 RGBLookupFD Same as Palette [OJ except for pixel equal toOxFD 0x3F8 Palette[FE] 23: 0 RGBLookupFE Same as Palette[0] except for pixel equal to 0xFE Ox3F PalettefFFl c 23: 0 RGB Lookup FF Same as Palette [O) except for pixel equal to 0xFF

Note The above registers are addressable by CRISP via Data Transfer Operations, primarily MOV instruction.

0xC00xxx, 0xD00xxx Analog Controller Registers See the section entitled Analog Controller, below 0xC80xxx, 0xD80xxx Display Registers See the section on Display Registers, above.

0xF00xxx CRISP Registers Table 25: (8-bit)-Device Control and Status Addr Bits Init Name Description Ox000 DevCon Device Control Register. 7 GPDD 3GPIO 3 Data Direction 1=Output 6GPDD2GPIO 2 Data Direction 1=Output 5 GPDD_1 GPIO 1 Data Direction 1=Output 4 GPDD0 GPIO 0 Data Direction 1=Output 3 DBT Bus clock: 0 = SysClk/2, 1= SysClk/3 2 CLK Clock Enable for Bus Devices 1 RST Reset Control for Bus Devices 0 PWR Power Enable for Bus Devices 0x004 AuxCon 7GPIO3General Purpose I/O 3 6GPIO2General Purpose I/O 2 5 GPIO_1 General Purpose I/O 1 4GPIO0General Purpose I/O 0 3 WDRUN Watch-Dog/Timer Run 2 WDENBL WatchDog Abort Enable Abort-time out causes abort and sets WD IRQ to host, if enabled. When this bit is false the WatchDog Status is mapped to AUX cc, allowing WD as a general purpose program timer 1AICRSTReset Control for Analog Controller-if dual Displays are used, this signal may be used for both Analog Controllers 0AICPWRPower Enable for Analog Controller-if dual Displays are used, this signal may be used for both Analog Controllers. 0x008 ConStat 7 DGPC2 When a 1 is written, General Purpose Counter 2 is decremented 6 DGPC_1 When a 1 is written, General Purpose Counter_1 is decremented. 5ZGPC 2When General Purpose Counter2-0, this register is set to"1". 4 ZGPC_1 When General Purpose Counter_1 = 0, this register is set to "1". 3 WDTO WatchDog Timer Timed Out (This signal may or may not be assigned an external CRASIC pin. It may be cleared by writing to the WDCNT register.) 2 CIRQ CRASIC Interrupt line level 1 OIRQ_2 MicroDisplay Interrupt line level [Right] 0 OIRQ1 MicroDisplay Interrupt line level [Left] OxO0C WDCnt WatchDog Timer 7: 0 WDCnt The WatchDog/Timer count is an 8 bit register whose value is transferred to a countdown counter each time a CRISP instruction with WDR=True is executed, or when this register is written.

Note The above 8-bit registers are also addressable with CRISP immediate instructions.

OxFOOxxx CRISP Registers Table 26: (8-bit)-Semaphore Registers Addr Bits Init Name Description Ox010 SEM o Semaphore Register 0 7: 0 SET0 Each bit may be read or written by either the Host Processor or CRISP through SEMx registers. 0x014 SEM 1 Semaphore Register 0 7: 0 SEM1 Each bit may be read or written by either the Host Processor or CRISP through SEMx registers. Ox018 SEM 2 Semaphore Register 0 7: 0 SEM2 Each bit may be read or written by either the Host Processor or CRISP through SEMx registers. OxOlC SEM 3 Semaphore Register 0 7: 0 SEM3 Each bit may be read or written by either the Host Processor or CRISP through SEMx registers. Ox020 GP Counter 1 7: 0 GPC_1 General purpose count-down counter 0x024 GP_Counter_2 7 GPC_2 General purpose count-down counter 0x028 Reserved 8-bit OxOF C

Note The above 8-bit registers are also addressable with CRISP immediate instructions.

OxFOOxxx CRISP Registers Table 27: (32-bit)-Data Processing Addr Bits Init Name Description Ox 100 TranSourceStart 31: 0 *TSA Transfer Start Address, used with Data Transfer Operations (MOV, AND, ORR, NOT and XOR instructions). Address must be on a 32-bit boundary (23: 2), but all 32 bits may be used when the register is used for ADD operations. Ox 104 TranSourceEnd 31: 0 *TEA Transfer End Address, used with Data Transfer Operations (MOV, AND, ORR, NOT and XOR instructions). This address must be greater than, or equal to, TSA for predictable operation. Address must be on a 32-bit boundary (23: 2), but all 32 bits may be used when the register is used for ADD operations. 0x 108 TranDestination 31: 2 RowLen This 8-bit value specifies the number of 4 32-bit transfers for each row when the destination is MicroDisplay memory (or memories. It has no effect when transferring to SDRAM or other devices. 23: 2 TDA Transfer Destination Address, used with Data Transfer Operations (MOV, AND, ORR, NOT and XOR instructions). Address must be on a 32-bit boundary. Ox 1 OC CrispPC Register 22: 2 CPC This register holds the current program counter of a CRISP program. When CRISP writes to this register (LDR), it effectively acts as a jump instruction. Address must always be on a 32-bit boundary, and it must be a valid SDRAM address.

Note The above registers are also addressable by CRISP LDR, STR, and ADD* instructions.

OxFOOxxx CRISP Registers Table 28: (32-bit)-RESERVED Addr Bits Init Name Description Oxl 10 Reserved OxlF C

OxFOOxxx DAPPER Registers Table 29: (32-bit)-Control and Dither Grid Addr Bits Init Name Description 0x200 Dapper Control 7 Dapper-Enable When set to"1"Dapper Enable directs all transfers to SDRAM to be processed through the DAPPER. Multiples of 32 pixels, written to 32-pixel boundaries, are required for Dapper to properly process bit plane data through to SDRAM. Note: This bit must be"0"when loading CRISP programs and data to the SDRAM. 6 No Original Data Original pixel data is not written to the SDRAM when this bit is"1"only Dapper processed bit-planes are written. This bit is only valid when Dapper Enable is"1" 5 Palette-Bypass When this bit is"1", the Palette (Color Lookup Table) is bypassed and each 32- bit write to SDRAM is treated as a single 24-bit pixel. When this bit is"O", 32-bit data is treated as four 8-bit pixels to be processed through the Palette. 4 FIFO Reset Writing a"1"to the Dapper FIFO Reset causes the Dapper to reset the FIFO without writing its content to SDRAM. This bit should be toggled back to'0'to make the FIFO functional again. There is no minimum time interval required between bringing the bit to'1'and back tao 0'. 3: 0 Bit Planes Number of bit planes (-1) generated by Dapper. Bit planes are generated R-msb, G-msb, B-msb, and so on through to R- lsb, G-lsb, and B-lsb. Up to 15 bit planes may be generated corresponding to 5 bits per color. 0x204 Plane Base Addr 22: 2 Plane Base The Plane Base register specifies the address of the first"bit-plane"for Dapper processed image data. 0x208 Plane Length 17: 1 Plane Len Plane Length is specified in 1K 0 increments. Ox20C Reserved 0x210 Gird_Bias_Row_0 23: 1 Grid_Col_2 This value is added to Row (mod (3)) =0 6 Column (mod (3)) =2 8-bit color values (R, G, & B) before the data is truncated into N bit planes. This provides the "Dither"dapper does... 15: 8 Grid Col_1 This value is added to Row (mod (3)) =0 Column (mod (3)) =1 8-bit color values (R, G, & B) before the data is truncated into N bit planes. This provides the "Dither"dapper does... 7: 0 Gird_Col_0 This value is added to Row (mod (3)) =0 Column (mod (3)) =0 8-bit color values (R, G, & B) before the data is truncated into N bit planes. This provides the "Dither"dapper does... 0x214 Grid Bias Row 1 23: 0 Grid_Col_2. 0 Same as Grid Bias Row 0, except values are used for Row (mod (3)) =1 data 0x218 Grid Bias Row 2 23: 0 Grid_Col_2. 0 Same as Grid Bias Row 0, except values are used for Row (mod (3)) =2 data. 0x2 1 C Reserved

Note The above registers are addressable by CRISP via Data Transfer Operations, primarily MOV instruction.

OxFOOxxx DAPPER Registers Table 30: (32-bit)-Gamma Waveform Control Addr Bits Init Name Description Ox220 Gamma Wave [00. 031 31: 2 not used 9 28: 2 Form03 5-bit substitution value used when bit- 4 plane data equals binary 00011. 23: 2 not used 1 20: 1 Wform-02 5-bit substitution value used when bit- 6 plane data equals binary 00010. 15: 1 not used 3 12: 8 Form01 5-bit substitution value used when bit- plane data equals binary 00001. 7: 5 not used 4: 0 Form00 5-bit substitution value used when bit- plane data equals binary 00000. 0x224 Gamma Wave[04.07] 31: 0 Wform 07 : 04 Same as Gamma_Waveform [0.3], except these are substitution values for 00111, 00110,00101, and 00100. Ox228 Gamma Wave 08. OB 31: 0 Wform OB: 08 Same as Gamma_Waveform [0.3], except these are substitution values for 01011, 01010, 01001, and 01000. 0x22C Gamma Wave[0C.0F 31: 0 Wform OF: OC Same as Gamma_Waveform [0. 3], except these are substitution values for 01111, 01110, 01101, and 01100. 0x230 Gamma Wave[10.13] 31: 0 Wform 13: 10 Same as Gamma_Waveform[0.3], except these are substitution values for 10011, 10010, 10001, and 10000. 0x234 Gamma Wave[14.17] 31: 0 Wform 17: 14 Same as Gamma-Waveform [0. 3], except these are substitution values for 10111, 10110,10101, and 10100. 0x238 Gamma Wave[18.1B 31: 0 Wform 1B : 18 Same as Gamma_Waveform [0.3], except these are substitution values for 11011, 11010,11001, and 11000. Ox23C Gamma Waver1C. 1F 31: 0 Wform 1F : 1C Same as Gamma-Waveform [0. 3], except these are substitution values for 11111, 11110,11101, and 11100.

Note The above registers are addressable by CRISP via Data Transfer Operations, primarily MOV instruction.

OxFOOxxx DAPPER Registers Table 31: (32-bit)-RESERVED Addr Bits Init Name Description Ox240 Reserved Ox3F C

Note The above registers are addressable by CRISP via Data Transfer Operations, primarily MOV instruction OxF80xxx CRASIC Registers Table 32: CRISP and IRQ control Addr Bits Init Name Description Ox000 Central Control Read and Write Register 0 31 GLU_Logic-Reset 0->1->0 resets misc logic unused 5 14 BackEnd FIFO Re 0->1->0 resets"backend"FIFO set (Analog Controller bus) 13 Extended_RDY l=Hold RDY (ready) true'til WE# OE# goes inactive, O=RDY is a single doc wide 12: 9 OxE Refresh Rate SDRAM refresh 8 : 3 unused 2 STEP Write of 1 executes a single Opcode self-clearing upon completion. (HOLD is s = 1) 1 1 HOLD 0=RUN (continue), 1=HOLD (halt) CRISP process 0 Crisp Reset 0->1 Resets Crisp processor 1->0 transition Starts CRISP at address held in Crisp_PC register 0x004 Timer_Tick 15: 0 TMR Clock scaler for CRISP watchdog/ timer. This value is transferred to an intern Counter. Each time the counter"rolls over' one"tick"is applied to CRISPs watchdog timer register. Ox008 IRQEN Read and Write register 8 Error_IRQ_Enable 1=Enabled. This IRQ may occur if bad transaction to the Display System is detected. (Primarily an unexpected nReady during a row burst). Toggle this bit to'0'tc clear the interrupt. 7 GPIO_3_IRQ_Enab 1=Enabled. When the general- le purpose I/O-3 line transitions from high to low, an interrupted will be generated. Togg this bit to'0'to clear the interrupt. 6 GPIO_2_IRQ_Enab 1=Enabled. When the general- le purpose I/O-2 line transitions from high to low, an interrupted will be generated. Togg this bit to'0'to clear the interrupt. 5 GPIO_1_IRO_Enab 1=Enabled. When the general- le purpose I/O-1 line transitions from high to low, an interrupted will be generated. Togg this bit to'0'to clear the interrupt. 4 GPIO_0_IRQ_Enab 1=Enabled. When the general- le purpose I/O-0 line transitions from high to low, an interrupted will be generated. Togg this bit to'0'to clear the interrupt. 3 WDog_IRQ_Enable 1=Enabled. If the watchdog functio is enabled, and the watchdog timer times-o an interrupt is generated. Toggle this bit to '0'to clear the interrupt. 2 PS2_IRQ_Enable 1=Enabled. When PS/2 port is set t ( automatically poll, and data is available, an interrupt will be generated. Toggle this bit '0'to clear the interrupt. 1 OS2R_IRQ Enable 1=Enabled. When the Right MicroDisplay interrupt line transitions fro high to low, an interrupted will be generate Toggle this bit to'0'to clear the interrupt. 0 052L_IRQ_Enable 1=Enabled. When the Left MicroDisplay interrupt line transitions fron high to low, an interrupted will be generate Toggle this bit to'0'to clear the interrupt. Ox00C IRQSC Read Only Register 8 Error_IRQ_Status 1=True. Buss Error IRQ status. A write of"1"clears. 7 GPIO_3_IRQ_Statu 1=True. GPIO_3 IRQ status. s 6 GPIO_2_IRQ_Statu 1=True. GPIO_2 IRQ status. s 5 GPIO_l_IRQStatu 1=True. GPIO_1 IRQ status. s 4 GIPO_0_IRQ_Statu 1=True. GPIO_0 IRQ status. s 3 WDog_IRQ_Status 1=True. Watchdog overrun IRQ status. 2 PS2_IRQ_Status 1-True. PS/2 IRQ status. 1 OS2R_IRQ_Status 1-True. Right MicroDisplay IRQ status. 0 052L_IRQ_Status 1=True. Left MicroDisplay IRQ status. Ox010 PS2 Control Statu Read and Write Register s 7 PS2 Enable Setting this bit to"1"enables PS/2 port operation, a"0"will disable all PS/2 related functions. 6 PS2Clk_Enable When set to"1"device's Clock is allowed t run. This bit must be set to"1"if PS/2 port interrupts are to be used. This bit must be s to"0"when sending commands to the PS/2 device. 5 Send Command Writing a"1"will cause a PS/2 command sequence to commence, e. g. the PS2Clk is un-inhibited, command and data will be sent, then the"PS2ClkEnable"is set to"1"to allow the PS/2 device to send any data it may have ready. This bit should be cleared before clearing the corresponding interrupt-to ensure no erroneous data is transmitted to the PS2 port. 4 PS2_Reset Resets the PS/2 port when toggled from'0' '1'. Should be normally'0'. As long as this bit is set to'1', the ps2_port will stay in 'inhibit'state. 3 Pus/2 terror Set to"1"if an error is detected on the PS/2 port. The error bit will become'0' automatically after initiating the next send/receive command to the PS2 port. 2 1 : 0 0x014 PS2 Command Read and Write Register 7: 0 PS2 Command To send a command to the PS/2 device, this register is written to the command byte before setting the PS2 send command bit. Ox018 PS2 Data Read Only Register 7: 0 PS2_Data This is a single byte register, which contains the last data received from the PS2 port. 0x1C ID Register Read Only Register 7: 0 The upper nibble of the byte [7: 4] defines the product. The lower nibble [3: 0] defines the revision number of the ASIC. ASIC ID: D [7: zu 15: 8 FPGA ID: D [15: 12] LP1M4-0 FPGA Revision Number : D [11 : 8] : As new features are added for e. g. palette bypass, this number will be incremented. For LP1M4 board the rev. 1 i to be released. (There is no revision 0.) Ox020 Addressing Control Read and Write Register 31 Auto Increment When this bit is set to"1", the DPTR is automatically incremented by 4 bytes subsequent to each transfer, otherwise the indirect R/W will reflect a single CRASIC address location. 30RedirectSDRAM When this bit is set to"1", all Host SDRAM write accesses are redirected using DPTR. 29 CntSkipEnable Used in conjunction with "RedirectSDRAM", setting this bit to "1"enables the use of the Rectangle Control register values to adjust DPTR addressing (see Rectangle-Control register, below). This is the key to handling DVI data automatically. 28: 2 not used 4 23: 2 DPTR The data pointer provides a R/W address for transfers to and SDRAM when Redirect_SDRAM is enabled. All transfers are 32-bit aligned. 1: 0 not used Ox24 Row Control Read and Write Register 31: 2 not used 4 23:1 Row Skip When"Cnt Skip Enable"in 6 Addressing Control register is set to"1"a "RowCount"is exhausted, this value is added to DPTR (23: 2) to form the address the next logical row write to SDRAM. This the number of 32-bit words to skip at the ei of each row. 15: 1 not used 0 9:0RowCountWhen"CntSkipEnable"in Addressing Control register is set to"1"th value provides a count of 32-bit transfers fi each logical image row. Used in conjunctio with"RowSkip", any rectangle up to 102 words wide can be directed into SDRAM Ox24 DE Control Read and Write Register 25 DE Enable When 1', DE signal is asserted periodically. When'0', DE is disabled and never asserted. Note: DE signal is used only when using panel link interface. 24 DE_Polarity When'O', DE will be active low. When'1', DE will be active high. 23 : 1 DE Width The pulse width of DE signal is equal to the number of - 6 clocks of this hex value. 15:0 DE_Frequency This is the hex value of the number of clocks before which the DE signal will be asserted. Ox2C: Reserved for OxOF Debug Registers C OxF80xxx CRASIC Registers Table 33: (8-bit)-Serial EEPROM Addr Bits Init Name Description Ox100 SB Control 7 sbStart Writing a"1"initiates a new command transaction. This is cleared automatically when target device has acknowledged the command and is ready for data transfers, if any. 6: 4 sbCommand Commands supported are as follows 000 = Nop 001 = Read Address 010 = Read 011 = Write 100-111 = not defined 3 sbAddrMode 0 = small address, 1 = extended address 2: 0 sbDevAddr Device address Ox 104 SB Status 7 sbReady This bit goes high when the data register is ready for a transaction. If reading, it goes high when data is available, for write it goes high when the device is ready to receive the next data. 6 sbDone This bit goes high when all data has been transferred, e. g. the Read Address command will return 1 or 2 bytes (depending on device addressing capabilities) and then will be"done". 5 sbError If any transaction is unacknowledged by target device, this bit is set to indicate an error. It is automatically cleared when the next istart is issued. Ox108 SB_ AddrLow 7: 0 sbAddrL For small address devices (such as Analog Controller), this provides addressing of up to 256 locations OX10C SB-AddrHigh 7: 0 sbAddrH Extends the addressing to 64K-byte (such as the CRISP boot ROM) Ox 110 SB Data 7: 0 sbData Data register for I^2C transactions 0x114 - Reserved 0x11C

Dithering and Planerization Processing Engine & (data) Router (DAPPER) The MicroDisplay supports a native color depth of one bit per color, e. g. 8 possible colors for each pixel. In order to generate higher color depths, image data must be separated into color planes for each bit of color depth. Each image color plane is written to the MicroDisplay once per frame, reproducing the image on the display.

The color separation process is somewhat time consuming and inconvenient on most microprocessors. The DAPPER relieves the host processor from this chore by providing an 8-bit per pixel display buffer interface. In addition, DAPPER allows the representation of even higher color depths through the use of an 8-to-24-bit color palette and spatial dithering.

Figure 16 shows how each byte of image data is processed through the palette, adjusted by the"grid" (see Dithering, below), and separated into individual bit planes. Up to 5 bit planes per color can be generated automatically.

Dithering Dithering is achieved with the use of a 3 by 3 noise injection grid. Each (RGB) color of pixel is rounded up or down according to the grid, producing on average the approximate original color when viewed over a group of adjacent pixels. Such spatial dithering improves color fidelity by 3 bits per color at the expense of absolute image resolution.

Either the host or CRISP processor can initialize the three'grid' (g) registers. Each register holds three"noise"values corresponding to pixel column modulo-3. The register used for a given row is selected by the value of row modulo-3.

Data Router The 32 bit values of the unprocessed, and as well as the processed, data (Plane- registers) are written into the local memory (SDRAM). The address generated by the router is stored in an address FIFO (32 deep) known as AFIFO. The corresponding data is stored in DFIFO, as shown in the table below. The router calculates the address for each plane register-write as follows: Destination address = Plane Base Address + (A19-A3 of the first write address remapped as A1 6-AO) + (Plane number x Plane-Length) The Plane number is the value referred by the Row Address A17-A14. The 'Plane_Base_Address'&'Plane_Length'are registers initialized by either the Host or CRISP processor.

The data and addresses are written sequentially into the AFIFO and DFIFO as the plane registers are filled up. The address for the unprocessed data is retained as it is from the processor. The address and the unprocessed data also are written into the AFIFO and DFIFO.

The router arbitrates for the local memory along with CRISP and the Host microprocessor (it only reads in the regular address space and not in the DAPPER address space). When it wins the arbitration, it writes the data into the corresponding address of the local memory (SDRAM).

Table 34 AFIFO (0) DFIFO (0) AFIFO (1) DFIFO (1) AFIFO (2) DFIFO (2) I AFIFO (31) DFIFO (31) The AFIFO and DFIFO are designed into the system to reduce the latency unprocessed data writes of the processor, into the DAPPER. If the AFIFO/DFIFO is full, and the CRISP processor is moving data from the SDRAM to the MicroDisplay, the host processor may be held off from completing a write operation for up to N micro-seconds.

Color Rich Internal System Processor (CRISP) The Color Rich Internal System Processor, or CRISP, is a very small instruction set processor used, primarily, to drive DMA transfers from memory to the MicroDisplay. CRISP is the part of the Color Rich Controller that programmatically controls the operation of the MicroDisplay and Analog Controller.

CRISP is designed to handle a simple 512-Color mode of operation, but is flexible enough to manage higher color operations, as well as stereo imaging on dual displays. With its simple instruction set, it can simplify cursor tracking, fonts, and multiple screen and window management for the host processor.

It is possible to create a simplified (host) software interface to enable display module customers to develop products without having to understand the details of CRISP programming, MicroDisplay, or Analog Controller. The CRISP programming may be auto-loaded from a serial EEPROM, or downloaded by the host driver at initialization.

Note, this simplified interface and CRISP programming is under development and is not yet available.

Design Considerations To deliver a complete Color Rich solution for MicroDisplay customers, it must simplify both hardware and software development. For the hardware, it can standardize the way Color Rich mode is implemented. For the software, it can eliminate the need for separating color planes, and provide automatic conversion from industry standard pixel definitions to MicroDisplay's way of doing things.

The following is a list of functional desires and/or requirements.

Support for varied color depth, allowing tradeoff between color depth and power usage. Minimum of 512 colors.

'Conversion from 8,16, and 24 bits per pixel to MicroDisplay Color Rich.

'Color Palettes and real time Dithering.

Support for two Displays, simultaneously.

System must be capable of video frame-rate image throughput.

Low power consumption while display image is static.

Support for Fonts, Cursors, Windows, etc.

32-bit bus to maximize throughput.

Buffer memory addressing is flexible, avoiding hard coded address maps.

An integrated"Watch-Dog"to protect the liquid crystal display.

CRISP Instruction Start Conditions The majority of CRISP instructions have a field called Start Conditions. See the table below. This field specifies which signal (s) must be"true"before the instruction is allowed to execute. The instruction halts the CRISP processor until the conditions are satisfied. Note that a WatchDog timer can prevent the processor from hanging indefinitely in the event that the specified signals never come"true".

The start condition (s) to be tested are specified in the instruction as a"1"or"true", while conditions to be ignored are"0". Interrupt signals are"true"when they are "asserted"by the MicroDisplay.

This capability allows for precise synchronization of display data transfers between buffer memory, and the MicroDisplay (s). Registers for all devices may also require synchronous updates according to the state of the MicroDisplay.

Table 35 Start Source Signal Notes Condition GPIO3GPIO3"True"when high. GPIO2GPIO2"True"when high. GPIO_1 GPIO_1"True"when high. GPIO0GPIO0"True"when high. [not applicable] WDTO WD timer WatchDog time-out. This signal is useful only if AUXCON register, WDENBL = 0, otherwise time-out will cause an ABORT interrupt. ORQ2 Interrupt"True"when Right MicroDisplay interrupt line is low. ORQ1 Interrupt"True"when Left MicroDisplay interrupt line is low.

CRISP Branch Conditions The CRISP flow control instructions have Branch Conditions, instead of Start Conditions. Branch Conditions are immediately tested, and the instruction executes according to the results of the test. Flow control instructions allow for more complex "real-time"programs, such as automatically updating a cursor's screen position or preparing new host data for display utilizing the time between MicroDisplay field updates.

The branch condition (s) to be tested are specified in the instruction as a"1", while conditions to be ignored are"0". Unlike Start Conditions, Branch Conditions are tested as high or low, not true or false. The actual state of a tested flag or signal is important.

Table 36 Branch Source Signal Notes Condition GPIO3 GPI3 General Purpose Input/Output 3 state. GPIO2GPI2General Purpose Input/Output 2 state. GPIO_1 GPI1 General Purpose Input/Output 1 state. GPIO0 GPI0 General Purpose Input/Output 0 state. TCC-Condition code set by TST instruction. WDTO WD timer WatchDog time-out. This signal is useful only if AUXCON register, WDENBL = 0, otherwise time-out will cause an ABORT interrupt. ORQ2 Interrupt Right MicroDisplay IRQ state. ORQ1 Interrupt Left MicroDisplay IRQ state.

Instruction Set Summary Table 37: Load CRA register

Instruction CRA Reg Address (31: 27) (26: 24) (23: 0) 00000-LDR nnn aaaa aaaa aaaa aaaa aaaa aaaa 00001-STR nnn aaaa aaaa aaaa aaaa aaaa aaaa These instructions move the data found at the specified Address to/from one of up to eight registers in the CRISP.

Data Transfer Operations These instructions move data from one part of memory to another, or to memory mapped devices such as MicroDisplay. Some of the instructions combine source and destination data using Boolean operators.

Table 38 Instruction WDR Addr Move Skip Start Conditions (31: 27) (26) Mode Count Count (7: 0) (25: 24) (23: 16) (15: 8) 00010-MOP b Src Dst cccc cccc ssss ssss g3 g2 gl gO na wd q2 ql 00011-b Src Dst cccc cccc ssss ssss g3 g2 gl gO na wd MOV q2 ql 00100-b Src Dst cccc cccc ssss ssss g3 g2 gl gO na wd NOT q2 ql 00101-b Src Dst cccc cccc ssss ssss g3 g2 gl g0 na wd AND q2 ql 00110-b Src Dst cccc cccc ssss ssss g3 g2 gl g0 na wd XOR q2 ql 00111-b Src Dst cccc cccc ssss ssss g3 g2 gl g0 na wd ORR q2 ql

Immediate Data Operations These instructions operate on 8-bit registers found on the MicroDisplay, Analog Controller, CRISP, and an additional (TBD) device.

Table 39 Instruction Dev Sel Device Address Data Start Conditions (31: 27) (26: 24) (23: 16) (15: 8) (7: 0) 01000-LDC nnn aaaa aaaa dddd g3 g2 gl gO na wd dddd q2 ql 01001-SET nnn aaaa aaaa dddd g3 g2 gl gO na wd dddd q2 ql 01010-CLR nnn aaaa aaaa dddd g3 g2 gl gO na wd dddd q2 ql 01011-TST nnn aaaa aaaa dddd g3 g2 gl gO na wd dddd q2 ql

Flow Control Operations These instructions provide program flow control to create loops and conditional execution.

Table 40 Instruction WDR Offset Address Conditions (31: 27) (26) (25: 8) (7: 0) 01100-BCL b aa aaaa aaaa aaaa aaaa g3 g2 gl gO tc wd q2 ql 01101-b aa aaaa aaaa aaaa aaaa g3 g2 gl gO tc wd BCH q2 ql 01110-HBL b aa aaaa aaaa aaaa aaaa g3 g2 gl gO tc wd q2 ql 01111-b aa aaaa aaaa aaaa aaaa g3 g2 gl gO tc wd HBH q2 ql

Timing Control Operations This instruction provides precise inline timing for display field control.

Table 41 Instruction [not [not used] Delay Count [not used] (31: 27) used] (25: 18) (17: 8) (7: 0) (26) 11111-DLY b-cc cccc cccc g3 g2 gl g0 na wd q2 q1 Note Instructions 10000 through 11110 are undefined and should not be used.

Instruction Set Details Table 42: Load CRA Register Instruction CRA Reg Address (31: 27) (26: 24) (23: 0) 00000-LDR nnn aaaa aaaa aaaa aaaa aaaa aaaa

LDR reg, address 11 address (data)->reg Data found at the Address location specified is loaded to specified register. The details of this instruction's fields may be found below.

Table 43: Store CRA Register Instruction CRA Reg Address (31: 27) (26: 24) (23: 0) 00001-STR nnn aaaa aaaa aaaa aaaa aaaa aaaa

STR reg, address//reg->mem address The contents of the specified register are written to memory at the specified address.

The values designating CRA registers (nnn) are as follows.

Table 44 nnn Register Actual Size-Usage Name 000 Source 24 bits-Starting address value used by"Memory to Pointer MicroDisplay"and"Memory to Memory"operations as the data source memory address. Note: the low order two bits of loaded value are ignored because all transfers must be 32-bit word aligned. 001 Destination 24 bits-Starting address value used by"Memory to Pointer MicroDisplay"and"Memory to Memory"operations as the data destination memory address. Note: the low order two bits of loaded value are ignored because all transfers must be 32-bit word aligned. 010 Transfer 22 bits-Used by"Memory to MicroDisplay"and Count"Memory to Memory"operations as a byte count for the entire transfer operation. The upper two unused bits (of the 24-bit field) are ignored, but should be set to zero for future compatibility. Note: the low order two bits of loaded value are ignored because all transfers are 32-bit data operations. 011 Program 24 bits-This is the CRISP Program counter. An LRI to Counter this register is effectively a JMP indirect. Note: the low order two bits of loaded value are ignored because instructions must be 32-bit word aligned. lxx Undefined These four register addresses are reserved for future expansion.

Data Transfer Operations Table 45 Instruction WDR Addr Mode Move Skip Start Conditions (31: 27) (26) (25: 24) Count Count (7: 0) (23: 16) (15: 8) 00010-MOP b x x cccc cccc ssss ssss na g2 gl wd i2 q2 il ql

MMD wdr, smode=l, dmode=0, mCount, sCount, conditions//Src->Dst Move memory to MicroDisplay. This instruction copies 32-bit words from"Source Address"memory through"End Address", to"Destination Address" MicroDisplay (s). This instruction invokes an optimized data path between the SDRAM and the MicroDisplay. Instruction WDR Addr Mode Move Skip Start Conditions (31: 27) (26) (25 : 24) Count Count (7: 0) (23: 16) (15: 8) 00011-b Src Dst cccc cccc ssss ssss na g2 gl wd i2 q2 MOV il ql

MOV wdr, smode, dmode, mCount, sCount, conditions//Src->Dst This general purpose data move instruction copies 32-bit words from"Source Address"to"Destination Address"according to SRC and DST"Addr Mode" settings.

Below is a breakdown of each of the fields for MOP and MMD.

Table 47 Field Value Function Name or Range WDR b=0 Watchdog is not reset on this operation b=l Watchdog is reset to last loaded full count, WD status is cleared Addr Mode SRC = Source/Destination Addresses are incremented normally 0 DST (e. g. +4 bytes for each 32-bit transfer) for the duration of = 0 the transfer. After each 32-bit move the Source/Destination Address SRC=1 is incremented by 4 (to next word address), and the DST=1"Move Count"is decremented by 1. When"Move Count"is exhausted, it is reset to it's original value, and the Source/Destination Address is incremented by"Skip Count". Move Count 0-OxFF When SRC and/or DST Addr Mode is"1", this value is used to load a countdown counter to track 32-bit transfers. Upon reaching zero, each affected Address register is incremented by"Skip Count", and the counter is reloaded to begin countdown again. Skip Count 0-OxFF When SRC and/or DST Addr Mode is"1", this value is used to increment each affected Address register. Start Condition s xx For each of the 8 start conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must be met (condition is"True") before the instruction is executed. Note: some conditions cannot happen simultaneously. For more information, see"Start Conditions"discussion on page nn.

Table 48 Instruction WDR Addr Mode Move Skip Start Conditions (31: 27) (26) (25: 24) Count Count (7: 0) (23: 16) (15: 8) 00100-b Src Dst cccc cccc ssss ssss na g2 gl wd i2 q2 NOT il ql NOT wdr, smode, dmode, mCount, sCount, conditions// ! Src->Dst

This instruction transfers 32 bit words from"Source Address"through"End Address"to"Destination Address"according to SRC and DST"Addr Mode" settings. The destination data is inverted from the source data. Details of this instruction's fields found below.

Table 49 Instruction WDR Addr Mode Move Skip Start Conditions (31: 27) (26) (25: 24) Count Count (7: 0) (23: 16) (15: 8) 00101-b Src Dst cccc cccc ssss ssss na g2 gl wd i2 q2 AND il ql

AND wdr, smode, dmode, mCount, sCount, conditions//Src & Dst->Dst This instruction transfers 32 bit words from"Source Address"through"End Address"to"Destination Address"according to SRC and DST"Addr Mode" settings. The prior data at the destination address is ANDed with the source data, then stored at the destination address.

Below is a breakdown of each of the fields for NOT and AND.

Table 50 Field Value Function Name or Range WDR b=O Watchdog is not reset on this operation b=l Watchdog is reset to last loaded full count, WD status is cleared Addr Mode SRC = Source/Destination Addresses are incremented normally 0 DST (e. g. +4 bytes for each 32-bit transfer) for the duration of = 0 the transfer. After each 32-bit move the Source/Destination Address SRC=1 is incremented by 4 (to next word address), and the DST=1"Move Count"is decremented by 1. When"Move Count"is exhausted, it is reset to it's original value, and the Source/Destination Address is incremented by"Skip Count". Move Count 0-OxFF When SRC and/or DST Addr Mode is"1", this value is used to load a countdown counter to track 32-bit transfers. Upon reaching zero, each affected Address register is incremented by"Skip Count", and the counter is reloaded to begin countdown again. Skip Count 0-OxFF When SRC and/or DST Addr Mode is"1", this value is used to increment each affected Address register. Start Condition s xx For each of the 8 start conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must be met (e. g. condition is"True") before the instruction is executed. Note: some conditions cannot happen simultaneously. For more information, see"Start Conditions"discussion on page nn.

Table 51 Instruction WDR Addr Move Skip Start Conditions (31: 27) (26) Mode Count Count (7: 0) (25: 24) (23: 16) (15: 8) 00110-b Src Dst cccc cccc ssss ssss na g2 gl wd i2 q2 XOR il ql

XOR wdr, smode, dmode, mCount, sCount, conditions 11 Src ^ Dst-> Dst This instruction transfers 32 bit words from"Source Address"through"End Address"to"Destination Address"according to SRC and DST"Addr Mode" settings. The prior data at the destination address is EXCLUSIVE-ORed with the source data, then stored at the destination address. Details of this instruction's fields found below.

Table 52 Instruction WDR Addr Move Skip Start Conditions (31: 27) (26) Mode Count Count (7: 0) (25: 24) (23: 16) (15: 8) 00111-b Src Dst cccc ssss ssss na g2 gl wd i2 q2 ORR cccc il ql ORR wdr, smode, dmode, mCount, sCount, conditions 11 Src I Dst-> Dst This instruction transfers 32 bit words from"Source Address"through"End Address"to"Destination Address"according to SRC and DST"Addr Mode" settings. The prior data at the destination address is ORed with the source data, then stored at the destination address.

Below is a breakdown of each of the fields for XOR and ORR Table 53 Field Value Function Name or Range WDR b=O Watchdog is not reset on this operation b=l Watchdog is reset to last loaded full count, WD status is cleared Addr Mode SRC = Source/Destination Addresses are incremented normally 0 DST (e. g. +4 bytes for each 32-bit transfer) for the duration of = 0 the transfer. After each 32-bit move the Source/Destination Address SRC=1 is incremented by 4 (to next word address), and the DST=1"Move Count"is decremented by 1. When"Move Count"is exhausted, it is reset to it's original value, and the Source/Destination Address is incremented by"Skip Count". Move Count 0-OxFF When SRC and/or DST Addr Mode is"1", this value is used to load a countdown counter to track 32-bit transfers. Upon reaching zero, each affected Address register is incremented by"Skip Count", and the counter is reloaded to begin countdown again. Skip Count 0-OxFF When SRC and/or DST Addr Mode is"1", this value is used to increment each affected Address register. Start Condition s xx For each of the 8 start conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must be met (e. g. condition is"True") before the instruction is executed. Note: some conditions cannot happen simultaneously. For more information, see"Start Conditions"discussion on page nn.

Immediate Data Operations Table 54

Instruction Dev Sel Device Data Start (31: 27) (26: 24) Address (15: 8) Conditions (23: 16) (7: 0) 01000-LDC nnn aaaa aaaa dddd na I Q W L R dddd G B LDC dSel, dAddr, data, conditions//data-> regAddr This instruction writes the immediate Data to the specified device register address (regAddr). Details of this instruction's fields found below.

Table 55 Instruction Dev Sel Device Data Start Conditions (31: 27) (26 : 24) Address (15: 8) (7 : 0) (23: 16) 01001-SET nnn aaaa aaaa dddd na g2 gl wd i2 q2 dddd il ql SET dSel, dAddr, data, condition 1/data I regAddr--> regAddr

This instruction ORs the immediate Data with the specified device register address (regAddr), then writes the result to that same device register address.

Below is a breakdown of each of the fields for LDC and SET.

Table 56 Field Value Function Name or Range DevSel OOx Registers selected (see also CRISP Control & Status registers, A UXCON) x = left or right device Olx Analog Controller parallel interface selected (see also CRISP Control & Status registers, AUXCON) x = left or right device 100 MicroDisplay Registers selected (see also CRISP Control & Status registers, AUXCON) Both L & R devices are selected. 101 Analog Controller parallel interface selected (see also CRISP Control & Status registers, A UXCON) Both L & R devices are selected. 110 Auxiliary device selected 111 Color Rich ASIC selected (internal 8-bit registers-see page nn) Device Address 0-OxFF 8 bit address of device. (see also CRISP Control & Status registers, A UXCON) Data 0-OxFF 8 bit data to be used with device register. For LDC, this value is written to the device register. For SET, each bit that is a"1"is set in the device register, while"0"bits are left unchanged. Start Condition s xx For each of the 8 start conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must be met (condition is"True") before the instruction is executed. Note: some conditions cannot happen simultaneously. For more information, see"Start Conditions"discussion on page nn.

Table 57 Instruction Dev Sel Device Data Start Conditions (31: 27) (26: 24) Address (15: 8) (7: 0) (23: 16) 01010-CLR nnn aaaa aaaa dddd na g2 gl wd i2 q2 dddd il ql

CLR dSel, dAddr, data, condition// ! data & regAddr--> regAddr This instruction inverts the immediate Data; ANDs the result with the specified device register address (regAddr), then writes the result to that same device register address. Details of this instruction's fields found below.

Table 58 Instruction Dev Sel Device Data Start Conditions (31: 27) (26: 24) Address (15: 8) (7: 0) (23: 16) 01011-TST nnn a aaaa aaaa dddd na g2 gl wd i2 q2 I dddd il ql

TST dSel, dAddr, data, conditions 11 data & regAddr--> TCC (affects This instruction ANDs the immediate Data with specified device register. When the results are OxOO, the"TCC"bit of the condition register is cleared (false). Otherwise it is set (true). The specified device register is unchanged by this operation.

Below is a breakdown of each of the fields for CLR and TST.

Table 59 Field Value Function Name or Range DevSel OOx MicroDisplay Registers selected (see also CRISP Control & Status registers, A UXCON) x = left or right device Olx Analog Controller parallel interface selected (see also CRISP Control & Status registers, A UXCON) x = left or right device 100 MicroDisplay Registers selected (see also CRISP Control & Status registers, A UXCON) Both L & R devices are selected. 101 Analog Controller parallel interface selected (see also CRISP Control & Status registers, A UXCON) Both L & R devices are selected. 110 Auxiliary device selected 111 Color Rich ASIC selected (internal 8-bit registers-see page nn) Device Address 0-OxFF 8-bit address of device registers. (see also CRISP Control & Status registers, AUXCON) Data 0-OxFF 8 bit data to be used with device register. For CLR, each bit that is a"1"is cleared in the device register, while"0" bits are left unchanged. For TST, this value is ANDed with the device register and the TCC bit is set if the result is not zero, otherwise TCC is cleared. Start Condition s xx For each of the 8 start conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must be met (e. g. condition is"True") before the instruction is executed. Note: some conditions cannot happen simultaneously. For more information, see"Start Conditions"discussion on page nn.

Flow Control Operations Table 60

Instruction WDR Offset Address Conditions (31: 27) (26) (26: 8) (7: 0) 01100-BCL b aa aaaa aaaa aaaa g3 g2 gl gO tc wd aaaa q2 ql BCL (Branch if Condition LOW)//if LOW: PC += Offset Address The selected conditions are tested immediately against the source signals to determine if the branch is to be taken or not. If all conditions are not met, the branch

is taken. Non-selected conditions are ignored; thus a BCL with no conditions would always branch. If the selected conditions are met, the instruction processing continues at the next subsequent instruction.

Table 61 Instruction WDR Offset Address Conditions (31: 27) (26) (26: 8) (7: 0) 01101-BCH b aa aaaa aaaa aaaa g3 g2 gl gO tc wd aaaa q2 ql

BCH (Branch if Condition HIGH)//if HIGH: PC += Offset Address The selected conditions are tested immediately against the source signals to determine if the branch is to be taken or not. If all conditions are met, the branch is taken. Non-selected conditions are ignored; thus a BCH with no conditions would always branch. If the selected conditions are not met, the instruction processing continues at the next subsequent instruction.

Here is a description for each of the fields of BCH and BCL.

Table 62 Field Value Function Name or Range WDR b=O Watchdog is not reset on this operation b=l Watchdog is reset to last loaded full count, WD status is cleared Offset Address OxOOOO Positive offset, added to the Program Counter (PC) if the 0 conditions of the instruction are met. Since the through instructions are 32-bit word aligned, this value represents OxIFFF the word, not byte, offset. Thus a OxOOOOO would not F increase the (post increment) PC, while Ox00002 would add 8 to the (post increment) PC, skipping two instructions. Ox2000 Negative offset, added to the Program Counter (PC) if 0 the conditions of the instruction are met Since the through instructions are 32-bit word aligned, this value represents Ox3FFF the word, not byte, offset. Thus a Ox3FFFF would F decrease the (post increment) PC by 4, causing the Branch instruction to loop until the conditions were invalid. A value of Ox3FFFO would adjust the PC to point to the 1 5th prior instruction. Condition s xx For each of the 8 conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must be met if the Branch is to be executed, otherwise the PC continues with the next sequential instruction. Note: This is different from"Start Conditions"used in other instructions in that the condition test is immediate and only happens once. For more information, see"Flow Control Conditions"discussion on page nn.

Table 63 Instruction WDR Offset Address Conditions (31: 27) (26) (26: 8) (7: 0) 01110-HBL b aa aaaa aaaa aaaa g3 g2 gl gO tc wd aaaa q2 ql

HBL (Halt if Condition LOW)//if LOW: HALT, ELSE PC +=Offset The selected conditions are tested immediately against the source signals to determine if the branch is to be taken or not. If all conditions are not met, the instruction processing is halted. Non-selected conditions are ignored; thus a HBL with no conditions would halt instruction processing. If the selected conditions are met, the instruction processing continues at the offset address.

Table 64 Instruction WDR Offset Address Conditions (31: 27) (26) (26: 8) (7: 0) 01111-HBH b aa aaaa aaaa aaaa g3 g2 gl gO tc wd aaaa q2 ql HBT (Halt if Condition HIGH) 11 if HIGH: HALT, ELSE PC +=Offset The selected conditions are tested immediately against the source signals to determine if the branch is to be taken, or not. If all conditions are met, the instruction processing is halted. Non-selected conditions are ignored; thus a HBH with no conditions would halt instruction processing. If the selected conditions are not met, the instruction processing continues at the offset address.

Below is a description for each of the fields of HBH and HBL.

Table 65 Field Value Function Name or Range WDR b=O Watchdog is not reset on this operation b=l Watchdog is reset to last loaded full count, WD status is cleared Offset Address OxOOOO Positive offset, added to the Program Counter (PC) if the 0 conditions of the instruction are met. Since the through instructions are 32-bit word aligned, this value represents Ox 1 FFF the word, not byte, offset. Thus a OxOOOOO would not F increase the (post increment) PC, while Ox00002 would add 8 to the (post increment) PC, skipping two instructions. Ox2000 Negative offset, added to the Program Counter (PC) if 0 the conditions of the instruction are met. Since the through instructions are 32-bit word aligned, this value represents Ox3FFF the word, not byte, offset. Thus a Ox3FFFF would F decrease the (post increment) PC by 4, causing the Branch instruction to loop until the conditions were invalid. A value of Ox3FFFO would adjust the PC to point to the 15th prior instruction. Condition s xx For each of the 8 conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must if the Halt is to be executed, otherwise the PC continues with the instruction at PC+Offset Address. Note: This is different from"Start Conditions"used in other instructions in that the condition test is immediate and only happens once. For more information, see"Flow Control Conditions" discussion on page nn.

Table 66 Instruction [not [not used] Delay Count [not used] (31: 27) used] (25: 18) (17: 8) (7: 0) (26) 11111 - DLY - - cc cccc cccc -

DLY (Delay)//Wait awhile before proceeding.

The delay instruction is used when the CRISP processor, rather than the MicroDisplay's built-in timing parameters are controlling all display timing.

Below is a description for each of the fields of DLY.

Table 67

Field Value Function Name or Range Delay Time Ox000 The period for which the flow of instructions is delayed through = ( (this hex value+1) x 256 clocks); under normal Ox3FF operating conditions this is about 4 uSec per count.

Color Rich Internal System Processor (CRISP) Introduction The Color Rich Internal System Processor, or CRISP, is a very small instruction set processor used primarily to drive DMA transfers from memory to the Microdisplay.

CRISP is the part of the Color Rich ASIC (CRASIC) that programmatically controls the operation of the Microdisplay and AIC.

Experience with the Dol-plus demonstrated the power and flexibility of"DMA List Processing". CRISP leverages that experience and expands processing capabilities with the addition of a few Boolean operations and a simplified dithering algorithm. These new operations can be applied to image data as it is moved from memory to memory or memory to the Microdisplay.

CRISP is designed to handle directly a simple 512-Color mode of operation, but is flexible enough to also manage higher color operations. With its simple instruction set it can also vastly simplify cursor tracking, fonts, multiple screen and window management for the Host.

In addition, it may be possible to create a simplified (host) software interface to enable display module customers to develop products without having to understand the"nitty- gritty"details of the Microdisplay or AIC. They wouldn't even need to learn the internal operation of CRISP, only the external interface of the CRASIC. The CRISP programming can either be"booted"from an I^2C EEPROM, or downloaded by the host driver at initialization.

Design Considerations For the CRASIC to deliver a complete Color-Rich solution for customers, it must simplify both hardware and software development. For the hardware side, it can standardize the way Color Rich mode is implemented. For the software side, it can eliminate the need for separating out color plains and provide automatic conversion from industry standard pixel definitions to the Display Sytem's way of doing things.

The following is a list, in no particular order, of functional desires and/or requirements: Support for at least 512 Colors (minimum to claim"Color Rich") Conversion from 8,16, and 24 bits per pixel to Color Rich.

Color Palettes and real time Dithering Support for two Displays, simultaneously

System must be capable of video frame-rate image throughput Low power consumption while display image is static Support for Fonts, Cursors, Windows, etc...

Flexible enough to allow higher-than-512 Color modes (e. g. more than 3 bits used per color plain).

Other considerations, in no particular order: Schedule (e. g. keep it simple, stupid) 32-bit bus to maximize throughput-instructions would ideally be 32-bit as well.

Keeping buffer memory addressing flexible (e. g. avoiding hard coded address maps for what part of RAM is used for what.) An integrated"Watch-Dog"to avoid burning the display Avoid necessity of complex calculations, if possible.

CRISP Memory Map The CRISP memory map refers to RAM and memory mapped devices (AIC, etc.) controlled directly by the Color Rich ASIC. RAM is used primarily for CRISP programs, cursors and display buffers, but could also be used as buffer by the host processor for fonts, audio data, etc.

The external interface for the Color Rich ASIC provides the means for a host processor to access the entire contents of the CRISP memory. The CRASIC external interface registers (some of which control CRISP itself) are also fully accessible to CRISP programs.

The memory map is as follows: Table 68 Addr Range Device Name Description 0x000000 SDRAM Must be at least 512K to support "Color Rich", but can thru be as large as 8Mb. Used for display buffers, cursors, palettes, etc. 0x7FFFFF 0x800000 Display memory [Right Organization of memory depends on the mode. thru Device] Ox8FFFFF 0x900000 Display memory [Left Organization of memory depends on mode. thru Device] Ox9FFFFF OxAOOOOO Display memory [Both Organization of memory depends on mode. thru Devices-Write Only] OxAFFFFF OxBOOOOO Auxiliary device This is undefined at this time. It might be used for audio, or as a high-speed data transport device such as thru Fire-Wire that could benefit from CRISP's DMA OXB7FFFF capabilities. It could be serviced by display memory FIFO, using all the same signals. OxB80000 DAPPER Color Palette (8-All 256 locations are used to hold the palette. These registers are write-only. thru bit to 24-bit Lookup Table) OxFFFFFF OxCOOOOO AIC registers Up to 256 8-bit registers. Registers appear in least significant byte of 32-bit word access. thru [Left Device] OxC7FFFF OxC80000 Registers up to 256 8-bit registers. Registers appear in least significant byte of 32-bit word access. thru [Left Device] 0xCFFFFF OxDOOOOO AIC registers up to 256 8-bit registers. Registers appear in least significant byte of 32-bit word access. thru [Right Device] OxD7FFFF OxD80000 Registers Up to 256 8-bit registers. Registers appear in least significant byte of 32-bit word access. thru [Right Device] OxDFFFFF OxEOOOOO AIC registers up to 256 8-bit registers. Registers appear in least .. significant byte of 32-bit word access. thru [Both Devices-Write Only] OxE7FFFF OxE80000 Registers Up to 256 8-bit registers. Registers appear in least significant byte of 32-bit word access. thru [Both Devices-Write Only] OxEFFFFF OxFOOOOO CRISP and DAPPER Up to 256 registers. Registers may be up to 32-bits. While primarily used by CRISP, these registers are thru registers fully visible to the Host processor. OxF7FFFF OxF80000 CRASIC External interface These are always readily accessible to the host processor. They are only accessible to CRISP thru Data thru and GLU peripheral registers Transfer operations such as the MOV instruction. OxBFFFFF

CRISP Instruction Start Conditions Note: the pin definition for the FPGA that is to emulate the Color Rich ASIC does not include all the specific display system signals discussed here. It does, however, include eight generic I/O bits that could be used for these signals. All of these signals could be detected using the system's interrupt capability, but that increases the complexity of CRISP programming.

The majority of CRISP instructions have a field called Start Conditions. This field specifies which signal (s) must be"true"before the instruction is allowed to execute.

The instruction, in essence, halts the CRISP processor until the conditions are satisfied. Note, however, that a Watch-Dog timer can prevent the processor from hanging indefinitely in the event that the specified signals never come"true".

The start condition (s) to be tested are specified in the instruction as a"1", while conditions to be ignored are"0".

With the exception of the ITO signals, all signals are"true"when they are"asserted" by the Display System, and Watch-Dog timer. The ITO signals are considered "true"each time it changes polarity (see table below).

This capability allows for precise synchronization of display data transfers between buffer memory and the Microdisplay (s). Registers for all devices may also require synchronous updates according to the state of the Display System.

Table 69 Start Source Signal Notes Condition GPIO3GPIO3"True"when high. GPIO2GPIO2"True"when high. GPIO_1 GPIO_1"True"when high. GPIO0GPIO0"True"when high. WDTO WD timer Watch-Dog timeout. This signal is useful only if AUXCON register, WDENBL = 0 ; otherwise timeout will cause an ABORT interrupt. ORQ2 Interrupt"True"when Secondary interrupt line is low. ORQ1 Interrupt"True"when Primary interrupt line is low.

CRISP Branch Conditions Note: the pin definition for the FPGA that is to emulate the Color Rich ASIC does not include all the specific signals discussed here. It does, however, include eight generic I/O bits that could be used for these signals. All of these signals could be detected using register accesses with CRISP's TST instruction, but that increases the complexity of CRISP programming.

The CRISP flow control instructions have Branch Conditions, instead of Start Conditions. Branch Conditions are immediately tested and the instruction executes according to the results of the test (see Flow Control Instructions for details). Flow

control instructions allow for more complex"real-time"programs, such as automatically updating a cursor's screen position or preparing new host data for display, utilizing the time between field updates.

The branch condition (s) to be tested are specified in the instruction as a"1", while conditions to be ignored are"0". Unlike Start Conditions, Branch Conditions are tested as either"High"or"Low", not"True"or"False". The actual state of a tested flag or signal is what is important.

Table 70 Branch Source Signal Notes Condition TCC-Condition code set by TST instruction. GPIO3 GPI3 Genera ! Purpose Input/Output 3 state. GPIO2 GPI2 General Purpose Input/Output 2 state. GPIO_1 GPI_1 General Purpose Input/Output 1 state. GPIO_O GPI_O General Purpose Input/Output 0 state. WDTO WD timer Watch-Dog timeout. This signal is useful only if AUXCON register, WDENBL = 0 ; otherwise timeout will cause an ABORT interrupt. ORQ2 Interrupt Secondary IRQ state. ORQ1 Interrupt Primary IRQ state.

INSTRUCTION SET SUMMARY Load CRA register Table 71 Spar Instruction CRA Reg Address e (30 : 27) (26: 24) (23 : 0) (31) 0 0000-nnn aaaa aaaa aaaa aaaa aaaa aaaa LDR 0 0001-nnn aaaa aaaa aaaa aaaa aaaa aaaa STR

These instruction moves the data found at the specified Address to/from one of up to eight registers in the CRISP. In this discussion only 4 registers are defined.

Data Transfer Operations Table 72 Spar Instruction WD Addr Move Count Skip Start Conditions e (30: 27) R Mode (23: 16) Count (7: 0) (31) (26) (25: 24) (15: 8) 0 0010-b Src Dst cccc cccc ssss ssss na g3 g2 g1 g0 wd q2 q1 MOP 0 0011-b Src Dst cccc ccoc ssss ssss na g3 g2 gl gO wd q2 ql MOV 0 0100-b Src Dst cccc cccc ssss ssss na g3 g2 gl gO wd q2 ql NOT 0 0101-b Src Dst cccc cccc ssss ssss na g3 g2 gl go wd q2 ql AND 0 0110 - b Src Dst cccc cccc ssss ssss na g3 g2 g1 g0 wd q2 q1 XOR 0 0111-b Src Dst cocc cccc ssss ssss na g3 g2 gl gO wd q2 ql ORR

These instructions move data from one part of memory to another, or to memory mapped devices such as the Microdisplay. Some of the instructions combine source and destination data using Boolean operators. Immediate Data Operations (for control registers on the Microdisplay, AIC, CRISP, AUX) Table 73

Spar Instruction Dev Sel Device Address Data Start Conditions e (30: 27) (26: 24) (23: 16) (15: 8) (7: 0) (31) 0 1000-nnn aaaa aaaa dddd dddd na g3 g2 g1 g0 wd q2 q1 LDC 0 1001-nnn aaaa aaaa dddd dddd na g3 g2 g1 g0 wd q2 q1 SET 0 1010-nnn aaaa aaaa dddddddd nag3g2glgOwdq2ql CLR 0 1011-nnn aaaa aaaa dddd dddd na g3 g2 g1 g0 wd q2 q1 TST These instructions operate on 8-bit registers found on the Microdisplay, AIC, CRISP, and an additional. (TBD) device.

Flow Control Operations Table 74 Spar Instruction WD Offset Address Conditions e (30: 27) R (25: 8) (7: 0) (31) (26) 0 1100-b aa aaaa aaaa aaaa aaaa tc g3 g2 gl go wd q2 ql BCL 0 1101-b aa aaaa aaaa aaaa aaaa tc g3 g2 gl gO wd q2 ql BCH

These instructions provide program flow control to create loops and conditional execution.

Timing Control Operations Table 75 Spar Instruction WD [not used] Count Start Conditions e (30: 27) R (25: 16) (15: 8) (7: 0) (31) (26) 0 110 - b - cccc cccc na g3 g2 g1 g0 wd q2 q1 DLY 111 - b - - na g3 g2 g1 g-0 wd q2 q1 NOP

INSTRUCTION SET DETAILS Load CRA Register Table 76 Spar Instruction CRA Reg Address e (30: 27) (26: 24) (23: 0) (31) 0 0000-nnn aaaa aaaa aaaa aaaa aaaa aaaa LDR

LDR reg, address 11 address (data)->reg Data found at the Address location specified is loaded to specified register. The details of each of this instruction's fields may be found below.

Store CRA Register Table 77

Spar Instruction CRA Reg Address e (30: 27) (26: 24) (23: 0) (31) 0 0001-nnn aaaa aaaa aaaa aaaa aaaa aaaa STR STR reg, address//reg->mem address The contents of the specified register is written to memory at the specified address.

Illustrative values designating CRA registers (nnn) are as follows: Table 78 nnn Register Actual Size-Usage Name 000 Source 24 bits-Starting address value used by"Memory to Pointer Microdisplay"and"Memory to Memory"operations as the data source memory address. Note : the low order two bits of loaded value are ignored because all transfers must be 32-bit word aligned. 001 Destination 24 bits-Starting address value used by"Memory to Pointer Microdisplay"and"Memory to Memory"operations as the data destination memory address. Note : the low order two bits of loaded value are ignored because all transfers must be 32-bit word aligned. 010 Transfer 22 bits-Used by"Memory to Microdisplay"and Count"Memory to Memory"operations as a byte count for the entire transfer operation. The upper 2 unused bits (of the 24-bit field) are ignored, but should be set to zero for future compatibility. Note : the low order two bits of loaded value are ignored because all transfers are 32- bit data operations. Oil Program 24 bits-This is the CRISP Program counter. An LRI to Counter this register is effectively a JMP indirect. Note : the low order two bits of loaded value are ignored because instructions must be 32-bit word aligned. lxx Undefined These four register addresses are reserved for future expansion.

Data Transfer Operations Table 79 Spar Instruction WD Addr Move Count Skip Start Conditions e (30: 27) R Mode (23: 16) Count (7: 0) (31) (26) (25: 24) (15: 8) 0 0010-b Src Dst cccc cccc ssss ssss na g2 gl wd i2 q2 il ql MOV

MOV wdr, smode, dmode, mCount, sCount, conditions//Src->Dst This instruction copies"Transfer Count"bytes from"Source Address"memory to "Destination Address"memory according to SRC and DST"Addr Mode"settings.

The destination data should be the same as the source data. The details of each of this instruction's fields may be found below.

Table 80 Spar Instruction WD Addr Move Count Skip Start Conditions e (30: 27) R Mode (23: 16) Count (7: 0) (31) (26) (25: 24) (15: 8) 0 0011-b Src Dst cccc cccc ssssssss na g2 gl wd i2 q2 il ql MPI

MPI wdr, smode, dmode, mCount, sCount, conditions//Src->Dst "Move and Process Image"passes source data through the Palette and Bias-grid before depositing results at the destination. Note that the destination will contain only half the number of words as the source, due to this process (see details of Pallets and Dithering on page nn).

Here is a breakdown of each of the fields for MOV and MPI: Table 81 Field Value Function Name or Range WDR b=O Watchdog is not reset on this operation b=l Watchdog is reset to last loaded full count, WD status is cleared Addr Mode SRC = Source/Destination Addresses are incremented normally 0 DST (e. g. +4 bytes for each 32-bit transfer) for the duration of = 0 the transfer. After each 32-bit move the Source/Destination Address SRC=1 is incremented by 4 (to next word address), and the DST=1"Move Count"is decremented by 1. When"Move Count"is exhausted, it is reset to it's original value, and the Source/Destination Address is incremented by"Skip Count". Move Count 0-OxFF When SRC and/or DST Addr Mode is"1", this value is used to load a count-down counter to track 32-bit transfers. Upon reaching zero, each affected Address register is incremented by"Skip Count", and the counter is reloaded to begin count-down again. Skip Count 0-OxFF When SRC and/or DST Addr Mode is"1", this value is used to increment each affected Address register. Start Condition s xx For each of the 8 start conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must be met (e. g. condition is"True") before the instruction is executed. Note : some conditions cannot happen simultaneously. For more information, see"Start Conditions"discussion on page nn.

Table 82 Spar Instruction WD Addr Move Count Skip Start Conditions e (30 : 27) R Mode (23 : 16) Count (7: 0) (31) (26) (25: 24) (15: 8) 0 0100-b Src Dst ccoc cccc ssss ssss na g2 gl wd i2 q2 il ql NOT NOT wdr, smode, dmode, mCount, sCount, conditions// ! Src->Dst

This instruction transfers"Transfer Count"bytes from"Source Address"memory to "Destination Address"memory according to SRC and DST"Addr Mode"settings.

The destination data is INVERTed from the source data. The details of each of this instruction's fields may be found below.

Table 83 Spar Instruction WD Addr Move Count Skip Start Conditions e (30: 27) R Mode (23: 16) Count (7: 0) (31) (26) (25: 24) (15 : 8) 0 0101-b Src Dst cccc cccc ssss ssss na g2 gl wd i2 q2 a ql AND

AND wdr, smode, dmode, mCount, sCount, conditions//Src & Dst->Dst This instruction transfers"Transfer Count"bytes from"Source Address"memory to "Destination Address"memory according to SRC and DST"Addr Mode"settings.

The prior data at the destination address is ANDed with the source data, then stored at the destination address.

Here is a breakdown of each of the fields for NOT and AND: Table 84 Field Value Function Name or Range WDR b=O Watchdog is not reset on this operation b=l Watchdog is reset to last loaded full count, WD status is cleared Addr Mode SRC = Source/Destination Addresses are incremented normally 0 DST (e. g. +4 bytes for each 32-bit transfer) for the duration of = 0 the transfer. After each 32-bit move the Source/Destination Address SRC=1 is incremented by 4 (to next word address), and the DST=1"Move Count"is decremented by 1. When"Move Count"is exhausted, it is reset to it's original value, and the Source/Destination Address is incremented by"Skip Count". Move Count 0-OxFF When SRC and/or DST Addr Mode is"1", this value is used to load a count-down counter to track 32-bit transfers. Upon reaching zero, each affected Address register is incremented by"Skip Count", and the counter is reloaded to begin count-down again. Skip Count 0-OxFF When SRC and/or DST Addr Mode is"1", this value is used to increment each affected Address register. Start Condition s xx For each of the 8 start conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must be met (e. g. condition is"True") before the instruction is executed. Note : some conditions cannot happen simultaneously. For more information, see"Start Conditions"discussion on page nn.

Data Transfer Operations (continued) Table 85

Spar Instruction WD Addr Move Count Skip Start Conditions e (30: 27) R Mode (23: 16) Count (7: 0) (31) (26) (25: 24) (15: 8) 0 0110-b Src Dst cccc cccc ssssssss na g2 gl wd i2 q2 il ql XOR XOR wdr, smode, dmode, mCount, sCount, conditions 11 Src A Dst-> Dst This instruction transfers"Transfer Count"bytes from"Source Address"memory to "Destination Address"memory according to SRC and DST"Addr Mode"settings.

The prior data at the destination address is EXCLUSIVE-ORed with the source data, then stored at the destination address. The details of each of this instruction's fields may be found below.

Table 86 Spar Instruction WD Addr Move Count Skip Start Conditions e (30: 27) R Mode (23: 16) Count (7: 0) (31) (26) (25: 24) (15: 8) 0 0111-b Src Dst cccc cccc ssss ssss na g2 gl wd i2 q2 il ql ORR ORR wdr, smode, dmode, mCount, sCount, conditions 11 Src I Dst-> Dst This instruction transfers"Transfer Count"bytes from"Source Address"memory to "Destination Address"memory according to SRC and DST"Addr Mode"settings.

The prior data at the destination address is ORed with the source data, then stored at the destination address. Here is a breakdown of each of the fields for XOR and ORR Table 87 Field Value Function Name or Range WDR b=O Watchdog is not reset on this operation b=l Watchdog is reset to last loaded full count, WD status is cleared Addr Mode SRC = Source/Destination Addresses are incremented normally 0 DST (e. g. +4 bytes for each 32-bit transfer) for the duration of = 0 the transfer. After each 32-bit move the Source/Destination Address SRC=1 is incremented by 4 (to next word address), and the DST=1"Move Count"is decremented by 1. When"Move Count"is exhausted, it is reset to it's original value, and the Source/Destination Address is incremented by"Skip Count". Move Count 0-OxFF When SRC and/or DST Addr Mode is"1", this value is used to load a count-down counter to track 32-bit transfers. Upon reaching zero, each affected Address register is incremented by"Skip Count", and the counter is reloaded to begin count-down again. Skip Count 0-OxFF When SRC and/or DST Addr Mode is"1", this value is used to increment each affected Address register. Start Condition s xx For each of the 8 start conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must be met (e. g. condition is"True") before the instruction is executed.Note : some conditions cannot happen simultaneously. For more information, see"Start Conditions"discussion on page nn.

Immediate Data Operations Table 88

Spar Instruction Dev Sel Device Address Data Start Conditions e (30: 27) (26: 24) (23: 16) (15: 8) (7: 0) (31) 0 1000-nnn aaaa aaaa dddd dddd na I Q W L R G B LDC LDC dSel, dAddr, data, conditions//data-> regAddr This instruction writes the immediate Data to the specified device register address (regAddr). The details of each of this instruction's fields may be found below.

Table 89 Spar Instruction Dev Sel Device Address Data Start Conditions e (30: 27) (26: 24) (23: 16) (15: 8) (7: 0) (31) O 1001-nnn aaaa aaaa dddd dddd na g2 gl wd i2 q2 il ql SET

SET dSel, dAddr, data, conditions 11 data I regAddr--> regAddr This instruction ORs the immediate Data with the specified device register address (regAddr), then writes the result to that same device register address.

Here is a breakdown of each of the fields for LDC and SET: Table 90 Field Value Function Name or Range DevSel OOx Registers selected (see also CRISP Control & Status registers, A UXCON) x = left or right device Olx AIC parallel interface selected (see also CRISP Control & Status registers, A UXCON) x = left or right device 100 Registers selected (see also CRISP Control & Status registers, A UXCON) Both L & R devices are selected. 101 AIC parallel interface selected (see also CRISP Control & Status registers, A UXCON) Both L & R devices are selected. 110 Auxiliary device selected 111 Color Rich ASIC selected (internal 8-bit registers-see page nn) Device Address 0-OxFF 8 bit address of device. (see also CRISP Control & Status registers, A UXCON) Data 0-OxFF 8 bit data to be used with device register. For LDC, this value is written to the device register. For SET, each bit that is a"1"is set in the device register, while"0"bits are left unchanged. Start Condition s xx For each of the 8 start conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must be met (e. g. condition is"True") before the instruction is executed. Note : some conditions cannot happen simultaneously. For more information, see"Start Conditions"discussion on page nn.

Immediate Data Operations (continued) Table 91

Spar Instruction Dev Sel Device Address Data Start Conditions e (30: 27) (26: 24) (23: 16) (15: 8) (7: 0) (31) 0 1010-nnn aaaa aaaa dddd dddd na g2 gl wd i2 q2 il ql CLR CLR dSel, dAddr, data, conditions// ! data & regAddr--> regAddr This instruction inverts the immediate Data, ANDs the result with the specified device register address (regAddr), then writes the result to that same device register address. The details of each of this instruction's fields may be found below.

Table 92 Spar Instruction Dev Sel Device Address Data Start Conditions e (30: 27) (26: 24) (23: 16) (15: 8) (7: 0) (31) 0 1011 - nnn a aaaa aaaa dddd dddd na g2 gl wd i2 q2 il ql TST

TST dSel, dAddr, data, conditions 11 data & regAddr--> TCC (affects This instruction ANDs the immediate Data with specified device register. When the results are Ou00 the"TCC"bit of the condition register is cleared (false), otherwise it is set (true). The specified device register is unchanged by this operation.

Here is a breakdown of each of the fields for CLR and TST: Table 93 Field Value Function Name or Range DevSel OOx Registers selected (see also CRISP Control & Status registers, A UXCON) x = left or right device Olx AIC parallel interface selected (see also CRISP Control & Status registers, A UXCON) x = left or right device 100 Registers selected (see also CRISP Control & Status registers, A UXCON) Both L & R devices are selected. 101 AIC parallel interface selected (see also CRISP Control & Status registers, A UXCON) Both L & R devices are selected. 110 Auxiliary device selected 111 Color Rich ASIC selected (internal 8-bit registers-see page nn) Device Address 0-OxFF 8 bit address of device registers. (see also CRISP Control & Status registers, A UXCON) Data 0-OxFF 8 bit data to be used with device register. For CLR, each bit that is a"1"is cleared in the device register, while"0" bits are left unchanged. For TST, this value is ANDed with the device register and the TCC bit is set if the result is not zero, otherwise TCC is cleared. Start Condition s xx For each of the 8 start conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must be met (e. g. condition is"True") before the instruction is executed. Note: some conditions cannot happen simultaneously. For more information, see "Start Conditions"discussion on page nn.

Flow Control Operations Table 94

Spar Instruction WD Offset Address Conditions e (30: 27) R (26: 8) (7: 0) (31) (26) 0 1100-b aa aaaa aaaa aaaa aaaa tc g2 gi wd i2 q2 in qu BCL

BCL (Branch if Condition LOW)//if LOW: PC += Offset Address The selected conditions are tested immediately against the source signals to determine if the branch is to be taken or not. If all conditions are not met, the branch is taken. Non-selected conditions are ignored, thus a BCL with no conditions would always branch. If the selected conditions are met, the instruction processing continues at the next subsequent instruction.

Table 95 Spar Instruction WD Offset Address Conditions e (30: 27) R (26: 8) (7: 0) (31) (26) 0 1101-b aa aaaa aaaa aaaa aaaa tc g2 gl wd i2 q2 i 1 q I BCH BCH (Branch if Condition HIGH) 11 if HIGH: PC += Offset Address The selected conditions are tested immediately against the source signals to determine if the branch is to be taken or not. If all conditions are met, the branch is taken. Non-selected conditions are ignored, thus a BCH with no conditions would always branch.. If the selected conditions are not met, the instruction processing continues at the next subsequent instruction.

Here is a description for each of the fields of BCH and BCL: Table 96 Field Value Function Name or Range WDR b=O Watchdog is not reset on this operation b=l Watchdog is reset to last loaded full count, WD status is cleared Offset Address OxOOOO Positive offset, added to the Program Counter (PC) if the 0 thru conditions of the instruction are met. Since the OxIFFF instructions are 32-bit word aligned, this value represents F the word, not byte, offset. Thus a OxOOOOO would increase the (post increment) PC not at all, while Ox00002 would add 8 to the (post increment) PC... skipping 2 instructions. Ox2000 Negative offset, added to the Program Counter (PC) if 0 thru the conditions of the instruction are met. Since the Ox3FFF instructions are 32-bit word aligned, this value represents F the word, not byte, offset. Thus a Ox3FFFF would decrease the (post increment) PC by 4, causing the Branch instruction to loop until the conditions were invalid. A value of Ox3FFFO would adjust the PC to point to the 15th prior instruction. Condition s xx For each of the 8 conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must be met if the Branch is to be executed, otherwise the PC continues with the next sequential instruction. Note : This is differentfrom"Start Conditions"used in other instructions in that the condition test is immediate and only happens once. For more information, see"Flow Control Conditions"discussion on page nn.

Table 97 Spar Instruction WD Offset Address Conditions e (30: 27) R (26: 8) (7: 0) (31) (26) 0 1110-b aa aaaa aaaa aaaa aaaa tc g2 gl wd i2 q2 i I q I HBL

HBL (Halt if Condition LOW)//if LOW: HALT, ELSE PC +=Offset Address The selected conditions are tested immediately against the source signals to determine if the branch is to be taken or not. If all conditions are not met, the instruction processing is Halted. Non-selected conditions are ignored, thus a HBL with no conditions would Halt instruction processing. If the selected conditions are met, the instruction processing continues at the offset address.

Table 98 Spar Instruction WD Offset Address Conditions e (30: 27) R (26: 8) (7: 0) (31) (26) 0 1111-b aa aaaa aaaa aaaa aaaa na g2 gl wd i2 q2 i I q I HBH HBT (Halt if Condition HIGH) 11 if HIGH: HALT, ELSE PC +=Offset Address The selected conditions are tested immediately against the source signals to determine if the branch is to be taken or not. If all conditions are met, the instruction processing is Halted. Non-selected conditions are ignored, thus a HBH with no conditions would Halt instruction processing. If the selected conditions are not met, the instruction processing continues at the offset address. Here is a description for each of the fields of HBH and HBL : Table 99 Field Value Function Name or Range WDR b=O Watchdog is not reset on this operation b=l Watchdog is reset to last loaded full count, WD status is cleared Offset Address OxOOOO Positive offset, added to the Program Counter (PC) if the 0 thru conditions of the instruction are met. Since the OxIFFF instructions are 32-bit word aligned, this value represents F the word, not byte, offset. Thus a OxOOOOO would increase the (post increment) PC not at all, while Ox00002 would add 8 to the (post increment) PC... skipping 2 instructions. Ox2000 Negative offset, added to the Program Counter (PC) if 0 thru the conditions of the instruction are met. Since the Ox3FFF instructions are 32-bit word aligned, this value represents F the word, not byte, offset. Thus a Ox3FFFF would decrease the (post increment) PC by 4, causing the Branch instruction to loop until the conditions were invalid. A value of Ox3FFFO would adjust the PC to point to the 1 5th prior instruction. Condition s xx For each of the 8 conditions, a"0"denotes a"Don't Care"while a"1"indicates the condition must if the Halt is to be executed, otherwise the PC continues with the instruction at PC+Offset Address. Note : This is different from"Start Conditions"used in other instructions in that the condition test is immediate and only happens once. For more information, see"Flow Control Conditions"discussion on page nn.

CRASIC External Addressing CRASIC Chip Select Asserted Table 100 A23 A22 A21 A20 A19 A18 Select Data Bits Output Addr. Lines 0 X X X X X SDRAM D31 : D0 A22 : A2 (8Mb) 1 0 0 0 X X SRAM [Right] D31 : D0 A19=0, A18: A2 1 0 0 1 X X SRAM [Left] D31 DO Al9=0, A18: A2 1 0 1 X X X SRAM [Both*] D31 : D0 A19=0, A18: A2 1 1 0 0 0 X CRISP** D31 : DO A11 : A2** 1 1 0 0 1 X CRASIC Regs D31 : D0 A5 : A2 1 1 0 1 0 X Auxiliary Dev D31 : DO A19 : A2 (512Kb) 1 1 0 1 1 X I^2C D31 : DO ? 1 1 1 0 0 0 Regs [Right] D7: D0 A19-1, A11:A2 1 1 1 0 0 1 Regs [Left] D7: D0 A19-1, A11 : A2 1 1 1 0 1 X Regs [Both*] D7: D0 A19-1, A11 : A2 1 1 1 1 0 0 AIC Regs [Right] D7: D0 A11-A2 1 1 1 1 0 1 AIC Regs [Left] D7: DOA11-A2 1 1 1 1 1 X AIC Regs [Both*] D7: D0 A11-A2

* Write Only ** Not normally accessible to external processor. These register are READ ONLY while CRISP is in RUN mode (see CCR). They

may be changed by the Host CPU only while CRISP is in HOLD mode.

CRASIC Registers Selected (All registers are"Little Endian") Table 101 A5 A4 A3 A2 Register Data Bits Remarks 0 0 0 0 0 DPTR D23 : D2 Auto Incr Addr Ptr used with IDAT 0001IDATD31 : DOData R/W to Addr pointed to by DPTR 0010CSEMD31 : DOCRISP Operations Semaphore 0 0 1 1 CCR D2: DO Control Register 0 1 0 0 IRQEN D7 : DO Interrupt Enables Register 0 1 0 1 IRQSC D7 : D0 Interrupt Status (read), Clear (Write Is) 0 1 1 0 TMR D15 : DO WatchDog/Timer Clock Tick 0 1 1 1 SPC D23: D2 CRISP Program Counter Start Address 1 X X X <Spare>--Spare register space reserved for future expansion.

DPTR and IDAT The data pointer (DPTR) provides a R/W address for transfers to and from SDRAM via the IDAT register. The DPTR is automatically incremented by 4 bytes subsequent to each transfer.

CSEM General purpose semaphore register set/cleared by both CRISP and system processor CCR The Control register controls the activity state of the CRISP processor.

Table 102

Bit Name Function 2 STEP Write of 1 executes a single Opcode, self clearing upon completion. (HOLD is set = 1) 1 HOLD 0=RUN (continue), 1=HOLD (halt) CRISP process 0 START Start CRISP at new address (held in CPC register) IRQEN Enables for interrupt sources IRQSC Status of IRQs, a write of 1 resets/clears interrupt source TMR Clock scaler for CRISP watchdog/timer. This value is transferred to an internal Counter. Each time the counter "rolls over", one"tick"is applied to CRISPs watchdog/timer register.

SPC This register provides the start address for CRISP programs when enabled by the CCR's START bit (see CCR, above).

All CRISP instructions are 32 bits, so this start address will always fall on a 4-byte boundary (e. g. Al and AO will always = 0).

CRISP Control & Status (8-bit registers) Please note: the following are addressable with LDC, SET, CLR, or TST instructions. A10 : A5 may be used for future expansion addressing, and should generally be programmed as zeros; the hardware may or may not decode these address bits in the various implementations.

Table 103

All : A10 : A4 A3 A2 Register Data Bits Remarks A5 0X000DEVCOND7 : DO Device Control 0 X 0 0 1 AUXCON D6 : DO Auxiliary Control 0 X 0 1 0 LSTAT DS : DO Line Statuses (read- only) 0 X 0 1 1 WDCNT D7 : DO WatchDog/Timer Count 0 X 1 0 0 SEM0 D7:D0 Semaphore Register 0 0"X101SEMID7 : DOSemaphore Register 1 0 X 1 1 0 SEM2 D7:D0 Semaphore Register 2 0 X 1 1 1 SEM3 D7 : DO Semaphore Register 4 DEVCON The Device Control register controls power, reset and clocks for the Display (s). All 8 bits are also mirrored in a CRASIC accessible (see CRASIC External Registers, CCR).

Table 104 Bit Name Function 7 DBT 2 Clock Divide By 2 for Microdisplay [Left] 6 CLK 2 Clock Enable for Microdisplay [Left] 5 RST2 Reset Control for Microdisplay [Left] 4 PWR2Power Enable for Microdisplay [Left] 3 DBT_1 Clock Divide By 2 for Microdisplay [Right] 2CLK1Clock Enable for Microdisplay [Right] 1 RST-1 Reset Control for Microdisplay [Right 0 PWR1Power Enable for Microdisplay [Right

AUXCON The Auxiliary Control register controls power and reset for the AIC, 2 GPOs (for an additional device), and behavior of the Watch-Dog/Timer function.

Table 105 Bit Name Function 5 WDRUN Watch-Dog/Timer Run 4 WDENBL Watch-Dog Abort Enable Abort-time out causes abort and sets WD IRQ to host, if enabled. When this bit is false the Watch-Dog Status is mapped to AUX cc, allowing WD as a general purpose program timer 3 GPO 2 General Purpose Output 2 (could control reset for additional device) 2 GPO_1 General Purpose Output 1 (could control power for additional device) 1 AICRST Reset Control for AIC-if dual Displays are used, this signal may be used for both AICs 0 AICPWR Power Enable for AIC-if dual Displays are used, this signal may be used for both AICs.

LSTAT The Line Status register is read only states of various external signals from the Display (s) and the CRASIC itself.

Table 106 Bit Name Function 7 GPI_2 General Purpose Input 2 6 GPI 1 General Purpose Input 1 5 WDTO Watch-Dog Timer Timed Out (This signal may or may not be assigned an external CRASIC pin. It may be cleared by writing to the WDCNT register.) 4 CIRQ CRASIC Interrupt line level 3 OITO 2 ITO line level [Left] 2 OIRQ2 Interrupt line level [Left] 1 OITO_1 ITO line level [Right] 0 OIRQ1 Interrupt line level [Right]

WDCNT The Watch-Dog/Timer count is an 8 bit register whose value is transferred to a count-down counter each time a CRISP instruction with WDR=True is executed, or when this register is written.

SEMx Semaphore registers. SEMO corresponds to D7: DO ; SEMI to D15 : D8; SEM2 to D23: D16 ; SEM3 to D31: D24 of the CSEM register (see CRASIC External Registers. CSEM). Each bit may be read or written by either the Host Processor through CSEM, or CRISP through SEMx registers.

CRISP Process Registers (32-bit registers) Please note: the following are NOT addressable with LDC, SET, CLR, or TST instructions. A10 : A9 may be used for future expansion addressing, and should generally be programmed as zeros; the hardware may or may not decode these address bits in an implementation. The first 8 locations are addressable by the LDR and STR instructions.

Table 107 All A10: A8 A7 A6 : A2 Register Data Remarks A9 Bits IX00XOOOO TSAD23 : D2 Transfer Source Address 1 X 0 0 X0001 TDA D23 : D2 Transfer Destination Address 1 X 0 0 X0010 TCNT D21 : D2 Transfer Count 1 X 0 0 X0011 CPC D23 : D2 Current CRISP Program Counter 1 X 0 0 XO1XX <spare>---Spare register locations that may be loaded by the LDR instruction. 1 X 0 0 X1000 OPR D23 : D2 Opcode-NOTE: not loadable using LDR instruction. 1 X 0 1 Reg BIAS [0: 3] D31 : DO Dither Grid Bias Registers Addr 1: 0 (Actual organization to be determined) 1 X 1 Reg Reg CLUT [0: 6 D31: DO 256 byte (64x4 bytes) Color Addr Addr 3] Look-Up Table. Data is 5 4: 0 arranged"Little Endian", e. g. D7: DO comprise the first byte; D31 : D24 is the third byte.

TSA Transfer Source Address. Pointer to source data to be used during execution of Data Transfer Opcodes. This may register may be loaded using the LDR instruction or externally addressed when CRISP is in HOLD mode.

TDA Transfer Destination Address. This may register may be loaded using the LDR instruction or externally addressed when CRISP is in HOLD mode.

TCNT Transfer Count. This value is the number of 32-bit words to be processed during execution of Data Transfer Opcodes. This may register may be loaded using the LDR instruction or externally addressed when CRISP is in HOLD mode.

CPC This register holds the current program counter of a CRISP program. This may register may be loaded using the LDR instruction or externally addressed when CRISP is in HOLD mode.

OPR While CRISP is running, this register is automatically loaded from the address pointed to by the CPC. When CRISP is in the HOLD state, this register may be written externally by the Host CPU with a CRISP Opcode and executed via the STEP function of the CCR (see CRASIC External Registers, CCR).

BIAS See next section: CRISP Palettes and Dithering CLUT See next section: CRISP Palettes and Dithering CRISP Palettes and Dithering When processing 256"mapped"colors the same source data is used for each color plain, but the values in the palette are reloaded for the color being processed.

Changing the RAM-based copies of the palette (one each for Red, Green, and Blue) will result in color changes in the image, even though the source image data is unchanged.

When processing 24-bit color the same palette may be used, but the source data comes from separate (planerized) buffers for Red, Green, and Blue. Separate color palettes may be used to correct gamma for individual colors.

Figure 17 shows the process 1700 to convert 8-bit pixel data into pixels in a format amenable to the Display System. In operation 1702, the original 4 pixels (8 bits each) are identified. A palette lookup (256 entries x 7 bits) is performed in operation 1704. Each pixel is now 3 bits pixel data + 4 bits remainder. In operation 1706, grid- bias (4 bits) is added to each pixel according to its row-column. Remainder bits are removed in operation 1708, leaving only 3 bits per pixel data. In operation 1710, the pixels are then packed into the format of the Display System. When the second group of 4 pixels have been processed, they are packed into the other half of the 32- bit word, and all 8 pixels are then written to the Display System or memory.

Analog Controller Overview Description The Analog Controller (OAC) implements all power management functions. This includes power efficient DC to DC conversions needed for driving the Liquid Crystal and LED for the Display System, and the programmability of electrical parameters. This ensures the most optimal settings regardless of the operating temperature and unit variation. The temperature sensor is on-chip and the compensation is done automatically by the internal state machine. An internal booster converter generates the voltages and currents necessary to bias three separate LEDs. The charge pump and the voltage regulator generate and regulate the Liquid Crystal ITO voltages. The voltage and current for each color are controlled individually for optimal performance and power savings. The chip also monitors the temperature and, as the temperature reading changes, reads the corresponding table values from the separate EEPROM. It also recalculates and programs the LED currents and LC ITO voltages through the parallel, or I2C, interface.

Features Requires only single 3.3 V supply Fully compatible to with Backplane ICs Can drive both Liquid Crystal and three color LEDs with a few external passive components Highly power efficient operation Supports three color LED field sequential illumination Supports high refresh rate (more than 100 Hz) as well as low refresh rate g Supports gray scale mode Supports both binary and RMS mode operations to drive Liquid Crystal On-chip voltage boosters for Liquid Crystal and LED drivers Programs precision ITO voltages Matching ITO inversion voltage automatically generated Can program Liquid Crystal DC offset voltage Single common LED anode pin supplies all three LEDs 'Separate common anode voltages for individual LED can be set sequentially 'Each LED intensity is controlled by sinking different currents at the split LED cathodes Monitors the chip junction temperature with offset trim capability 'Supports external thermistor option Can automatically correct the temperature dependencies of Liquid Crystal and LED related parameters.

Parallel interface-compatible to Backplane parallel interface System Interface and Timing Pin Assignment Figure 18 is an illustration of an Analog Controller Chip 1800. The pinout for the chip is set forth in the table below.

Table 108: Pin Out by Pin Number PAD PAD PAD PAD # NAME DESCRIPTION TYPE 1 csN Chip select. An external host CPU, Digital, In Controller or Color Rich ASIC generates this signal. 2 a [5] OAC Register Address bit 5. Connect Digital, In together with a [9] of Display chip. 3 a [4] OAC Register Address bit 4. Connect Digital, In together with a [8] of Display chip. 4 a [3] OAC Register Address bit 3. Connect Digital, In together with a [7] of Display chip. 5 a [2] OAC Register Address bit 2. Connect Digital, In together with a [6] of Display chip. 6 a [l] OAC Register Address bit 1. Connect Digital, In together with a [5] of Display chip. 7 a [O] OAC Register Address bit 0. Connect Digital, In together with a [4] of Display chip. 8 Vddd Positive digital supply voltage. A regulated Power Supply, 3.3V system power is supplied from the In host system. 3.3V is provided through a supply pin of 80-pin connector. LED and Backplane chip also powered from the same source. 9 vssd Digital ground. All Logic circuits and Power Supply, digital pads are connected to this ground. In 10 d [7] MSB of data byte. These 8 bits of data Digital, Bidir lines are connected to the least significant 8 bits of data bus. 11 d [6] Bit 6 of data byte Digital, Bidir 12 d [5] Bit 5 of data byte Digital, Bidir 13 d [4] Bit 4 of data byte Digital, Bidir 14 d [3] Bit 3 of data byte Digital, Bidir 15 d [2] Bit 2 of data byte Digital, Bidir 16 d [1] Bit 1 of data byte Digital, Bidir 17 d [O] LSB of data byte Digital, Bidir 18 Reserved for the future use 19 Reserved for the future use 20 Reserved for the future use 21 Reserved for the future use 22 Reserved for the future use 23 blue Input signal from Display chip. This signal Digital, In will control on/off timing of Blue LED. 24 red Input signal from Display chip. This signal Digital, In will control on/off timing of Red LED. 25 green Input signal from Display chip. This signal Digital, In will control on/off timing of Green LED. 26 led LED'on'indicator from Display chip. Digital, Bidir During test mode this pin outputs clkito 27 Test pin : vdacled Analog, OUT Test Pin 28 comled Voltage booster2 output. A low pass filter Analog, Out capacitor for the LED supply voltage booster is connected. Common anode of LED is connected to this pin. 29 vsse Ground pad for the LED driver circuitry. Power Supply, In 30 lled Connected to one end of inductor to the Analog, Out pch rectifier and nch switch. 31 vdde Positive supply voltage for the LED driver Power Supply, circuitry. A regulated 3.3V system power In is supplied from the host system. 3.3V is provided through the supply pin of 80-pin connector. 32 rled Red LED On switch. This pad is connected Analog, In to the cathode of terminal red LED. 33 gled Green LED On switch. This pad is Analog, In connected to the cathode of terminal green LED. 34 bled Blue LED On switch. This pad is Analog, In connected to the cathode of terminal blue LED. 35 Test pin : idacled. Reflecting the voltage Analog, OUT across the internal current monitoring Test Pin resistor. 36 vssa Analog ground. This ground is connected Power Supply, only to the charge pump, A/D & D/A, In operational amplifiers and filters, POR, voltage reference and regulators. 37 vdda Power supply pin to the analog circuitry. Power Supply, Analogous to vssa. In 38 resetN Reset input pin. This signal is'or'ed with Digital, In internal POR signal. Once reset, the chip Active Low will go through the boot-up procedure including down-loading of default parameters from EEPROM 39 porstN Power On Reset. This signal goes to the Digital, Out, Microdisplay (or Display chip) and other Active Low chips in the system, and resets all internal circuitry. 40 rref Precision reference current generator Analog, In resistor. A resistor with 1% tolerance and low temperature coefficient must be used. 41 Testpin : vbg. Bandgap voltage. Analog, OUT TestPin 42 Reserved for future use 43 vito ITO voltage output pin. Liquid Crystal ITO Analog, Out electrode is connected to this pin. This voltage changes from Vtrip to Vtrin as the digital input the'ito'pin switches from'1' to 0'or vice versa. The voltage levels are programmed in the register. 44 Test pin : dacos Analog, OUT Test Pin 45 ito Digital input signal from Display chip. Digital, In Synchronized with the pixel voltage polarity and indicating the polarity of'vito' pin. 46 Test pin : dacres Analog, OUT Test Pin 47 Testpin : dacvito Analog, OUT Test Pin 48 tsense External temperature sensor input. Test pin Analog, In used as an internal sensor monitoring. 49 Reserved for future use 50 vdblp Double booster output. A holding capacitor Analog, Out is connected (plus voltage). 51 cppl Voltage double booster capacitor terminal Analog, Out (positive) for plus Vito generation. 52 cpnl Voltage double booster capacitor terminal Analog, Out (negative) for plus Vito generation. 53 cmpl Voltage double booster capacitor terminal Analog, Out (negative) for minus Vito generation. 54 vdbln Negative double booster output. A holding Analog, Out capacitor is connected (minus voltage). This is generated from Vdblp. 55 cmnl Voltage double booster capacitor terminal Analog, Out (positive) for minus Vito generation. This voltage is generated from the positive boosted voltage. 56 cpn2 Voltage triple booster capacitor terminal Analog, Out (positive) for plus Vito generation. 57 cpp2 Voltage triple booster capacitor terminal Analog, Out (negative) for plus Vito generation. 58 vtrip Triple booster output. A holding capacitor Analog, Out is connected (plus voltage). 59 mclk Master clock from the host CPU or Digital, In controller. This clock will be divided into a slower clock by 2"where n is in the Input Clock Divider register. 60 1 cmst Serial communication master mode. Digital, In Setting this to'1'will put OAC into a master mode for the EEPROM serial communication. Otherwise, the system CPU will be the master. 61 Reserved for future use 62 scl I'C interface serial interface clock. Digital, Bidir Frequency is up to 100K. 63 sda I C interface serial interface data. DIgital, Bidir Frequency is up to 100K. 64 wrN Register write Digital, In Active Low Pin Description Table 109: Pin Description Name Direction Type Description A [5: 0] Input Digital System address. csN Input Digital System chip select. Active low. D [7: 0] Bi-Digital System data. directional mclk Input Digital System clock (master clock). rstN Input Digital System reset. Active low. wrN Input Digital System write. Active low. Drive high to indicate a read. bled Output Analog Red LED control. Blue timing indicator from blue Input Digital Backplane IC. comled Output Analog Common LED control. gled Output Analog Green LED control. Green timing indicator from green Input Digital Backplane IC. ITO Timing indicator from ito Input Digital Backplane IC. Normally, LED timing indicator from Backplane IC. Becomes the Bi- led Digital charge pump clock diagnostic directional output when the cp clk test bit (Ox7: 0) = 1. lled Output Analog LED Inductor control. mled Input Analog LED Monitor input. Red timing indicator from red Input Digital Backplane IC. rled Output Analog Red LED control. vito Output Analog ITO Voltage output.

System Bus The Bus has a parallel address, data, and control signal organization. The OAC and the Backplane IC receive separate chip selects. While several other signals are shared, the interface to the Backplane IC is intended to operate at much higher data rates. The Backplane IC also has a more complicated protocol.

The table above includes only those pins of the OAC belonging to the parallel interface, along with the timing signals from the Backplane IC, the ITO voltage, and LED current pins. Refer to the System Interface and Timing section for a complete listing of the OAC pins.

A [6: 0] is used for register addresses when A [7] is high and ignored otherwise.

Bus Protocol Bus Handshake Figure 19 illustrates a transaction waveform 1900 during parallel write timing.

Figure 20 illustrates a transaction waveform 2000 during parallel read timing. Read and write accesses from the host system to the OAC consist of driving the chip select csN low, setting the address bus A [S : Oy for the duration of the access, and driving or floating the data and write signals appropriately. The chip select signal must be pulled high a minimum of 5 cycles between accesses.

Write accesses must last for a minimum of 9 clock cycles. Any access of shorter duration may lead to an indefinite setting of the configuration registers. There is no maximum access duration.

Read accesses must last for a minimum of 9 clock cycles before valid data is returned. The read data will remain valid until the chip select is deactivated.

ITO Voltage Figure 21 depicts an ITO voltage generation waveform 2100. The voltage output to the common counter electrode of the LCM is a function of the polarity of the ito signal from the Backplane IC, and the color of the current field. This is determined by the red, green, and blue signals from the Backplane IC. The magnitudes of the ITO voltage relative to the power rails are the programmable parameters red ito, green-ito, and blue ito.

LED Currents Figure 22 depicts an LED current generation waveform 2200. The currents driven to the individual LED's are, in general, a function of the polarity of the led signal from the Backplane IC, and the color of the current field. The magnitudes of the currents through, and voltages across, the red, green, and blue LED's are the programmable parameters ired, igreen, iblue, vred, vgreen, and vblue, respectively.

ITO and LED Timing Figure 23 illustrates an LED timing waveform 2300 where itorst = 0. The timing for the generation of the ITO voltage and the LED currents is determined by the ito, led, red, green, and blue signals from the Backplane IC. Led always falls coincident with the fall of red, green, or blue. A special case arises when led falls later, which is possible through the programming of the flash delay registers in the Backplane IC.

The ITO reset pulse enable bit ito rst, determines how the overlap of the led signal onto the next color field is interpreted. When dito-rut is off the driving of the LED corresponding to the original color field, it is prolonged until the overlap ends. The green LED is driven well into the blue field and the blue LED into the red field.

Figure 24 depicts a waveform 2400 for ITO and LED Timing with ito-rst = 1. When the ito rst bit is on, the driving of the current LED is not prolonged, but instead a

special ITO voltage is generated. This voltage is called a zap, or reset voltage, and has a magnitude given by the programmable parameter reset-ito.

Analog Controller Configuration Figure 25 is a block diagram of an analog controller 2500 according to an embodiment of the present invention.

Configuration Registers On the parallel bus the OAC configuration registers are accessed according to the value of A [5: 0]. For address compatibility with the mapping for the Backplane Integrated Circuit (Backplane IC) configuration registers, it is assumed that A [5: 0] of the OAC is connected to A [9: 4] of the Backplane IC.

On the I2C bus, the OAC configuration registers are accessed according to the value of the word address.

All 8 bits are not always defined for a particular configuration register. Where a bit is undefined, it should be considered reserved, and a 0'should be written to it for compatibility with future versions of the chips. Reading an undefined bit always returns a'O'.

Not all addresses within a configuration register space map into a defined configuration register. Such addresses should be considered reserved. Only values of '00'should be written to reserved register addresses, to maintain compatibility with future versions of the chips. Reading a reserved register always returns'00'.

All registers below are read/write accessible except where noted. All address values are given in hexadecimal. The Init column indicates the values of the respective variables at reset.

Configuration Register Address Map Table 110: Configuration Register Summary Address Range Register Group Description Chip Revision, EEProm, POR status and power down, Bandgap voltage OxOO-Ox07 System Interface trim, clock divider for charge pump and led booster clocks, Backplane chip status Temperature sensor control, OxOC-OxOF Temperature Control temperature sensor offset adjust, temperature monitor. ITO voltage offset, red OxlO-Oxl7 ITO Control green blue ito voltages, reset voltage. LED current set, led anode Ox 18-Ox 1 E LED Control voltage set, and led power control.

Chip Revision Register (OxO) To determine feature set and compatibility, the host CPU reads the Chip Revision register. This register can be overwritten internally at power-up when the OAC reads the I2C EEPROM. This register will change to reflect the first byte of read from the EEPROM.

System Registers (Ox01-Ox07) EEPROM ID (Ox01)

This byte, the EEPROM ID register, can also indicate the manufacturer and the size of EEPROM used.

Chip Power Control (Ox02) Table 111 Bit Reference Description Software power-on reset for the Backplane IC. The porstN output 2 sporstN pin is pulled low whenever this bit is low, or the OAC detects a powering up condition. Power down of the analog circuitry that generates the LED currents 1 led_pdwnN (rled, gled, and bled). Does not affect the generation of power-on reset (porstN). Active low. Power down of the analog circuitry that generates the ITO voltage (vito) and the LED currents (rled, 0 pdwnN gled, and bled). Does not affect the generation of power-on reset (porstN). Active low.

Reference Voltage Trim (Ox03) Bits 5-0 of the Vbg register are a value to trim the output voltage of internal bandgap circuitry.

Charge Pump Clock Divider (Ox04 & Ox05 [3: 0])

Bits 7-0 of the Charge Pump Clock Divider register are a divisor value to scale the input clock to generate OAC internal clock references for the charge pump.

Assuming OAC runs from the same 66Mhz clock as the Backplane chip, for example, a value of 3fh would provide a 1.476Mhz internal clock. Nominally the divided value should be as close as possible to 25KHz.

LED Booster Clock Divider (Ox05 [7: 4] & Ox06) Bits 7-0 of the LED Booster Clock Divider register are a divisor value to scale the input clock to generate OAC internal clock references for the LED voltage booster.

This clock is used for the internal PWM circuitry. Assuming OAC runs from the same 66Mhz clock as an Backplane chip, for example, and a value of 3fh would provide a 1.476Mhz internal clock.

Backplane Chip Status (Ox08) The Display System reports the state of all inputs from the Backplane chip. The purpose of this register is primarily for diagnostic, calibration, and production QA testing of assembled modules.

Table 112 Bit Reference Description 4 itop ITO input pin state. 3 led LED input pin State. 2 red Red LED input pin state. 1 green Green LED input pin state. 0 blue Blue LED input pin state.

Temperature Sensor Configuration (OxOC) Table 113 Bit Reference Description 7 temp test Temperature test enable via pin tsense. For diagnostics only: temp-test should normally be left low. External temperature sensor enable 6 ets via pin tsense. Enables use of an external look-up 5 lut-enable table based on the sampled temperature. Enables temperature sampling. Set to 0'whenever temp interval is to 4 temp enable be changed and then reset to'1' afterwards. Temperature sampling interval. Enter 0'for every 32 cycles,'1' 3: 0 temp interval for every (222 + 32) cycles,'2'for every (2 * 222 + 32 cycles), etc.

Temperature Sensor Offset (OxOD) Bits 7-0 of the Temperature Sensor Offset register sets and adjusts the base of the temperature readings. Ideally, the lowest sensed temperature will read as FFh, and the highest as OOh.

Temperature Value (OxOE) & Sample (OxOF) The Temperature Sample register reports the temperature at the OAC chip, or external sensor, depending on the System Configuration register setting. The value can be used to fine tune ITO voltages and LED currents going to the LCM.

Temperature Sample reports the average of eight most samples.

ITO Mode Control (Ox 10) Table 114 Bit Reference Description 1 itorst ITO reset pulse enable. Setting this bit will enable the generation of ITO reset (zap) voltages. The reset pulse will position the LC into a fully saturated state before settling into a certain RMS state.

ITO Reset (Zap) Voltage (Ox12) Bits 7-0 of the ITO Reset (Zap) Voltage register sets the level for ITO between fields when"Reset Mode"is true (see ITO Mode Register). The range for ITO Zap Voltage is 0 through 5.7 volts above VDD and below GND ITO Offset Voltage 1-Viton trim (gain_balance) (Ox13) Bits 7-0 of the ITO Baseline Voltage register sets the offset voltage for red, green, and blue ITO voltages. This register trims the mismatch between the resistors used to generate Viton. The trim range for Viton is +/-2% of Viton below GND. OOh will produce the highest offset voltage, and FFh will be the lowest offset voltage.

ITO Offset Voltage 2-Vcenter trim (dc_bias) (Ox14) Bits 7-0 of the ITO Baseline Voltage register sets the offset voltage for red, green, and blue ITO voltages. This register trims the mismatch between the resistors used to generate Vcenter, that is the middle point between Vdd and GND. The trim range for Viton is +/-22% of Viton below GND. This register, in conjunction with Register 06, can generate an arbitrary offset voltage on Viton. OOh will produce the highest offset voltage, and FFh will be the lowest offset voltage.

ITO Voltage-Red (Oxl 5) Bits 7-0 of the ITO Red Set register sets the ITO voltage during a red field. The range for ITO Red set is from 3.3V to Vito max. While ITO is positive, Vito is switched to the positive ITO voltage. While ITO is negative, Vito is switched to the

Vito negative. Vito negative is also generated by the value set by this register. The range for the Vito negative is from GND to- (Vito-3. 3V).

ITO Voltage-Green (Ox16) Bits 7-0 of the ITO Green Set register sets the ITO voltage during a Green field. The range for ITO Red set is from 3.3V to Vito max. While ITO is positive, Vito is switched to the positive ITO voltage. While ITO is negative, Vito is switched to the Vito negative. Vito negative is also generated by the value set by this register. The range for the Vito negative is from GND to- (Vito-3. 3V).

ITO Voltage-Blue (Ox 17) Bits 7-0 of the ITO Blue Set register sets the ITO voltage during a Blue field. The range for ITO Red set is from 3.3V to Vito max. While ITO is positive, Vito is switched to the positive ITO voltage. While ITO is negative, Vito is switched to the Vito negative. Vito negative is also generated by the value set by this register. The range for the Vito negative is from GND to- (Vito-3. 3V).

Red LED Current (Ox18 [5: 0]) Bits 5-0 of the Red LED Current register sets the amount of current drawn through the red LED while"flashing"a red field. The range for Red LED Current is 0 through 120mA, each increment representing approximately 2mA.

Green LED Current (Oxl9 [3: 0] & Oxo8 [7: 6]) The 6 bit Green LED Current register sets the amount of current drawn through the green LED while"flashing"a green field. The range for Green LED Current is 0 through 120mA, each increment representing approximately 2mA Blue LED Current (oxo9 [3: 0] & Ox I A [1: 0]) The 6 bit Blue LED Current register sets the amount of current drawn through the blue LED while"flashing"a blue field. The range for Blue LED Current is 0 through 120mA, each increment representing approximately 2mA.

Red LED Common Anode Voltage (OxlA [6: 2]) The 5 bit Red LED Voltage register sets the common anode voltage while color inputs from the Backplane chip activates from one color to the next. Only one color is enabled with field sequential display. At this time the LED signal from the Backplane chip is disabled. The purpose of this is to settle the common anode voltage before the LEDs are activated and start drawing current.

Green LED Common Anode Voltage (OxlB [3: 0] & OxlA [7]) The 5 bit Green LED Voltage register sets the common anode voltage while color inputs from the Backplane chip activates from one color to the next. Only one color is enabled with field sequential display. At this time the LED signal from the Backplane chip is disabled. The purpose of this is to settle the common anode voltage before the LEDs are activated and start drawing current.

Blue LED Common Anode Voltage (OxlB [7: 4] & Ox 1C [0]) The 5 bit Blue LED Voltage register sets the common anode voltage while color inputs from the Backplane chip activates from one color to the next. Only one color is enabled with field sequential display. At this time the LED signal from the Backplane chip is disabled. The purpose of this is to settle the common anode voltage before the LEDs are activated and start drawing current.

Summary Table 115: Configuration Register Summary Address Range Register Group Description Chip Revision, EEProm, POR status and power OxOO-Ox07 System Interface down, Bandgap voltage trim, clock divider for charge pump and led booster clocks, Backplane chip status Temperature sensor control, OxOC-OxOF Temperature Control temperature sensor offset adjust, temperature monitor. ITO voltage offset, red, OxlO-Oxl7 ITO Control green, blue ito voltages, reset voltage. LED current set, led anode Ox 18-Ox 1 C LED Control voltage set, and led power control.

Detailed Description Table 116 Addr Bits Init Name Description Ou00 Chip identification (read only). Returns OxOI, 0x02, etc., to 7: 0 2 Chip_ID indicate the version number of the chip. 0x01 Serial EEPROM identification. Following a configuration load 7: 0 0 ROM m from serial EEPROM, returns OxOI, 0x02 etc., to indicate the version number of the ROM. Ox02. Software power-on reset for the Backplane IC. The porstN output 2 1 sporstN pin is pulled low whenever this bit is low, or the OAC detects a powering up condition. Power down of the analog circuitry that generates the LED currents (rled, gled, and bled). 1 1 led_pdwnN Does not affect the generation of power-on reset (porstN). Active low. Power down of the analog circuitry that generates the ITO voltage (vito) and the LED 0 1 pdwnN currents (rled, gled, and bled). Does not affect the generation of power-on reset (porstN). Active low. Ox03 5: 0 0 band gap Band gap trimming voltage. OxO4 Lower eight bits of divider used to generate the charge pump clock from the system clock. Least significant bit, div cp [0], is read 7: 0 ff dix cep [7: 0] only and always returns'1'. Default value produces a # 2048 clock. Charge pump clock is idled whenever pdwnN is low. OxO5 Lower four bits of divider used to generate the LED clock from the system clock. Least significant bit, div led [0], is read only and 7: 4 f Dialed [3: 0] _ always returns'1'. Default value produces a 0 64 clock. LED Clock is idled whenever led_pdwnN or pdwnN is low. Upper four bits of charge pump 3: 0 7 div_cp[11 : 8] clock divider. 0x06 Upper eight bits of LED clock 7: 0 03 div_led[11:4] divider. OxO7 Charge pump clock test enable. When '1', routes internal charge pump clock out through the led pin. For diagnostics only: 0 0 cp_clk_test cpclktest should normally be left low, in which case led performs as an input from the Backplane IC. Backplane IC status register. Status of incoming control bits Oxo8 from the Backplane IC. Read only. 4-ito Input ito pin value. 3-led Input led pin value. 2-red Input red pin value. 1-green Input green pin value. 0-blue Input blue pin value. DAC Test register. For oxo9 diagnostics only. LED Voltage. When enabled, 4 0 vled test output at pin led test. LED Current. When enabled, 3 0 iled test output at pin led test ITO Voltage. When enabled, 2 0 vito test output at pin led test. Internal resistor balance 1 0 gain-test adjustment for vito. When enabled, output at pin ito_test. DC Bias voltage adjustment for 0 1 bias-test vito. When enabled, output at pin ito test. OxOc Temperature test enable via pin tsense. For diagnostics only: 7 0 temp test - temp_test should normally be left low. External temperature sensor 6 0 ets enable via pin tsense. Enables use of an external look-up 5 1 lut-enable table based on the sampled temperature. Enables temperature sampling. Set to 0'whenever temp interval 4 1 temp enable is to be changed and then reset to '1'afterwards. Temperature sampling interval. Enter 0'for every 32 cycles,'1' 3: 0 8 temp interval for every (222 + 32) cycles,'2'for every (2 * 222 + 32 cycles), etc. OxOd 7: 0 80 temp offset Temperature offset. OxOe Temperature value. Average of 5: 0 - temp_val eight most recent samples. Read only. OxOf Temperature sample. Most recent 5:0 temp samp temperature reading. Read only. Ox01 0 ITO Reset pulse enable. A'1'on this bit enables the generation of an ITO reset voltage on the continuance of the led signal into a new field. A'0'disables the 1 0 ito rst generation of the reset voltage, and allows the continuance of the led signal into a new field to be interpreted as keeping the previous field's LED on. Ox12 ITO Voltage during a reset 7: 0 00 reset ito interval. Analogous to red-ito. Oxl3 vito_gain balance Internal resistor balance 7: 0 80 (riton) adjustment. Oxl4 vito dc bias 7: 0 80 DC Bias voltage adjustment. (rvcenter) OxlS ITO Voltage during a red field. Value during a positive ITO field 7: 0 00 Red ito is relative to the power supply VDD. Value during a negative ITO field is relative to ground. Oxl6 ITO Voltage during a green field. 7: 0 00 green ito Analogous to red-ito. Oxl7 ITO Voltage during a blue field. 7: 0 00 blue ito Analogous to red-ito. Ox18 Lower two bits of current drawn 7: 6 0 igreen [1 : 0] by the green LED during the flash region of the green field. Current drawn by the red LED 5 : 0 00 Ired [5 : 0] during the flash region of the red field. Oxl9 Lower four bits of current drawn 7: 4 0 iblue [3: 0] by the blue LED during the flash region of the blue field. Upper four bits of current drawn 3: 0 0 igreen [5: 2] by the green LED during the flash region of the green field. OxlA Lower bit of voltage applied at the 7 0 vgreen [O] green LED during the flash region of the green field. Voltage applied at the red LED 6: 2 00 vred during the flash region of the red field. Upper two bits of current drawn 1: 0 0 iblue [5: 4] by the blue LED during the flash region of the blue field. OxlB Lower four bits of voltage applied 7: 4 0 vblue [3: 0] at the blue LED during the flash region of the green field. Upper four bits of voltage applied 3: 0 0 vgreen [4: 1] at the green LED during the flash region of the green field. OxlC Upper bit of voltage applied at the 0 1 vblue [4] blue LED during the flash region of the green field. I C Slave device 0 configuration ou20 register 1. Slave address for device 0. Reset 7: 4 A devaddr [0] value is correct for serial EEPROM. Upper 3 bits of word address of the next byte access for device 0. word addr [O] [10: 8 3: 1 0 In conjunction with the lower 8 bits (Ox21 : 7-0), automatically increments after each access. 12 C r/wN Bit for device 0: determined by the operation O O rwN [0] (0x30 : 1-0). Always returns 0'on a read. Ox21 I2C Slave device 0 configuration 0x21 register 2. Lower 8 bits of word address of the next byte access for device 0. 7: 0 0 word_addr[0] [7: 0] In conjunction with the upper 3 bits (Ox20 : 3-1), automatically increments after each access. I2C Salve device 0 configuration ou22 register 3. Device 0 error. A'1'indicates the 7 0 err [O] occurrence of an error during a write or read access. Device 0 present. Automatically set or reset by the I2C bus 6 0 present [0] initialization operation. A'1' indicates the device was detected. Maximum number of tries before timing out on a write to device 0. 5 : 0 0 time_out[0] A try fails on a negative acknowledge to the transmission of the slave address. Ox24 I2C Slave device 1 configuration 0x24 register 1. Slave address for device 1. Reset 7: 4 8 devaddr [l] value is correct for the Backplane IC. word_addr[1][10: 8 Upper 3 bits of word address of 3: 1 0 ] the next byte access for device 1. 0 0 rwN[1] I2C r/wN Bit for device 1. I C Slave device 1 configuration Ox25 registers 2-3. Analogous to - 26 device 0 configuration registers 2 - 3 (0x21 - 22). Oxo Number of the device to undergo 4 0 op_ndev the I C operation. I C operation err. A'1'on this bit 3 0 Operr indicates an error was detected. I2C operation request. A'1'on this bit requests an I2C operation 2 0 Opreq to start. The bit is automatically cleared at the conclusion of the operation. 1: 0 0 op_mode Mode of operation. Ox31 Data. The contents are written to the I2C bus on a write operation. 7: 0 0 data The contents are updated with read data on the conclusion of a successful I2C read operation. 0x32 I C error during DMA access. A '1'on this bit indicates an error 6 0 dma err was detected during an I2C boot or LUT access. C DMA access request. A'1'on this bit requests a DMA access to 5 0 dmareq start. The bit is automatically cleared at the conclusion of the access. 12 C DMA access size. Enter'0' 4: 0 1B dma siz - for 1 byte, '1' for 2 bytes, etc. Ox33 4 : 0 0 dmaaddr [4: 0] I2C DMA access starting address. Ox34 Lower 3 bits of clock low time. Enter 0'for 1 cycle, 1'for 2 7: 5 7 tlow [2: 0] cycles, etc. Reset value is correct for 3.3V, 66 kHz I 2C operation. Data set up time. Enter 0'for 1 cycle,'1'for 2 cycles, etc. Reset 4: 0 10 tset value is correct for 3. 3V, 66 kHz I2C operation. Ox35 Lower 2 bits of clock high time. Enter '0' for 1 cycle, '1' for 2 7: 6 3 thigh [l : 0] cycles, etc. Reset value is correct for 3.3V, 66 kHz I2C operation. Upper 6 bits of clock low time for 5: 0 27 tlow [8: 3] device 0. Ox36 Lower bit of read time. Enter'0' for 1 cycle, 1'for 2 cycles, etc. 7 1 tread [0] Reset value is correct for 3.3V, 66 kHz I2C operation Upper 7 bits of clock high time 6: 0 43 thigh [8: 2] for device 0. for device 0. 0x37 Upper 8 bits of read time for 0 9F tread [8 : 1] device 0. Ox38 Enables one-clock buffering of 0 1 filter incoming I2C signals to filter out noise. Ox3C 7 : 4 7 device address Device address for the OAC. Upper 3 bits of word address of the next byte access. In conjunction with the lower 8 bits 3: 1 0 word addr [10: 8] (Ox3D : 7-0), automatically increments after each access. Read only. I2C r/wN Bit. Writing to it has no 0 0 rwN effect. Always returns'1'on a read. Ox3D Lower 8 bits of word address of the next byte access. In conjunction with the upper 3 bits 7: 0 0 word addr [7: 0] (Ox3C : 3-1), automatically increments after each access. Read only. Ox3E Lower 8 bits of bus turn around time. Enter'O'for 1 cycle,'1'for 7: 0 E9 tbta [7: 0] 2 cycles, etc. Reset value is correct for 3.3V, 66 kHz IZC operation. Ox3F 0 0 tbta [8] Upper bit of bus turn around time.

Additional Functions Pin Out The following table lists the pin out for the OAC pins used for additional functions.

There are a total of 6 pins. The led pin is also listed in this table.

Table 117 : Additional Function Pins Name Direction Type Description Diagnostic output for ITO-related ito test Output Analog DAC's. Normally, LED timing indicator Bi-from Backplane IC. Becomes the led Digital directional charge pump clock diagnostic output when the cp clk test bit (Ox7 : 0) = 1. Diagnostic output for LED-related led test Output Analog DAC's. porstN Output Digital Power-on reset output. Generates the reference current. rref Input Analog Nominally set to 100KQ Bi-External temperature sensor input tsense Analog directional and temperature diagnostic output.

Temperature Sensor

Owing to the properties of the liquid crystal itself, a number of the programmable parameters have a strong dependence on temperature. For this reason, a temperature sensor is included. It is accessible to the host system as a read-only configuration register. The parameters that are temperature-sensitive are listed in the following table.

Table 118 vito dc bias red ito green ito blue ito ired igreen iblue vred vgreen vblue

With the exception of vito_dc_bias, the other parameters have been introduced above. Note that reset ito is not a relevant function of temperature.

Table 119 Bit Number Address 7 6 5 4 3 2 1 0 Ox14 vitodcbias Ox15 red-ito Ox16 green ito Oxl 7 blue-ito Ox18 igreen [l : 0] Ired [5 : 0] 0x19 iblue[3:0] igreen[5:2] vgrn[ 0x1a Vred[5:0] iblue[5:4] 0] Oxlb vblue [4: 1] vgreen [4: 1] Vblue Oxlc [0]

LUT Entry Mapping

Power-On Reset On power-up, or on the detection of a voltage spike during normal operation, on- board circuitry implements a prolonged reset, lasting approximately 100,000 cycles.

This starts from the rising edge of the voltage on the system power bus. In normal operation, without any power glitch, the negative activation of the reset pin rstN generates an internal reset that is released on the deactivation of rstN.

A power-on reset signal porstN is output from the OAC for use by the system, especially the Backplane IC.

Diagnostics Table 120: Diagnostic Configuration Addr Bits Init Name Description Ox07 == Charge pump clock test enable. When '1', routes internal charge pump clock out through the led pin. For 0 0 cpclktest diagnostics only: cpclktest should normally be left low, in which case led performs as an input from the BackplaneIC. Backplane IC status register. Status of Ox08 incoming control bits from the Backplane IC. Read only. 4-ito Input ito pin value. 3-led Input led pin value. 2-red Input red pin value. 1-green Input green pin value. 0-blue Input blue pin value. DAC Test register. For diagnostics 0x09 only. LED Voltage. When enabled, output 4 O vled_test at pin led test. LED Current. When enabled, output 3 0 iled test at pin led test. ITO Voltage. When enabled, output at 2 0 vito test pin led_test. Internal resistor balance adjustment 1 0 gain-test for vito. When enabled, output at pin ito_test. DC Bias voltage adjustment for vito. 0 1 bias test When enabled, output at pin ito test. OxOc Temperature test enable via pin tsense. For diagnostics only: 7 0 temp test temp test should normally be left low. External temperature sensor enable 6 0 ets via pin tsense. OxOf Temperature sample. Most recent 5: 0 temp_samp temperature reading. Read only.

Table 121: Temperature Sensor Diagnostics ets temp test Description 0 0 Normal mode: internal sensor. Diagnostics only: internal sensor voltage output to 0 1 pin tsense. 1 0 External sensor enabled via input pin tsense. 1 1 Diagnostics only: internal temperature DAC voltage output to pin tsense.

Electrical Characteristics General Specifications Table 122 Parameter Symbol Conditions Min Typ Max Units Absolute Maximum 5.5 V Supply Voltage Supply Voltage Vd33 3. 0 3. 3 3. 6 V Range (from host system) Operating Voltage V33 3. 0 3.3 3.6 V Power dissipation Internal only 10 mW Supply Current (Idle) 10 µA Operating-10 25 60 °C Temperature Storage Temperature-20 25-80 °C LatchUp 200 mA ESD See note 1 150 V 0

Note: Measured using the human body model per MIL-STD-883.

DC Characteristics Table 123 Parameter Symbol Condition Min Typ Max Units Input Low VIL 0 0. 3xVDD v Level Input High VIH 0. 7XV V Level DD Output Low VoL 0.4xVDD V Level Output High VoH 0. 8xV V Level DD POR Threshold 0.6 1.0 V POR TBD V Hysteresys POR Pulse 1.5 msec Width Bandgap Voltage Reference Table 124

Parameter Symbol Condition Min Typ Max Units Reference Voltage Vref Cref = 0.1 V OF Before trim @25°C 1.13 1.215 1.300 After trim 0 1.215 1.225 1.20 5 Output Impedance 1 PSRR @IKHz 70 dB Rref (current -1% 100 +1% K# generator) Output Voltage From-10°C-100 100 ppm/° Temperature to 60°C C Coefficient

Note: All current sources are referenced to reference current generated by the band- gap voltage generator, and the external precision resistor (Rref).

Master Clock Frequency Table 125 Parameter Symbol Conditions Min Type Max Units Master Clock MCLK 1 66 MHz Frequency Master Clock 0. 01 % Frequency tolerance 0.01 Master Clock Duty-10 50 10 % Cycle Master Clock Phase 5 % Jitter Charge Pump Clock 25 KHz (Driven from Master Clockinternally) LED Booster Clock 100 KHz Frequency (Driven from Master Clock internally)

ITO voltage Table 126 Parameter Symbol Conditions Mi Type Max Units n Positive ITO Voltage Vitop @25°C, 3.6 9. 0 V Vdd=3.3V Negative ITO Voltage Viton-5. 7-0. 3 V ITO voltage program Positive ITO 3.6 9. 0 V range # of programming bits 8 bits Vito tolerance-30 +30 mV Vito ripple Vitor ITO load = 5 mV lOuA Vito positive and (Vitop-1 5 mV negative voltage Vd3. 3) vs. mismatch Viton Vito positive and Vitoos (Vitop-300 mV negative DC voltage Vd3. 3/2) + 300 offset range Viton (Vito DC offset enabled) of programming bits Vitoos 8 bits for offset voltage generation # of programming bits Vitoos Adjust Center 8 bits for VCENTER (Vd33/2) using the trim (done in factory) same register as Vitoos (between Vd33 and gnd) to create OV Vitoos # of programming bits Vrmatch 8 bits for resistor mismatch cancellation (factory trim) Vito Temperature 200 ppm/° coefficient C

LEDs RED LED Table 127: Absolute Maximum Rating Parameter Symbol Conditions Mi Typ Max Units n Continuous Forward If r 50 mA Current Peak Forward Current Ifp_r Pulse width < 200 mA IOmS, Duty cycle=10% Power dissipation 120 mW Operating-10 25 80 °C Temperature Storage Temperature-30 85 °C Reverse Breakdown VRR_r @IR=100µA 5 V Voltage Table 128: Electrical Parameters Parameter Symbol Conditions Min Typ Max Units Red LED Forward VFR @IF3lmA 2. 47 V Voltage @IF=20mA (25 1.9 2.4 °C), 2.7 (10°C) Capacitance CLR @VF=0, 45 pF f=lMHz Forward voltage From 0. 5mA 1.6 3.0 range over operating to 120mA current (from-10 to 70°C) Full scale Output Vfs r 6 V voltage (Anode voltage DAC) Zero-Scale Output Vzs r 0 V voltage (Anode voltage DAC) Differential-1/2 1/2 LSB Nonlinearity (Anode voltage DAC) Maximum 5 Bits Resolution for voltage control (Anode voltage DAC) Monotonicity (Anode 5 Bits voltage DAC) Full scale Output Ifs r 120 mA Current (LED Current DAC) Zero-Scale Output Izsr 0 mA current (LED Current DAC) Differential-1/2 1/2 LSB Nonlinearity (LED Current DAC) Maximum 6 Bits Resolution for current control (LED CurrentDAC) Monotonicity(LED 6 Bits Current DAC) BLUE LED Table 129: Absolute Maximum Rating Parameter Symbol Conditions Mi Typ Max Units n Continuous Forward If b 30 mA Current Peak Forward Ifp_b Pulse width # 100 mA Current 10mS, Duty cycle=10% Power dissipation 120 mW Operating-20 25 80 °C Temperature Storage Temperature-30 85 °C Reverse Breakdown VRR_b @IR=100µA 5 V Voltage

Table 130: Electrical Parameters Parameter Symbol Conditions Min Typ Max Units Blue LED Forward VFB WI=62mA (m 3. 50 V Voltage easured) @IF=2OmA (Sp 3.6 4.0 ec'ed @25°C), (-10°C) 4.2 Capacitance CLR @VF=0, 45 pF f-1MHz Forward voltage From 0. 5mA 2.8 5.0 range over operating to lOOmA current (from-10 to 70°C) Full scale Output Vfs b 6 V voltage (Anode voltage DAC) Zero-Scale Output Vzs b 0 V voltage (Anode voltage DAC) Differential-1/2 1/2 LSB Nonlinearity (Anode voltage DAC) Maximum 5 Bits Resolution for voltage control (Anode voltage DAC) Monotonicity (Anode 5 Bits voltage DAC) Full scale Output Ifsb 120 mA Current (LED Current DAC) Zero-Scale Output Izsb 0 mA current (LED Current DAC) Differential-1/2 1/2 LSB Nonlinearity (LED Current DAC) Maximum 6 Bits Resolution for current control (LED Current DAC) Monotonicity (LED 6 Bits Current DAC) GREEN LED Table 131: Absolute Maximum Rating Parameter Symbol Conditions Mi Typ Max Units n Continuous Forward If_g 30 mA Current PeakForwardCurrent Ifp_g Pulse width # 100 mA 10mS, Duty cycle=10% Power dissipation 120 mW Operating-20 25 80 °C Temperature Storage Temperature-30 85 °C Reverse Breakdown Vnpg @lR=100iA 5 V Voltage

Table 132: Electrical Parameters Parameter Symbol Conditions Min Type Max Units Green LED forward Vfgrn gIf = 3. 74 V Voltage 34mA (measu red) 3.5 4.0 @IF=20mA (F rom data sheet 4.2 @25°C), (-10°C) Capacitance CLR @VF=0, 45 pF f-lMHz Forward voltage range From 0. 5mA 2.7 5.0 over operating current to 120mA (from-10 to 70°C) Full scale Output Vfs_g 6 V voltage (Anode voltage DAC) Zero-Scale Output Vzs g 0 V voltage (Anode voltage DAC) Differential-1/2 1/2 LSB Nonlinearity (Anode voltage DAC) Maximum Resolution 5 Bits for voltage control (Anode voltage DAC) Monotonicity (Anode 5 Bits voltage DAC) Full scale Output Ifs_g 120 mA Current (LED Current DAC) Zero-Scale Output Izs_g 0 mA current (LED Current DAC) Differential-1/2 1/2 LSB Nonlinearity (LED Current DAC) Maximum Resolution 6 Bits for current control (LED Current DAC) Monotonicity (LED 6 Bits Current DAC) Temperature Sensor Table 133 : LED

Parameter Symbol Conditions Min Typ Max Units Analog output Vrs-2 mV/° sensitivity 1 0°CSTopS60 C oc Analog Output VTOUT TA=25 C 600 mV voltage A/DSampling ftemp-1 +0.1 KHz Frequency 0.1% % Resolution 6 Bits Offset @25°C -70 70 mV Offset Trim Range-140 mV 140m V Offset Trim 2 mV Resolution A/DNonlinearity-1/2 1/2 LSB Monotonicity 6 Bits Mechanical Package Information The OAC is preferably assembled in a very low profile, surface mount plastic, 64- pin TQFP package. It can also be bumped with solder or gold, and then flip-chipped on a flexible Mylar or Kepton film. Soldering, welding, or gluing with conductive epoxy (for bumped dies) makes the connections. No bonding wires are used.

While various embodiments have been described above, it should be understood that they have been presented by way of example only, and not limitation. Thus, the breadth and scope of a preferred embodiment should not be limited by any of the above-described exemplary embodiments, but should be defined only in accordance with the following claims and their equivalents.