AN1368

Developing Embedded Graphics Applications using PIC®

Microcontrollers with Integrated Graphics Controller

FIGURE 1: RED + GREEN = YELLOW

Author: Pradeep Budagutta
Microchip Technology Inc.

INTRODUCTION
Graphic-enabled devices are used extensively in daily
life. They are found everywhere, including indoor
products, such as telephones, calculators, pagers, MP3
players, digital electric meters, smart remote and UPS
displays. They are also used in outdoor products, such
as traffic signals, taxi meters, bus displays, advertise- Human eyes perceive the light as a Yellow colored light
ment boards, etc. The list is virtually endless. A current instead of separate Red and Green colored lights. This
trend is that many existing devices are becoming color recognition property of the human eye is the
graphic-enabled because it is economically feasible, foundation of the RGB (Red, Green and Blue) model.
easy to use and the latest in technology. The model states that the human eye can be made to
This application note is intended to help engineers who perceive different colors by mixing appropriate
are designing their first graphic application. It describes proportions (intensities) of Red, Blue and Green colors.
the basic definitions and jargons of graphics applications Therefore, a ‘colored’ light can be formed by mixing
and it helps the engineer to understand the theory, different proportions of Red, Green and Blue colors.
necessary decision factors, hardware considerations, • Mixing the same proportions of three RGB colors
available microcontrollers and development tools. Soft- gives a Gray color
ware libraries and support are available from Microchip
• Mixing a zero amount of all RGB colors gives a
with further literature references for advanced users.
Black color
• Mixing a maximum amount of all RGB colors
BASICS OF COLOR SCIENCE gives a White color
In its purest form, color is associated with the Varying the intensity of light, while keeping the same
wavelength of light, within human visible range, from proportion of RGB, gives different shades of Gray,
about 400 nm (Violet) to 700 nm (Red), with Yellow which is also known as ‘Grayscale’. Using a single
centered at about 575 nm. That means, if a light of color (a fixed proportion of RGB) throughout an appli-
575 nm wavelength is incident on human eyes, it is cation gives a ‘Monochrome’ application, meaning a
perceived as a Yellow light. We have also learned that single color.
colors can be derived from three basic colors: Red, Since everything is represented in bits and bytes in a
Blue and Green. For example, Yellow can be derived digital system, then how can actual colors be repre-
by mixing Red and Green lights. Is this true? The sented as a number in the form of bits or bytes? Each
answer is both no and yes. It is no because mixing Red of these three basic colors (RGB) can represent a byte
and Green lights will constitute a mixture of lights with for a number ranging from 0 to 255. Therefore, with
wavelengths of 700 nm and 560 nm, and there is not a 3 bytes, we can represent 16 million colors (224) and
wavelength representing Yellow. The answer is yes this is termed as “True Color”. It is also common to use
because human eyes perceive this mixture as a Yellow 16 bits to represent colors. With 16 bits, we can
colored light. Therefore, we see the mixture of Red and represent 64K colors (216), which is sufficient for many
Green lights as a single Yellow light, as shown in graphics applications.
Figure 1. This is due to the color recognition properties
of the human eye.

 2011 Microchip Technology Inc. DS01368A-page 1

AN1368
In general, to divide 16 bits among Red, Green and FIGURE 3: GRAYSCALE VALUES OF
Blue, two schemes are used: 0 TO 225
• Scheme 1 (R<5> G<6> B<5>): In this scheme,
there are 5 bits of Red, followed by 6 bits of Green
and 5 bits of Blue. Green is given more bits
because of the property of the human eye, which
can distinguish more shades of Green than Red
and Blue. Figure 2 illustrates these 64K colors.
• Scheme 2 (T<1> R<5> G<5> B<5>): In this
scheme, there is one transparent bit, followed by
5 bits each of Red, Green and Blue. The
transparent bit indicates if the color should be
used or not.
Currently, the Microchip Graphics Library (Version 2.11)
supports only Scheme 1.

FIGURE 2: COLORS IN 16-BIT Alternatively, color may be represented using a Color

REPRESENTATION Look-up Table (CLUT), also called a palette table, where
the color is specified by the index of the table, as shown
in Figure 4. Depending on the size of the table, the bits
used to represent the index will vary as 256 entries of
RGB (8-bit index), 16 entries of RGB (4-bit index),
4 entries of RGB (2-bit index) and 2 entries of RGB (1-bit
index). This scheme is mainly used to save memory. For
more information on this scheme, see Appendix A:
“Color Look-up Table (CLUT)”.

FIGURE 4: COLOR LOOK-UP TABLE

(CLUT)

Grayscale is usually represented in a byte, with 0 as

Black, 1-254 as the shades of Black, getting lighter as
the number increases, and 255 as White, as shown in
Figure 3. Sometimes, only 4 or 2 bits are used to
represent 16 or 4 shades of Black, respectively. If only
one bit is used to represent either the on or off of a
color, then it is called ‘Monochrome’.
The number of bits required to represent a color is
called the ‘Color Depth’. For example, a color depth of
16 bits means it requires 16 bits to represent a color,
and therefore, we can represent 216 different colors.

DS01368A-page 2  2011 Microchip Technology Inc.

 AN1368
BASIC DISPLAY TERMINOLOGY A screen can be in Landscape mode (width > height) or
in Portrait mode (height > width). The ratio of the
A screen is made up of discrete elements, known as display screen’s visible width to its visible height is
pixels. Every pixel can show one point of color and called the ‘Aspect Ratio’. The most commonly used
each pixel is composed of three points: Red, Green aspect ratio is 4:3. The diagonal length of the display
and Blue. The colors are arranged next to each other screen is termed as the length of the display.
on a color screen, one point of intensity on a grayscale
For example, a display of 3.5'' means that the diagonal
screen or one point that can be set to on/off on a
length of the display is 3.5'', as shown in Figure 5.
monochrome screen. The number of such pixels in
horizontal and vertical directions is called the screen
resolution. For example, a resolution of 320x240
means there are 320 pixels horizontally (number of
columns) and 240 pixels vertically (number of rows).
Standard resolutions are given names, such as QCIF
(176x144), CIF (352x288),QVGA (320x240), WQVGA
(480x272), VGA (640x480) and WVGA (800x480), etc.
While mentioning the resolution, it is always better to
refer to the numbers instead of the names.

FIGURE 5: A 3.5'' QVGA DISPLAY IN LANDSCAPE MODE

320 Pixels

3.5''
240 Pixels

 2011 Microchip Technology Inc. DS01368A-page 3

AN1368
GRAPHICS SUBSYSTEM HARDWARE COMPONENTS OF A GRAPHICS
The hardware components required for a graphic appli- SYSTEM
cation, with their interconnection and design decisions, There are four basic components for any embedded
are described in the following subsections. graphics system, as illustrated in Figure 6. They
consist of the display glass, display controller, frame
buffer and the microcontroller.

FIGURE 6: THE FOUR BASIC COMPONENTS OF A GRAPHICS SYSTEM

Frame Display
Display Glass
Buffer Controller

Microcontroller

Display Glass All the LCD modules include gate and source drivers to
drive the voltage and current for displaying all the
Display glass is the device which displays a sequence of pixels. Table 1 gives a brief comparison of different
colors on the pixels and also converts the digital display technologies.
representation of colors to actual colors. The term, color,
includes grayscale. Generally, the types of displays used
are TFT LCDs, CSTN/MSTN LCDs or OLED/AMOLEDs.

TABLE 1: COMPARISON OF DIFFERENT DISPLAY TECHNOLOGIES (1)

Property TFT LCD STN (CSTN/MSTN) AMOLED
Frame Rate High Low High
Ghosting No Yes No
HBLANK and VBLANK Yes No Yes
Backlight Required Required Not Required
Cost Medium Low High
Typical Size for QVGA 1.5" to 5.7" 1.5" to 5.7" Up to 2.8"
(320x240) Resolution
Contrast Medium Low High
Power Consumption High Medium Low
Viewing Angle Medium Medium High
Note 1: Data mentioned in this table may change due to constant changes or advancements in the display
technology.

DS01368A-page 4  2011 Microchip Technology Inc.

 AN1368
Some of the important properties of display Frame Buffer
technologies are explained as follows:
The frame buffer is a memory (usually a RAM), which
• Frame Rate: The number of times the display holds the data to be shown on the display screen and
screen is refreshed in a second (this parameter acts as the data source for the display controller. The
does not reflect the refresh capacity of the required size of the frame buffer depends on the
microcontroller but only the capacity of the display resolution and color depth. The minimum requirement
glass). is that it should hold the data required to display one full
• Ghosting: When the screen is changed, the frame and must support the scan rate (preferred
previous frame is visible with lower intensity for a refresh rate as per the data sheet of the display glass)
fraction of time, which appears to be the ghost of of the display controller.
the current screen.
• HBLANK and VBLANK: Horizontal and vertical EQUATION 1:
blanking periods, where the display is not updated.
For more information, refer to Section 43. “Graph- Frame Buffer Size Required (Bytes) = Number of Pixels x
ics Controller Module (GFX)” (DS39731) in the Color Depth (Bits)/8
“PIC24F Family Reference Manual”.
• Backlight: For TFT/STN LCDs, backlight is EXAMPLE 1:
necessary to view the display. It could be either
CCFL or LED type. For a QVGA (320x240) display using a 16 BPP color
depth:
• Contrast: It is defined as the ratio of intensities of
white to black on the display. The higher the Frame Buffer = 320 x 240 x 16/8
contrast, the better is the display quality. = 153,600 Bytes
• Viewing Angle: It is the horizontal or vertical
= 150 Kbytes
angle within which the display is properly
viewable.
The display glass has no inherent memory and must be EXAMPLE 2:
constantly updated with pixel data in each row and
For a WQVGA (480x272) display using a 16 BPP
column, failing which, the display will go blank (similar
color depth:
to DRAM refresh). This refreshing of the display data is
handled by the display controller. Frame Buffer = 480 x 272 x 16/8
= 261,120 Bytes
Display Controller = 255 Kbytes
The display glass must be constantly refreshed by
feeding the horizontal and vertical pixel data repeatedly
EXAMPLE 3:
from the frame buffer. This task is performed by the
display controller. It fetches the data from the frame For a QCIF (176x144) display using an 8 BPP color
buffer, decodes it to the required bit format and feeds it depth:
to the display glass, along with proper control signals. Frame Buffer = 176 x 144 x 8/8
The display controller must adhere to the timing
requirements of the display glass. = 25,344 Bytes
= 24.75 Kbytes

 2011 Microchip Technology Inc. DS01368A-page 5

AN1368
Microcontroller controller keeps pumping data from the frame buffer
concurrently, with the microcontroller rendering pixels
The application code running inside the microcontroller into the frame buffer. However, the microcontroller
decides which data should be stored in the frame does not render any new shapes into the frame buffer
buffer, and as the frame buffer changes, the display if there is no change on the display screen. If there is a
content also changes. Each pixel’s color is calculated change on the screen, only the changed pixels need to
and stored in the frame buffer. The microcontroller and be sent to the frame buffer, thereby minimizing the data
the display controller must have the same settings, with transfer to the frame buffer.
respect to the color depth and memory range of the
frame buffer being used. The microcontroller must In Figure 7, only four pixels will be changed, and at a
have sufficient processing power (usually measured in 16-bit color depth, 4 * 16/8 = 8 bytes need to be sent
MIPS) to render the required shapes in the frame to the frame buffer.
buffer, such that it does not appear to be drawn slowly
on the display screen. This is because the display

FIGURE 7: PIXEL DATA UPDATE

Display Glass

Frame Display
Buffer Controller

Microcontroller

DS01368A-page 6  2011 Microchip Technology Inc.

 AN1368
INTEGRATION OF BASIC needs to understand the types of combinations of these
basic components that are possible in the form of ICs.
COMPONENTS
There are four types of possible combinations, as
The choice of how to integrate the four basic illustrated in Figure 8.
components is an important step in designing a Table 2 lists the advantages and disadvantages of the
graphics application. To make the choice, the designer four combinations of the basic components.

FIGURE 8: DIFFERENT WAYS OF INTEGRATING BASIC GRAPHICS’ COMPONENTS

Display Module

Parallel or Frame Display

Display Glass
Serial Buffer Controller

A. 2 Devices
Microcontroller

Graphic Controller Chip

Parallel or
Frame Display Display Glass
RGB
Serial Buffer Controller

B. 3 Devices Microcontroller

Frame Display
RGB Display Glass
Buffer Controller

Microcontroller
C. 2 Devices

SRAM

Frame Buffer
Parallel Display
RGB Display Glass
Buffer Controller

D. 3 Devices Microcontroller

 2011 Microchip Technology Inc. DS01368A-page 7

AN1368
TABLE 2: BASIC COMPONENTS
Options Advantages Disadvantages
A. The frame buffer and display • No specific IC is required for graphics • Generally higher cost
controller are housed in a single functionality • Usually forces a
module, called the ‘Display Module’. • Less system components and less PCB software driver change
The microcontroller and display space if the display module is
module interface through a serial or changed
parallel interface.(1) • May lack additional
memory required for
double-buffering,
animation, etc.
B. The frame buffer is housed • Software driver change is not required if • More system compo-
together with the display controller. the display glass is changed. Only a nents and more PCB
The microcontroller and graphics compile-time configuration change may be space
controller communicate through a needed. • Display size is limited
serial or parallel interface, whereas • Can be cheaper than Option A by the frame buffer
the graphics controller interfaces to inside the display
the display glass through an RGB controller
interface.(1)
C. The frame buffer and display • Only one IC is required for graphics • Display size is limited
controller are housed inside the functionality by the frame buffer
microcontroller. The microcontroller • Small form factor inside the
interfaces to the display glass • Usually the cheapest option microcontroller
through an RGB interface.
• Faster rendering since the memory is inside
Instead of a display controller, a
the microcontroller
combination of a parallel interface
and a DMA engine can be used as • Software driver change is not required if the
well. display module is changed. Only a
compile-time configuration change may be
needed.
D. The display controller is housed • The microcontroller can support the • Requires an extra IC
inside the microcontroller. Separate maximum display size possible as the size chip for the frame
RAM is used as the frame buffer. of the frame buffer can be selected by the buffer
The microcontroller interfaces to the user
display glass (display panel) through • Usually cheaper than Option A and B
an RGB interface and interfaces with
the frame buffer through a parallel
interface.
Instead of a display controller, a
combination of a parallel interface
and a DMA engine can be used as
well.
Note 1: A serial connection can be used on low resolution displays with low color depth (e.g., 1 BPP, 120x64). With
higher resolutions, the speed can be a bottleneck.

For Options A and B, the Microchip Graphics Library supports optional external RAM as a frame buffer with
currently supports PIC24, dsPIC® and PIC32 a parallel interface through the EPMP module. For
microcontrollers with PMP or EPMP. The more information on the device, refer to the respective
PIC24FJXXXDAXXX family can support Options C and device data sheet and also Section 43. “Graphics
D. PIC24FJ256DA210 contains 96 Kbytes of internal Controller Module (GFX)” (DS39731) from the
memory and has a graphics controller inside. It also “PIC24F Family Reference Manual”.

DS01368A-page 8  2011 Microchip Technology Inc.

 AN1368
POWER SEQUENCING IN DISPLAY Depending on the design of the display panel, all four
types of the power signals can be found in the panel
PANELS
data sheet. In some cases, only the digital power and
Different display panels will have different power supply backlight signals appear. This means that the panel
requirements and timing to enable each of the power has integrated an internal circuitry to generate the
signals of the panel. In some display panels, the analog and LCD power signals. This is usually true for
following power signals can be found: small displays (2'' to 4''). For larger displays, the analog
and LCD power signals tend to be higher in voltage
• Digital Power Signal
requirements. In these cases, it is not practical for the
• Analog Power Signal display panel to integrate such circuitry; therefore,
• LCD Power Signal those two power signals, that are specified as inputs,
• Backlight Power Signal must be externally provided.
The digital power signal is used to power-up the digital The different power signals of the display panel must
logic on the panel. The analog power signal is used to follow the power sequencing recommended by the
power the analog portion of the panel. The LCD power manufacturer. If the proper sequence is not followed,
signal, also known as the gate voltage, along with the the display panel’s life cycle can be reduced
digital data signals are primarily used to control the significantly. The typical power sequence of a panel is
pixel illumination. In some cases, the LCD power signal shown in Figure 9.
is composed of two power signals: the positive LCD Timing requirements are represented by t1, t2, t3, t4, t5
power signal and the negative LCD power signal. In and t6. In most cases, the requirements are
some cases, only one signal is available. The backlight represented as t1 = t4, t2 = t3 and t5 = t6.
illumination is controlled by the backlight power signal.

FIGURE 9: POWER SEQUENCE OF A PANEL

Voltage

t1 t2 t3 t4

VLCD

VA

VD

Time
t5 VBK t6

Data

Time
VD = Digital Supply VA = Analog Supply
VLCD = LCD Supply VBK = Backlight Supply
Data = Digital Signal from Controller

 2011 Microchip Technology Inc. DS01368A-page 9

AN1368
TOUCH SCREEN 4-WIRE RESISTIVE TOUCH SCREEN
Some applications require the support of a touch This touch screen has four signals, of which two are
screen for the display. This is achieved by using a purely digital signals. The other two signals are
separate touch screen on the display glass or by alternately configured as both digital and analog signals.
selecting a display module with a touch screen. In both The four signals can be directly connected to the
cases, the touch signals must be handled by either the microcontroller I/O pins with two digital inputs and two
microcontroller or a separate touch screen controller digital outputs, or analog pins. Figure 10 illustrates the
(such as Microchip’s AR1000 series touch screen connections for this scheme.
controllers). These touch signals are analog and digital
signals which must be decoded to sense the touch When the user touches the screen, the resistance of
coordinates. Transparent touch screens are usually of the screen changes. By measuring the resistance in
resistive type or capacitive type. Resistive touch horizontal and vertical directions, and comparing them
screens are the most commonly used and are with the calibrated values, the (x, y) coordinates of the
generally available in 4-wire or 5-wire configurations. point of touch can be obtained.
The touch point can be detected by measuring the When a point on the screen is touched, the x-coordinate
variation of the resistance of the touch screen. Only a voltage is obtained by applying voltages across the y-
4-wire touch screen is explained here. signal and measuring the analog voltage on the x-signal,
as shown in Figure 11. The y-coordinate voltage is
obtained by applying voltage across x-signals and
measuring the analog y-voltage, as shown in Figure 12.

FIGURE 10: 4-WIRE RESISTIVE TOUCH SCREEN

PIC® MCU
with ADC

Digital I/O

Digital I/O
X X
Digital I/O

Digital I/O

DS01368A-page 10  2011 Microchip Technology Inc.

 AN1368
FIGURE 11: MEASUREMENT OF THE X-VOLTAGE
Y
PIC®MCU
with ADC

Digital I/O
1
Digital I/O
Sense X
Dig-O/AN-I X
0
Dig-O/AN-I

FIGURE 12: MEASUREMENT OF THE Y-VOLTAGE

Y
PIC ® MCU
with ADC
1
Digital I/O

Digital I/O
0 X X
Dig-O/AN-I
Sense
Dig-O/AN-I

DECISION FACTORS Display Resolution and Size

After understanding the basic definitions and A particular resolution can be obtained in different
components of a graphics subsystem, the next step is display sizes. For example, QVGA (320x240) displays
to decide the specifications for the application. Some of are available in a size range of 1.5'' to 5.7''. As the size
the important factors that needs to be considered when increases, keeping the resolution constant, the pixels
deciding on specifications are as follows: will look coarser, that is, curved shapes on the screen
will appear blocky.
• Display Resolution and Size
• Display Orientation – Portrait or Landscape In an application, if the user needs to look at the display
from a short distance (e.g., hand held devices), higher
• Color Depth (BPP)
resolution displays are a better choice for larger
• Frame Buffer Size displays. If the user looks at the display from a long dis-
• Microcontroller Processing Power tance (e.g., token number of displays at banks), larger
• Configuration of Graphics Components sized displays with lower resolution may be used. If
• Frame Rate vs. MIPS pictures are being displayed, it is better to use a higher
• Interfacing with Unmatched Number of Display resolution. Figure 13 illustrates how ‘A’ appears on a
RGB Lines smaller sized lower resolution display, larger sized lower
resolution display and larger sized higher resolution
These decision factors are described in the following display, respectively.
sections.
FIGURE 13: DISPLAY OF ‘A’ AT
VARIOUS RESOLUTIONS

 2011 Microchip Technology Inc. DS01368A-page 11

AN1368
Display Orientation Color Depth Selection
Displays are available in Landscape (e.g., 320x240) or Along with the resolution of the display, the correct
Portrait (e.g., 240x320) modes. A landscape display choice of color depth is another decision factor since
can also be used in Portrait mode by setting a 90° this determines the size of the frame buffer (cost of
rotate function in the graphics library or display RAM). If natural photos are being displayed, it is better
controller. Similarly, a portrait display can also be used to go with 16 BPP or higher. If 256 different colors are
in Landscape mode. If rotating the pixels is enough for the application, then a color depth of 8 BPP
implemented by special hardware features inside the can be chosen (with the standard 256 colors provided
graphics controller, there is no penalty on the by the display controller or custom 256 colors using
performance. However, if the rotation is performed in CLUT (See Appendix A: “Color Look-up Table
software (such as the graphics library used), there is a (CLUT)”). This would reduce the RAM requirement by
penalty in the software performance. This is because 50%, compared to 16 BPP. If only 16 or 4 different
for every (x, y) point, a new rotated (x’, y’) point has to colors are sufficient, 4 BPP or 2 BPP can be used,
be calculated, which takes away some of the saving the RAM by 75% and 87.5%, respectively, as
processing power. compared to 16 BPP. Table 3 lists the RAM
Note the difference in the RGB strip alignment if the requirements for different color depths.
display is used in Rotated mode, as shown in
Figure 14.

FIGURE 14: LANDSCAPE AND PORTRAIT DISPLAYS USED IN LANDSCAPE MODE

TABLE 3: RAM SIZE REQUIREMENT FOR DIFFERENT COLOR DEPTHS

BPP for QVGA (320x240) 16 BPP 8 BPP 4 BPP 2 BPP
Number of Colors 65,536 256 16 4
RAM Size (Bytes) 153,600 76,800 38,400 19,200

DS01368A-page 12  2011 Microchip Technology Inc.

 AN1368
Frame Buffer Size Configuration of Graphics Components
The size of the frame buffer is calculated as follows: In Table 2, each configuration has its own advantages
and disadvantages.
EQUATION 2:
Most often, the decision to use one or another
Frame Buffer (Bytes) = Total_number_of_pixels x configuration is not influenced by the technical advan-
Color_Depth (in BPP)/8 tages or disadvantages, but rather by a supply chain
advantage or the availability of components. The
Table 3 shows an example for the QVGA (320x240) designer must balance between optimizing a design to
display. If the double-buffering technique is used, the meet the requirement and managing the supply chain.
frame buffer requirement will double (see Appendix B:
“Double-Buffering” for more information). Frame Rate VS. MIPS
If the frame buffer is inside the display controller or the
smart display module, and if the RAM is fixed, the Frame rate refers to the number of different frames that
maximum resolution that can be supported is limited. can be displayed in a second. This is a good
performance index for display of a video, but not for an
embedded GUI application. In general, an embedded
Processing Power (MIPS) application does not always change the entire screen,
The processing power required is application-specific. instead it changes a part of the screen, like a button or
It depends on how many graphic elements are a check box. The amount of change depends on the
displayed on the screen and the complexity of the size of the changed widget. The update time also
graphic elements. More processing power is required depends on factors, such as if the change belongs to a
to draw complex shapes, such as a circle, bevel, text, widget or an image. It is important to consider the
etc., rather than lines and rectangles. The processing worst-case scenario on the planned application. Initial
power requirements also depend on if a hardware calculation of frame rate and MIPS is important to get
graphics accelerator is available and used. Processing the preliminary requirements for the system. In addition
power requirements also depend on the update rate of to these calculations, it is recommended to evaluate
the screen elements. For many embedded graphics the system using development tools, such as
applications, ≥ 16 MIPS processing power could be evaluation kits.
sufficient. The best way to check the processing power
requirements is to evaluate using the standard
graphics development tools. (For more information on
development tools, see the “Development Tools”
section.

 2011 Microchip Technology Inc. DS01368A-page 13

AN1368
Interfacing with an Unmatched Number In Figure 15, the display panel has more RGB signal
of Display RGB Lines lines than the display controller. Here, all the RGB lines
of each color of the display controller are connected to
It is possible that the display controller’s number of the MSbs of the display’s display signal lines. The
RGB line outputs is different from the number of RGB unconnected LSbs may be connected to Ground or
line inputs of the display; it is still possible to interface VDD, or to the MSb of the same color. Connecting the
both of them. If the display’s RGB input lines are equal LSbs to the MSbs is the widely used method, since this
to the display controller’s RGB line outputs, there will enables the display to have a wider range of color
be no color degradation. If not, there may be a slight values.
color degradation because the display panel will be
unable to display all the colors generated by the display
controller. Usually, the former case is encountered
rather than the latter.

FIGURE 15: DISPLAY CONTROLLER’S DISPLAY SIGNALS LESS THAN LCD’S DISPLAY
SIGNALS

RGB 565 RGB 666 (Display)

Red [4] dp_Red [5]

Red [0] dp_Red [1]

Connect to VDD or GND or MSb dp_Red [0]

Green [5..0] dp _Green [5..0]

Blue [4..0] dp_Blue [5..1]

dp_Blue[5]

or dp_Blue[0]
Connect to VDD or
GND or MSb

DS01368A-page 14  2011 Microchip Technology Inc.

 AN1368
In Figure 16, the display LCD has less display signal The MSbs of the display lines of each color of the dis-
lines than the display controller. play controller are connected to all the display signal
lines of the LCD. The unconnected LSbs may be left
unconnected.

FIGURE 16: DISPLAY CONTROLLER’S DISPLAY SIGNALS ARE MORE THAN LCD’S DISPLAY
SIGNALS (POSSIBLE COLOR DEGRADATION)

RGB 666 RGB 585 (Display)

Red [5] dp_Red [4]

dp_Red [0]
Red [0] No Connect

Green [5..0] dp _Green [5..0]

Blue [4..0] dp_Blue[4..0]

Blue [0] No Connect

 2011 Microchip Technology Inc. DS01368A-page 15

AN1368
THE PIC24FJ256DA210 It can also interface with optional, external parallel
RAM through the Enhanced PMP module to increase
MICROCONTROLLER
the size of the frame buffer. The PIC24FJ256DA210
The PIC24FJ256DA210 device is a 16-bit microcon- graphics controller module is shown in Figure 17.
troller which supports a processing speed of up to
16 MIPS. The microcontroller includes 96 Kbytes of
internal RAM and a built-in display controller with
Graphics Processing Units (GPUs) to accelerate the
drawing of common 2D shapes.

FIGURE 17: PIC24FJ256DA210 GRAPHICS CONTROLLER MODULE

PIC24F Graphics
Controller Module
Display Interface
Clock (DISPCLK) Display Module
Interface

System Clock Registers GD<15:0>

and Control
Interface
HSYNC
Graphics
Controller Clock GPU Command

To Display
(G1CLK) Interface VSYNC

Glass
GEN
RCCGPU CHRGPU IPU

CLUT GPWR

Memory Request Arbiter GCLK

System
RAM

DS01368A-page 16  2011 Microchip Technology Inc.

 AN1368
• DISPCLK is the clock which drives the display • GPWR is the power supply control signal for the
glass. display glass. In some large displays, an external
• System clock is the clock speed at which the circuitry may be needed. Use this signal to enable
program accesses the Command/Control/Status or disable the external power circuitry. In displays
registers. that include an internal power circuitry, this signal
• G1CLK is the clock which drives the GPUs to can be connected to the display’s power enable
draw lines, rectangles, render characters and pin. This signal should not be used as a power
decode compressed data without the involvement supply line to the display glass.
of the processor. Table 4 lists the number of microcontroller pins
• External RAM, up to 16 MB, can be connected required for various display and RAM configurations.
through the EPMP module using a parallel inter- The Graphics Processing Units (GPUs) like the
face. The graphics module can use this on its own Character Graphical Processing Unit (CHRGPU),
without any involvement of the processor. The Rectangle Copy Graphics Processing Unit (RCCGPU)
interfaces allowed are limited to an 8-bit or 16-bit and Inflate Processing Unit (IPU) are the graphics
parallel connection. For more options and informa- accelerators. These accelerators are used for
tion, refer to Section 42. “Enhanced Parallel rendering characters, rectangles and to decompress
Master Port (EPMP)” (DS39730) in the “PIC24F the compressed data, respectively.
Family Reference Manual”.
These GPUs help to free the processing power of the
• HSYNC, VSYNC are the horizontal and vertical microcontroller, which can be used for the purpose of
synchronization signals to the display. the application. Instead of the CPU rendering the
• GCLK is the pixel clock. pixels, the application only needs to issue the
• GEN is a signal that varies in function for TFT and commands to draw primitive rendering functions (such
STN display types of interfaces. For TFT, this sig- as lines, bars and characters) to the screen. After
nal indicates that data lines are valid. For STN, issuing the commands, the CPU is free to perform
this signal toggles per line on the Line Toggle other application tasks. The application code runs in
mode and toggles per frame for the Frame Toggle parallel to the RCCGPU, which concurrently draws the
mode. For more information, refer to Section 43. line. However, care should be taken because returning
“Graphics Controller Module (GFX)” from a function, for example, Line(), need not imply
(DS39731) in the “PIC24F Family Reference that the line is completely drawn. This is called a non-
Manual”. blocking draw. The drawing can also be made blocking
• GD<15:0> carry the display RGB or Gray values by setting the proper compiler switch in the
as per the graphics module settings. Only the GraphicsConfig.h file, as explained in future
required number of lines is enabled, depending sections.
on the interface requirements of the display. (e.g.,
16 lined for TFT LCD’s RGB565 input or four lines
for MSTN’s grayscale input).

TABLE 4: MICROCONTROLLER PINS

Other (Clock
Display Data EPMP
Configuration and Sync) Total
Pins (RGB) Pins
Pins
A TFT LCD without External RAM (using CLUT 16 0 5 21
and 16-bit colors)
A 256-Color CSTN without External RAM 8 0 5 13
A TFT LCD with External 16-Bit Wide RAM of 16 37 5 58
256 Kbytes (using 16-bit colors)
A 16-Color MSTN without External RAM 4 0 5 9

 2011 Microchip Technology Inc. DS01368A-page 17

AN1368
The GPUs are briefly explained below: The RCCGPU can be used to achieve special effects,
• CHRGPU: Renders the characters on the display. such as screen animations, like scrolling, peeling, etc.
A font table must be loaded into the RAM and the For more information on the advanced usage of the
CHRGPU must point to that font table. The (x, y) RCCGPU, see Appendix C: “Advanced Usage of
coordinates must also be initialized on the appro- RCCGPU”.
priate CHRGPU registers. When a character code • IPU: Used to decompress a compressed data
and the draw command are given, the character using the DEFLATE algorithm with Fixed Huffman
will be rendered on the configured RAM area, codes. For example, images can be compressed
which can also be the frame buffer. The user must and kept in the internal Flash or external memory
take care of the display glass orientation as the and they can be decompressed into RAM during
characters cannot be rotated dynamically by the run time. Similarly, compressed user-specific data
CHRGPU. The CHRGPU does not support anti- can also be decompressed and used during run
aliased fonts; all the pixels on a character are of the time with the IPU. It should be noted that
same color. The background and foreground colors decompression can only commence from the
are set using the CHRGPU commands. If transpar- beginning of a compressed block and not from the
ency is enabled, only the foreground color is middle. For example, when storing multiple
drawn, and if the transparency is disabled, the images, compress each image to its own
background color is also drawn surrounding the compressed block. The IPU can be used to
character. To use the CHRGPU by default, uncom- decompress any images by specifying the
ment the line: #define USE_DRV_OUTCHAR in location of the desired compressed block. The
MicrochipGraphicsModule.h. Microchip Graphics Library will handle this
• RCCGPU: Used to draw horizontal or vertical scenario, making it transparent to the users.
lines, rectangles, filled rectangles, and to copy
rectangular regions. RCCGPU can perform the Note: Bit maps can be compressed by selecting
following three operations: the “IPU” option in the Graphics Resource
Converter (GRC) tool while converting the
- Copy – Copy a memory block from one part
images. GRC is a tool included in the
of the RAM to another. Depending on the
installation of the Graphics Library. The
command parameter, the block of memory
PutImage() API automatically decom-
can be a contiguous block or a rectangular
presses these compressed images using
block.
the IPU at run time. The user is required to
- Copy with Solid Fill – Fill a rectangular area allocate the required amount of RAM for
with a specific color. IPU operation, using compile-time options
- Copy with Transparency – Same as the copy as described in the Microchip Graphics
option, but a color set apart to indicate trans- Library Help file.
parency will not be copied to the destination,
leaving that part of the destination For more information on these GPUs and their
unchanged. registers, refer to Section 43. “Graphics Controller
Module (GFX)” (DS39731) in the “PIC24F Family
Each operation can use one of the 16 available logical Reference Manual”.
operations, called Raster Operations (ROPs), which is
applied while copying.
For example, source can be copied as is or the source
can be ORed with the destination area, or the source
can be ANDed with a separate region and copied to the
destination area. For more information on the Graphics
Controller Module (GFX) and the supported ROPs,
refer to the Section 43. “Graphics Controller Module
(GFX)” (DS39731) in the “PIC24F Family Reference
Manual”.

DS01368A-page 18  2011 Microchip Technology Inc.

 AN1368
DEVELOPMENT TOOLS • PIC24FJ256DA210 Development Board
(DM240312), which is a stand-alone board.
A fast and cost-effective way of evaluating the system
Both the boards require add-on display modules which
specification of an application is through the use of
are available with displays of various sizes. User-
existing development tools. Microchip has several
specific display panels can be used with the help of a
development tools supporting graphics design. Two
display prototype board. Figure 18, Figure 19, Figure 20
important tools that can be used for graphics
and Figure 21 illustrate these development tools. For the
development are:
latest tool set, visit: http://www.microchip.com/graphics.
• Graphics LCD Controller PICtail™ Plus SSD1926
Board (AC164127-5), which is an add-on board to
the Microchip’s generic development board for
16-bit and 32-bit microcontrollers, such as the
Explorer 16 board and PIC32 starter kits.

FIGURE 18: GRAPHICS PICtail™ PLUS DAUGHTER BOARD WITH 3.2'' DISPLAY KIT (AC164127-3)

FIGURE 19: DEVELOPMENT BOARD SUPPLIED WITH PIC24FJ256DA210 DEVELOPMENT KIT

(DV164039)

 2011 Microchip Technology Inc. DS01368A-page 19

AN1368
FIGURE 20: 4.3'' WQVGA POWERTIP TFT DISPLAY BOARD (AC164127-6)

FIGURE 21: GRAPHICS DISPLAY PROTOTYPE BOARDS (AC164139)

DS01368A-page 20  2011 Microchip Technology Inc.

 AN1368
SOFTWARE Features of the Microchip Graphics Library are:
• Works with 16-bit and 32-bit PIC® MCUs, as well
The basic software component required for any
as dsPIC® DSCs
graphics application is a Software Display Driver which
provides one basic operation (i.e., setting the color of a • Modular design, compile only what is required
pixel). A driver may also implement APIs to draw • Supports multiple user interfaces
fundamental shapes, for instance, a line, rectangle, • Not dependent on display size or resolution
bar, circle, text, image and so on. The Software Display • Low-cost, full-featured development tools
Driver must be written for every separate graphics
• Utilities to import fonts and images
driver used. More complex graphic elements, like
labels, buttons, check boxes, sliders and progress bars • Free to Microchip customers
are implemented in higher layers, which in turn, use the • Includes multiple low-level drivers
Software Display Driver. The structure of the Microchip Graphics Library is
Microchip provides a ‘free to use on PIC MCU’ software shown in Figure 22.
library, called “Microchip Graphics Library”, which The Microchip Graphics Library v2.11 is distributed
contains the above discussed drivers and higher along with the Microchip Applications Library and is
layers. Several demos are distributed with the graphics available for download at www.microchip.com/MAL.
library which the user can run out of the box on the
appropriate development tools.

FIGURE 22: STRUCTURE OF MICROCHIP GRAPHICS LIBRARY

Application Layer

User Message Interface

(Touch Screen, Keypad, and so on)

Graphic Object Layer

(Button, Slider, Edit Box, and so on)

Graphic Primitive Layer

(Line, Circle, Bar, and so on)

Display Device Driver Layer

(PutPixel, SetColor, and so on)

Graphic Display Controller/Display Panel

Note: The Microchip Graphics Library Version 2.11 is explained here. In future, the software structure may be modified. Refer
to the Help file of your Microchip Graphics Library for the latest information.

The Graphics Display Controller is the hardware module development board-specific files are in the folder,
consisting of the frame buffer and Display Controller. Microchip Solutions/Board Support Package. The
The remaining layers are the software layers. The project path must be set to point to these folders.
Microchip Graphics Library is organized in a set of Starting from the bottom-most Software layer to the
folders under the folder, ‘Microchip Solutions’. The ‘C’ top-most layer, the functionality of each layer and the
files are in the folder, Microchip Solutions/Microchip/ files responsible for those layers are explained further
Graphics, and the header files are in the folder, in the following sections. For more information, refer to
Microchip Solutions/Microchip/Include/Graphics. The the Help file of the Microchip Graphics Library.

 2011 Microchip Technology Inc. DS01368A-page 21

AN1368
DISPLAY DEVICE DRIVER LAYER GRAPHIC PRIMITIVE LAYER
Every hardware display controller has its own set of This is a layer above the Display Driver layer and
commands and status information. Therefore, separate provides most common services through APIs, which
software drivers are needed for each supported display are used to draw basic shapes, for instance, line
controller, which fulfills the requirements of the display (normal, thick, dashed), bar, rectangle, circle, polygon,
driver and the standard APIs defined by the Graphics bevel and arc. It also provides APIs for drawing text
Library. The list of supported drivers can be found in the and images. These are generic APIs which work with
Graphics Library Help file. To know if a display module is any given display driver. However, some of these APIs
supported, check if the display driver inside the display may be implemented by the Driver layer for optimized
module is available in the above mentioned folder. performance, especially if the driver supports 2D-
The main function provided by this layer is the Accelerations (For example, the Microchip Graphics
initialization of the driver using the API module and SSD1926). It is possible to write simple
ResetDevice(), painting a pixel using APIs, applications by using Primitive and Driver layers only
SetColor(color) and PitPixel(x, y), and and without using higher layers. To include this layer,
knowing the color of a pixel using the API the following files must be added to the project:
GetPixel(x, y). The other functionalities provided
HEADER FILES
are, for example, setting up of clipping area, getting the
maximum x and y values for a display screen, etc. An Primitive.h
application can be written using only this layer without
using any higher layers. In that case, all the shapes SOURCE FILES
must be drawn by the user. To include this layer, the Primitive.c
following files must be added to the project:
GRAPHIC OBJECT LAYER (GOL)
HEADER FILES
Graphics.h The GOL consists of many selectable objects, called
DisplayDriver.h or specific <Driver.h> ‘widgets’, such as Button, TextBox, Check Box,
(like SSD1926.h) ScrollBar, ProgressBar, Picture, ListBox, GroupBox,
Meter, DigitalMeter, Dial, Chart and Grid, which form
CONFIGURATION FILES the basic elements of a complex graphics application.
Each of these widgets is implemented in ‘C’, but with
HardwareProfile.h
basic object-oriented principles, and can be used as
GraphicsConfig.h modules. Use of this layer must be enabled at compile
These files are project-specific and must be inside the time in the ‘GraphicsConfig.h’ file. The use of
project folder. individual widgets can be enabled or disabled during
compile time in the ‘GraphicsConfig.h’ file, thereby
SOURCE FILES saving RAM and ROM. For more information, see the
Specific <Driver.c> (like SSD1926.c) “Configuration” section.
Each kind of widget has a default style scheme which
defines the font and the colors used for various parts
and states of the widget. For example, a button in a
pressed state has a different color than the released
button. The style scheme for each widget is explained
in detail in the Help file of the Microchip Graphics
Library. For each style scheme, some heap memory
(dynamically allocated memory) is required to store the
style scheme values. Heap is required to store the state
information for each enabled widget. The total heap
must be greater than the sum of heaps for all the
instances of used widgets. Aside from the heap
requirement, each widget type also needs to use RAM
space for variables when rendering and managing the
widgets. This additional RAM requirement is a constant
overhead for each type of widget. The difference
between the two is that the heap requirement is needed
for each instance of a widget, while the RAM
requirement is needed for each type of widget. The
RAM requirement is constant and not dependent on the
number of instances of one type of widget.

DS01368A-page 22  2011 Microchip Technology Inc.

 AN1368
For example, if the Release note says: Users can also create their own widgets and add to the
graphics library. See “References” for more details.
Module Button GOL
File Category Button
Heap for PIC24F 28 20 (per style
(per instance) scheme) Header Files • GOL.h
Heap for PIC32 44 24 (per style • Button.h
(per instance) scheme) • Chart.h
RAM for PIC24F 8 32 • CheckBox.h
RAM for PIC32 12 28 • DigitalMeter.h
ROM for PIC24F 1002 2076 • EditBox.h
ROM for PIC32 2748 5400 • Grid.h
• GroupBox.h
Note: The RAM and ROM requirements for • ListBox.h
PIC24F and PIC32 devices may be differ- • Meter.h
ent because of different microcontroller • Picture.h
architecture and different compilers. • ProgressBar.h
For a PIC24F application using only two buttons, a • RadioButton.h
RAM of 8 bytes, ROM of 1002 bytes and the required • RoundDial.h
heap memory would be 2 x 28 = 56 Bytes. • Slider.h
If one style scheme is used, then a heap memory of • StaticText.h
20 bytes would be required. • TextEntry.h
• Window.h
EQUATION 3:
• <CustomWidget.h>
Total Heap (Minimum Required Heap) = 20
Configuration • GraphicsConfig.h (to
(for the Style Scheme) + (2 x 28) = 76 bytes
Files select the usage of GOL and
Total RAM (for Graphics) = 32 (for GOL) + 8 = 40 bytes its individual widgets)
Source files • GOL.c
• GOLFontDefault.c
Note: This example is indicative only. It is
• Button.c
recommended to see the release notes of
the Microchip Graphics Library to derive • Chart.c
the appropriate values for that particular • CheckBox.c
release. • DigitalMeter.c
The GOL depends on the Primitive and the Display Driver • EditBox.c
layers. A function, GOLDraw(), must be called • Grid.c
continuously in a loop to simplify the drawing of widgets. • GroupBox.c
Additionally, a function, GOLDrawCallback(), must be • ListBox.c
implemented in the application code. This is used for
• Meter.c
custom drawing which is explained in the Help file.
Generally, this function can just return: TRUE. • Picture.c
• ProgressBar.c
To include the GOL, along with the files required for the
Primitive layer and Display Driver layer, the following • RadioButton.c
files must be added to the project. See the Help file for • RoundDial.c
the latest list of files. If the GOL is used, then in the • Slider.c
GraphicsConfig.h file, the macro, #define • StaticText.c
USE_GOL, must be defined. Individual macros for the • TextEntry.c
widgets used, such as #define USE_BUTTON, must
• Window.c
also be defined. If these individual macros are not
defined, the widgets will not be compiled even if they • <CustomWidget.c>
are included in the project.
Note: This list is for indication only. Refer to the
Microchip Graphics Library Help file for
the latest list of files.

 2011 Microchip Technology Inc. DS01368A-page 23

AN1368
USER MESSAGE INTERFACE After GOLDraw() is done, messages are received
from the touch screen driver and hard buttons driver.
The user message interface is a sublayer of the GOL The obtained message is passed to GOLMsg() to
which is enabled if the GOL is used. This sublayer is process and to output a widget-specific message. For
used to facilitate the message passing between example, it converts a “USER TOUCHED POSITION
widgets and user input. For example, if the user 100, 100” message to BUTTON1_PRESSED.
presses a button, then a message is sent to a call back
Additionally, the application must possess the
function, called GOLMsgCallback(), where the
GOLDrawCallback() and GOLMsgCallback()
message indicating that the button is pressed is
functions.
checked and an action is taken. This callback function
must be present in the application code if the GOL is If custom drawing is not done, then the draw callback is
being used, no matter if message passing is being used, as shown in Example 4.
used or not. If the message passing is not used, the
function body must return a ‘1’. EXAMPLE 4:
Similar to GOLDraw(), GOLMsg() must be called WORD GOLDrawCallback(void)
continuously in a loop inside the application code to {
facilitate message collection and passing. return (1);
}
The usage of GOLDraw(), GOLDrawCallback(),
GOLMsg() and GOLMsgCallback() are explained in The message callback handles the processed
Example 4, Example 5 and Example 6. message sent out by the widgets, as shown in
Example 6.
APPLICATION LAYER If the application already uses a main loop,
In this layer, the user has full control of the application. GOLDraw() and GOLMsg() can be called within the
Initially, the user must initialize the Microchip Graphics loop (see Example 5).
Library. The initialization is done by calling GOLInit() if
all the layers are being used, InitGraph() if the GOL Note 1: Refer to the application note, AN1136, “How
is not being used but the Primitive and Display Driver to Use Widgets in Microchip Graphics
layers are being used, or by calling ResetDevice() if Library” for creating a simple application.
only the Display Driver layer is being used. After the 2: Refer to the Microchip Graphics Library
initialization routine, the Primitive and Driver layers’ APIs Help file for the list of related application
can be called to achieve the required draw functionality. notes.
To use GOL objects (like buttons), the widgets must be
created by calling the widget’s create function (e.g.,
BtnCreate()), one by one, until all of the widgets are
created. This step will not display the widgets. The
created widgets are drawn on the screen when the
GOLDraw() function is called repeatedly in a while loop.
The messages are processed by calling the GOLMsg()
inside the same loop, as shown in Example 5.

DS01368A-page 24  2011 Microchip Technology Inc.

 AN1368
EXAMPLE 5:
while(1)
{
if(GOLDraw())
{ // Draw GOL objects
// Drawing is done here, process messages
TouchGetMsg(&msg); // Get message from touch screen
GOLMsg(&msg); // Process message
SideButtonsMsg(&msg); // Get message from side buttons
GOLMsg(&msg); // Process message
}
}

EXAMPLE 6:
WORD GOLMsgCallback(WORD objMsg, OBJ_HEADER *pObj, GOL_MSG *pMsg)
{
// beep if button is pressed
if(objMsg == BTN_MSG_PRESSED)
{
Beep();
}
}

 2011 Microchip Technology Inc. DS01368A-page 25

AN1368
CONFIGURATION 4. USE_FOCUS: If this option is defined, a dashed
outline (focus line) is displayed on the selected
The configuration of the Microchip Graphics Library is widget. This focus line is especially useful for
done through two files: navigation and selection of widgets if push
• GraphicsConfig.h buttons are used as user input.
• HardwareProfile.h 5. USE_TOUCHSCREEN: This option enables the
touch screen support for the application by
GraphicsConfig.h enabling the touch message processing part of
the GOL.
The software library related configurations are done in 6. USE_KEYBOARD: This option enables the
GraphicsConfig.h. The available options are: physical keys support for the application by
1. USE_NONBLOCKING_CONFIG: If this option is enabling the hard key message processing part
defined, then the Non-Blocking mode of the API of the GOL.
calls is used; that is, the APIs with return status 7. USE_GOL: This option enables the Graphic Object
values can return without completing their task. Layer (GOL). Initially, GOLInit() must be called.
The task may complete some time after GOLDraw() and GOLMsg() must be called
returning. This is especially useful if 2D- repeatedly in a loop and GOLDrawCallback()
Acceleration is supported by the display driver. and GOLMsgCallback() functions must be
In that case, the application can regain control implemented by the application.
while rendering is being processed in parallel. 8. USE_BUTTON, USE_CHECKBOX, USE_WINDOW,
This reduces Idle time for the microcontroller … , USE_CUSTOM: These options are valid only
waiting for the rendering of a primitive command if USE_GOL is defined. They enable the support
(such as line(), bar()). If this define is of each widget in the application and each
disabled (commented out), then all API calls enabled widget will use its share of RAM, ROM
return only after the completion of their task. For and heap.
non-blocking configuration, the user must check 9. USE_MULTIBYTECHAR: This will make the
the return value of the API to know if its XCHAR, 2 bytes long so that Unicode (UTF16)
execution got completed or not. is supported. Languages other than English
2. USE_DOUBLE_BUFFERING: If this option is characters can be supported by the usage of
defined, two buffers will be used. One buffer will Unicode. Enable this option if multiple language
be used as a draw buffer, where the next screen support is needed in the application and disable
is rendered, and the other as a frame buffer, (comment out) this option if only ASCII is used.
which holds the visible pixels of the screen. This If this option is disabled, XCHAR takes one byte
mode is used to avoid visible slow drawing on per character.
the screen, like rendering of a large area or 10. USE_FONT_FLASH: Supports fonts stored in
decoding and displaying of an image. This mode the internal Flash of the microcontroller to be
uses twice the amount of the frame buffer and used in the application.
the library currently supports this feature only for
11. USE_FONT_RAM: Supports fonts stored in the
certain graphics controllers (refer to the
RAM of the microcontroller to be used in the
documentation of the Microchip Graphics
application. This is used as font accelerations in
Library). If this option is commented out, then
PIC microcontrollers with a built-in display
only one buffer is used as the frame buffer and
controller.
the changes to the screen are visible at the
same time. 12. USE_FONT_EXTERNAL: Supports fonts
stored in the external memory (serial or paral-
3. USE_PALETTE: If this option is defined, Palette
lel Flash) to be used in the application. The
mode is enabled and the colors are taken from
application needs to implement a function,
the palette table. This option is available only to
WORD ExternalMemoryCallback(EXTDATA
controllers that have a built-in, programmable,
*memory, LONG offset, WORD nCount,
color look-up table (example: PIC24FJ256DA210
void *buffer), to get data from the external
has a graphics module with a color look-up table).
memory.
The table is required to initialize the palette
engine and set a valid palette table before 13. COLOR_DEPTH: Specifies the color depth used
displaying anything on the screen. The library in the demo in bits-per-pixel and it can take
supports this feature only for PIC microcontrollers values, such as 1, 2, 4, 8, 16 or 24 (limited by the
with a built-in display controller. The hardware capabilities).
COLOR_DEPTH setting can be 1, 2, 4 or 8 with the
palettes enabled. If disabled, Normal Color mode
will be used.

DS01368A-page 26  2011 Microchip Technology Inc.

 AN1368
14. USE_BITMAP_FLASH: Supports bit map 7. Touch Screen Definitions: These #defines
images stored in the internal Flash of the provide the definitions and setup information of
microcontroller to be used in the application. ADC channels used for sensing a touch in touch
15. USE_BITMAP_EXTERNAL: Supports bit map screen-enabled applications.
images stored in the external memory (serial or 8. RTCC Definitions: These #defines provide
parallel Flash) to be used in the application. The setup information for the RTCC module if an
application needs to implement a function, WORD internal Real-Time Clock (RTC) is used in the
ExternalMemoryCallback (EXTDATA application.
*memory, LONG offset, WORD nCount, 9. Communication Definitions: These
void *buffer), to get data from the external #defines provide the definitions and setup
memory. information of communication modules, such as
Note: Refer to the Microchip Graphics Library UART/USB channels used for communication
Help file for the latest set of configuration proposes in the application.
options. 10. Application-Specific Definitions: Any other
application-specific hardware definitions can be
HardwareProfile.h included here.
Refer to any graphics demo distributed with the
This file is used to configure hardware for an application Microchip Graphics Library for specific details or for
and is similar to the usage as in other Microchip software use as an example.
libraries. If a HardwareProfile.h already exists in the
application, the following definitions can be added to it, Note: Refer to the Microchip Graphics Library
and if not, a new file with the following definitions must Help file for the latest set of configuration
be created. This file mainly contains the following definitions.
sections:
1. GetSystemClock(): This macro must return OTHER GRAPHICS LIBRARIES FOR
the frequency of the system clock in Hertz.
PIC MICROCONTROLLER
2. GetPeripheralClock(): This macro must
return the frequency of the peripheral clock in Apart from Microchip, there are various third parties
Hertz. In PIC24 microcontrollers, this is half of that provide a graphics library for PIC MCUs. For
the system clock. For more information on example:
PIC32 devices, refer to the respective device • Segger’s emWin Graphics Library
data sheet.
• Micrium’s µC/GUI
3. GetInstructionClock(): This macro must
return the frequency of the instruction clock in
Hertz. In PIC24 microcontrollers, this is half of CONCLUSION
the system clock, and in PIC32 microcontrollers, Support for embedded graphics is becoming important
this is the same as the system clock. in recent user interface applications. Microchip
4. Display Related Settings: These settings supports their customers in multiple ways by providing
contain a set of #defines which defines the development tools, graphic libraries, graphic-enabled
various display related parameters, such as the microcontrollers (PIC24FJ256DA210) with extensive
type of the display (TFT or CSTN or MSTN), documentation, demo codes and support. Microchip’s
resolution of the screen, color depth, display graphics solutions can be successfully used for many
clock speed, various display timing parameters, embedded graphics applications resulting in less time
address of the draw buffer and frame buffer to market and lower cost.
among others.
5. I/O Ports for Keys: Provide definitions of the I/O
pins used in the application. These can be used
as inputs for switches or outputs to turn on LEDs,
or application controlled pins to enable specific
hardware.
6. External Memory Definitions: These
#defines provide the definitions of port pins
and/or module initialization, such as for SPI,
I2C™ or PMP to interface to the microcontroller
with the external memory (EEPROM, Flash or
SD Card).

 2011 Microchip Technology Inc. DS01368A-page 27

AN1368
APPENDIX A: COLOR LOOK-UP bytes, one for the RGB of each primary color. The other
colors are represented as a combination of the RGB
TABLE (CLUT)
values. Because each value is in the range of 0-255, it
A Recap of Basic Concepts provides a total of 16,777,216 possible colors; however,
this requires a huge storage memory (16 Mbytes).
Before understanding the concept of a CLUT, image
For example, for a resolution of 640x480, a 24-bit color
representations and their data types should be
image would require 900 Kbytes of memory (without
understood.
any compression).
A digital image consists of pixels, also known as Pels.
If memory space is a concern (which is generally the
This is a binary image (‘0’ or ‘1’) and is represented by
case), by quantizing the 24-bit color information,
a simple on/off of a pixel. It is also called a
reasonably accurate 8-bit color information can be
monochrome image, and for a 640x480 screen size,
achieved. This also means that we have only
the image size is (640x480/8) 37.5 Kbytes.
256 possible colors.
Image resolution refers to the number of pixels in an
image. Aspect ratio is the ratio of the column/row. In the 8-Bit Color Image Files
above example of 640x480, the aspect ratio is 4:3. This
Image files use a special concept to store color
above aspect ratio has been found to appear as a
information in a CLUT. The image is not represented by
natural image.
colors but a set of bytes. These bytes form the index to
8-Bit Gray Level Image a table, which has 3-byte values that specify the color
for a pixel. This means the user has to represent the
An 8-bit image is an image where each pixel has an 8-bit image by choosing the colors that best represent the
value (0-255) represented by a byte, which is also known image and does not exceed the 256 color
as Grayscale. Thus, the image can be a 2-dimensional combinations, as they are indexed by 8-bit values.
array of values, ranging between (0-255), which is also
One important savings of 8-bit representation over 24-bit
referred to as a bit map.
representation is in storage space, which is 300 Kbytes
For Example: An 8-bit Grayscale of VGA resolution vs. 900 Kbytes (with no compression applied).
would be the size of 300 Kbytes.
Figure A-1 can help in understanding the CLUT, which
is also known as a palette table.
Image Data Type for Color Images
The common data type for graphics and image file
formats is 24-bit color or 8-bit color. In a 24-bit color
image data type, each pixel is represented by three

FIGURE A-1: COLOR LOOK-UP TABLE (CLUT)

CLUT
0 Color 0

1
1 Pixel
Value = 21 2 210 200 9

3 R G B

.
.
21 Color 21 Color Value

DS01368A-page 28  2011 Microchip Technology Inc.

 AN1368
Figure A-1 shows a pixel which has RGB and it is Table A-1 represents the image using indexes instead
represented by a value of 21. The value, 21, in a look-up of RGB values from the palette table, as shown in
table, indexes the 24-bit RGB value for that pixel. Table A-2.

TABLE A-1: CONTENTS OF AN EXAMPLE IMAGE USING INDEXES INSTEAD OF RGB VALUES

0 0 0 0 1 1 1 1
0 0 0 0 1 1 1 1
0 0 0 0 1 1 1 1
0 0 0 0 1 1 1 1
2 2 2 2 3 3 3 3
2 2 2 2 3 3 3 3
2 2 2 2 3 3 3 3
2 2 2 2 3 3 3 3

TABLE A-2: PALETTE TABLE (24 BPP) The other way to device a CLUT is to scale the RGB
Color Index G B ranges to generate the 16-bit codes. However, since
R
the human eye is more sensitive to Green, we can give
0 0 255 0 Red and Blue 5 bits each, with 6 bits for Green, and
1 255 0 0 each cell in the image gets replaced by its index. This
may lead to edge artifacts in the image. Therefore, this
2 0 0 255 is not suitable for natural images but may be suitable
3 0 255 255 for artificially created images.
. . . . PIC24FJ256DA210 supports up to a 256-entry CLUT
. . . . with 16 BPP per entry.
255 219 200 198
Median Cut Algorithm
Note: A color index refers to one row of CLUT. It is an adaptive method that tries to put most bits where
A simple color changing type of animation process is colors are most clustered, so that they can be discrim-
possible by changing the CLUT in the above method. inated. This is the most general algorithm used while
This is also known as palette animation. converting a RGB image to a CLUT-based color image,
such as PC applications (e.g., GIMP).
In the above explanation, we saw a simple way to
cluster the image and device a CLUT, but clustering is
a slow process.

 2011 Microchip Technology Inc. DS01368A-page 29

AN1368
APPENDIX B: DOUBLE-BUFFERING 3. In the swap stage, the draw buffer is swapped
with the frame buffer (by exchange of pointers),
Manipulating pixels on the screen requires direct writes as shown in Figure B-3.
to the frame buffer. While these changes are being
executed, the screen is also refreshed. This means that FIGURE B-3: SWAP STAGE
the changes are displayed immediately as the frame
buffer is being updated. This is not a suitable scheme
when the changes are slow, for example, decoding an
image or having a large number of widgets on a screen. Exchange
The display may appear slow in that case and can Draw Buffer Frame Buffer
cause user dissatisfaction.
One solution to this problem is to use a double-buffering 4. In the final stage, the frame buffer is copied to
scheme supported by the Microchip Graphics Library. In the draw buffer, so the buffers have the same
this scheme, the changes are not directly written to the contents, as shown in Figure B-4.
frame buffer, but instead, they are written to a separate
buffer, called the ‘Draw Buffer’. After all the changes are FIGURE B-4: FINAL STAGE
made, the draw buffer and frame buffer are exchanged.
Now the draw buffer becomes the frame buffer, and
because of all the changes drawn there, the changes
appear spontaneous to the user. Of course, there will be Copy
a delay, as all the changes have to be written to the draw Draw Buffer Frame Buffer
buffer before displaying it. This delay is generally more
tolerable than displaying the changes slowly. 5. The workflow cycle continues.
After exchanging of the buffers, the latest buffer (which
is now the frame buffer) is copied to the new draw buf- When to Use Double-Buffering?
fer in the background, then the new draw buffer is in
sync with what is being displayed. New changes are Use double-buffering when:
then written to the draw buffer and the cycle continues. • There are large numbers of widgets on a screen
As the double-buffering scheme uses two buffers, the • The decoding and displaying of images are
RAM requirement will double. required
In the Microchip Graphics Library, if double-buffering is • The display resolution is larger than the WQVGA,
enabled, the frame buffer and draw buffer are even if the above points do not hold true
exchanged after the changes of all the widgets on a
screen are done (i.e., the new screen appears after the When Not to Use Double-Buffering?
whole screen is updated and not after updating an
individual widget). Do not use double-buffering when:

The workflow of double-buffering is as follows: • The immediate display of changes is required. For
example, plotting a graph.
1. In the initial stage, the draw buffer and frame buffer
• The RAM available for display is limited.
have the same contents, as shown in Figure B-1.
• The display resolution is smaller than the
FIGURE B-1: INITIAL STAGE WQVGA (this is not a hard rule).

How to Use Double-Buffering?

Double-buffering can be used as follows:
Draw Buffer Frame Buffer • Enable double-buffering by defining,
USE_DOUBLE_BUFFERING in the
2. In the drawing stage, only the draw buffer is GraphicsConfig.h.
modified, and therefore, the frame buffer is not • See the section on double-buffering in the
visible to the user, as shown in Figure B-2. Microchip Graphics Library Help file for code
examples.
FIGURE B-2: DRAWING STAGE

Draw Buffer Frame Buffer

DS01368A-page 30  2011 Microchip Technology Inc.

 AN1368
APPENDIX C: ADVANCED USAGE FIGURE C-2: EXPANDING RECTANGLE
OF RCCGPU 1

The RCCGPU can be used to achieve different kinds of

graphical effects, for instance, animations and scrolling
screens. These effects are achieved using a buffer of
memory, that is separate from the frame buffer, which
contains the final image to be displayed on the screen. Draw Frame
Buffer Buffer
The image from this buffer is then transferred to the 2
frame buffer in special ways which forms the various
visual effects on the screen. For example, transferring Algorithm:
the image vertically, line by line to the frame buffer,
Step i: The new screen is completely created in the
gives the effect of peeling, whereas moving the existing
draw buffer.
frame buffer, line by line, and filling the moved part with
the final image, gives the effect of scrolling the screen. Step ii: Move a rectangle of 2x2 from the middle of the
draw buffer to the middle of the frame buffer using the
Some of the effects and their algorithm are as follows:
RCCGPU rectangle copy command.
1. Move from Left to Right: This effect shows the
Step iii: Move a rectangle of 3x3 from the middle of the
new screen sliding from left to right.
draw buffer to the middle of the frame buffer using the
RCCGPU rectangle copy command.
FIGURE C-1: MOVE FROM LEFT TO RIGHT
Repeat the above steps until the whole screen is
1
transferred to the frame buffer, as shown in Figure C-2.
3. Expanding Line: This effect shows a vertical line
expanding from the middle to the end in a
horizontal direction.

FIGURE C-3: EXPANDING LINE

Draw Buffer Frame Buffer 1

2

Algorithm:
Step i: The new screen is completely created in the
Draw Buffer Frame Buffer
draw buffer.
Step ii: Move one right line (Height = Screen Height, 2
Width = One Pixel) from the draw buffer to the left side
of the frame buffer using the RCCGPU rectangle copy
command. Algorithm:

Step iii: Move two right lines (Height = Screen Height, Step i: The new screen to be shown is completely
Width = 2 Pixels) from the draw buffer to the left side of created in the draw buffer.
the frame buffer using the RCCGPU rectangle copy Step ii: Move two middle lines (Height = Screen Height,
command. Width = 2 Pixels) from the draw buffer to the middle of
Repeat the above steps until the whole screen is the frame buffer using the RCCGPU rectangle copy
transferred to the frame buffer, as shown in Figure C-1. command.

2. Expanding Rectangle: This effect shows a Step iii: Move four middle lines (Height = Screen Height,
rectangle expanding from the middle to the Width = 4 Pixels) from the draw buffer to the middle of
periphery of the screen. the frame buffer using the RCCGPU rectangle copy
command.
Repeat the above steps until the whole screen is
transferred to the frame buffer, as shown in Figure C-3.
Similarly, many screen transition effects can be
achieved easily.

 2011 Microchip Technology Inc. DS01368A-page 31

AN1368
APPENDIX D: ABBREVIATIONS REFERENCES
Table D-1 lists the abbreviations used in this document. For additional information on the use of the Microchip
Graphics Library, refer to following application notes:
TABLE D-1: ABBREVIATIONS • , AN1136, “How to Use Widgets in Microchip
Graphics Library” (DS01136), Paolo Tamayo,
Abbreviation Definition Microchip Technology Inc., 2007.
1 Kbyte 1024 Bytes • , AN1246 “How to Create Widgets in Microchip
AMOLED Active-Matrix Organic Graphics Library” (DS01246), Paolo Tamayo and
Light-Emitting Diode Harold Serrano, Microchip Technology Inc., 2009.
• AN1227, “Using a Keyboard with the Microchip
CCFL Cold Cathode Fluorescent Lamps
Graphics Library” (DS01227), Anton
CHRGPU Character Graphics Processing Alkhimenok, Microchip Technology Inc., 2008.
Unit
• AN1182 “Fonts in the Microchip Graphics Library”
CIF Common Intermediate Format (DS01182), Paolo Tamayo, Microchip Technology
CLUT Color Look-up Table Inc., 2010.
CSTN Color Super-Twisted Nematic • DEFLATE Compressed Data Format Specification
(RFC1951) — http://www.ietf.org/rfc/rfc1951.txt
DRAM Dynamic Random Access Memory
EPMP Enhanced Parallel Master Port For additional information on Microchip products
GPU Graphical Processing Unit or graphics applications, refer to the link:
GUI Graphical User Interface http://www.microchip.com/graphics
HBLANK Horizontal Blanking
IC Integrated Circuit
IPU Inflate Processing Unit
LCD Liquid Crystal Display
LED Light-Emitting Diode
MIPS Million Instructions per Second
MSTN Monochrome Super-Twisted
Nematic
OLED Organic Light-Emitting Diode
PCB Printed Circuit Board
PMP Parallel Master Port
QCIF Quarter Common Intermediate
Format
QVGA Quarter Video Graphics Array
RCCGPU Rectangle Copy Graphics
Processing Unit
RGB Red Green Blue
RTCC Real-Time Clock and Calender
STN Super-Twisted Nematic
TFT-LCD Thin Film Transistor Liquid Crystal
Display
UART Universal Asynchronous Receiver/
Transmitter
USB Universal Serial Bus
VBLANK Vertical Blanking
VGA Video Graphics Array
WQVGA Wide Quarter Video Graphics
Array

DS01368A-page 32  2011 Microchip Technology Inc.

Note the following details of the code protection feature on Microchip devices:
• Microchip products meet the specification contained in their particular Microchip Data Sheet.

• Microchip believes that its family of products is one of the most secure families of its kind on the market today, when used in the
intended manner and under normal conditions.

• There are dishonest and possibly illegal methods used to breach the code protection feature. All of these methods, to our
knowledge, require using the Microchip products in a manner outside the operating specifications contained in Microchip’s Data
Sheets. Most likely, the person doing so is engaged in theft of intellectual property.

• Microchip is willing to work with the customer who is concerned about the integrity of their code.

• Neither Microchip nor any other semiconductor manufacturer can guarantee the security of their code. Code protection does not
mean that we are guaranteeing the product as “unbreakable.”

Code protection is constantly evolving. We at Microchip are committed to continuously improving the code protection features of our
products. Attempts to break Microchip’s code protection feature may be a violation of the Digital Millennium Copyright Act. If such acts
allow unauthorized access to your software or other copyrighted work, you may have a right to sue for relief under that Act.

Information contained in this publication regarding device Trademarks

applications and the like is provided only for your convenience
The Microchip name and logo, the Microchip logo, dsPIC,
and may be superseded by updates. It is your responsibility to
KEELOQ, KEELOQ logo, MPLAB, PIC, PICmicro, PICSTART,
ensure that your application meets with your specifications.
PIC32 logo, rfPIC and UNI/O are registered trademarks of
MICROCHIP MAKES NO REPRESENTATIONS OR
Microchip Technology Incorporated in the U.S.A. and other
WARRANTIES OF ANY KIND WHETHER EXPRESS OR countries.
IMPLIED, WRITTEN OR ORAL, STATUTORY OR
OTHERWISE, RELATED TO THE INFORMATION, FilterLab, Hampshire, HI-TECH C, Linear Active Thermistor,
INCLUDING BUT NOT LIMITED TO ITS CONDITION, MXDEV, MXLAB, SEEVAL and The Embedded Control
QUALITY, PERFORMANCE, MERCHANTABILITY OR Solutions Company are registered trademarks of Microchip
FITNESS FOR PURPOSE. Microchip disclaims all liability Technology Incorporated in the U.S.A.
arising from this information and its use. Use of Microchip Analog-for-the-Digital Age, Application Maestro, CodeGuard,
devices in life support and/or safety applications is entirely at dsPICDEM, dsPICDEM.net, dsPICworks, dsSPEAK, ECAN,
the buyer’s risk, and the buyer agrees to defend, indemnify and ECONOMONITOR, FanSense, HI-TIDE, In-Circuit Serial
hold harmless Microchip from any and all damages, claims, Programming, ICSP, Mindi, MiWi, MPASM, MPLAB Certified
suits, or expenses resulting from such use. No licenses are logo, MPLIB, MPLINK, mTouch, Omniscient Code
conveyed, implicitly or otherwise, under any Microchip Generation, PICC, PICC-18, PICDEM, PICDEM.net, PICkit,
intellectual property rights. PICtail, REAL ICE, rfLAB, Select Mode, Total Endurance,
TSHARC, UniWinDriver, WiperLock and ZENA are
trademarks of Microchip Technology Incorporated in the
U.S.A. and other countries.
SQTP is a service mark of Microchip Technology Incorporated
in the U.S.A.
All other trademarks mentioned herein are property of their
respective companies.
© 2011, Microchip Technology Incorporated, Printed in the
U.S.A., All Rights Reserved.
Printed on recycled paper.

ISBN: 978-1-61341-123-0

Microchip received ISO/TS-16949:2002 certification for its worldwide

headquarters, design and wafer fabrication facilities in Chandler and
Tempe, Arizona; Gresham, Oregon and design centers in California
and India. The Company’s quality system processes and procedures
are for its PIC® MCUs and dsPIC® DSCs, KEELOQ® code hopping
devices, Serial EEPROMs, microperipherals, nonvolatile memory and
analog products. In addition, Microchip’s quality system for the design
and manufacture of development systems is ISO 9001:2000 certified.

 2011 Microchip Technology Inc. DS01368A-page 33

 Worldwide Sales and Service
AMERICAS ASIA/PACIFIC ASIA/PACIFIC EUROPE
Corporate Office Asia Pacific Office India - Bangalore Austria - Wels
2355 West Chandler Blvd. Suites 3707-14, 37th Floor Tel: 91-80-3090-4444 Tel: 43-7242-2244-39
Chandler, AZ 85224-6199 Tower 6, The Gateway Fax: 91-80-3090-4123 Fax: 43-7242-2244-393
Tel: 480-792-7200 Harbour City, Kowloon Denmark - Copenhagen
India - New Delhi
Fax: 480-792-7277 Hong Kong Tel: 45-4450-2828
Tel: 91-11-4160-8631
Technical Support: Tel: 852-2401-1200 Fax: 45-4485-2829
Fax: 91-11-4160-8632
http://www.microchip.com/ Fax: 852-2401-3431
India - Pune France - Paris
support
Australia - Sydney Tel: 91-20-2566-1512 Tel: 33-1-69-53-63-20
Web Address:
Tel: 61-2-9868-6733 Fax: 91-20-2566-1513 Fax: 33-1-69-30-90-79
www.microchip.com
Fax: 61-2-9868-6755 Germany - Munich
Atlanta Japan - Yokohama
China - Beijing Tel: 49-89-627-144-0
Duluth, GA Tel: 81-45-471- 6166
Tel: 86-10-8528-2100 Fax: 49-89-627-144-44
Tel: 678-957-9614 Fax: 81-45-471-6122
Fax: 86-10-8528-2104 Italy - Milan
Fax: 678-957-1455 Korea - Daegu
China - Chengdu Tel: 39-0331-742611
Boston Tel: 82-53-744-4301
Tel: 86-28-8665-5511 Fax: 39-0331-466781
Westborough, MA Fax: 82-53-744-4302
Fax: 86-28-8665-7889 Netherlands - Drunen
Tel: 774-760-0087 Korea - Seoul
Fax: 774-760-0088 China - Chongqing Tel: 82-2-554-7200 Tel: 31-416-690399
Tel: 86-23-8980-9588 Fax: 82-2-558-5932 or Fax: 31-416-690340
Chicago
Itasca, IL Fax: 86-23-8980-9500 82-2-558-5934 Spain - Madrid
Tel: 630-285-0071 China - Hong Kong SAR Tel: 34-91-708-08-90
Malaysia - Kuala Lumpur
Fax: 630-285-0075 Tel: 852-2401-1200 Tel: 60-3-6201-9857 Fax: 34-91-708-08-91
Cleveland Fax: 852-2401-3431 Fax: 60-3-6201-9859 UK - Wokingham
Independence, OH China - Nanjing Tel: 44-118-921-5869
Malaysia - Penang
Tel: 216-447-0464 Tel: 86-25-8473-2460 Fax: 44-118-921-5820
Tel: 60-4-227-8870
Fax: 216-447-0643 Fax: 86-25-8473-2470 Fax: 60-4-227-4068
Dallas China - Qingdao Philippines - Manila
Addison, TX Tel: 86-532-8502-7355 Tel: 63-2-634-9065
Tel: 972-818-7423 Fax: 86-532-8502-7205 Fax: 63-2-634-9069
Fax: 972-818-2924
China - Shanghai Singapore
Detroit Tel: 86-21-5407-5533 Tel: 65-6334-8870
Farmington Hills, MI Fax: 86-21-5407-5066 Fax: 65-6334-8850
Tel: 248-538-2250
China - Shenyang Taiwan - Hsin Chu
Fax: 248-538-2260
Tel: 86-24-2334-2829 Tel: 886-3-6578-300
Indianapolis Fax: 86-24-2334-2393 Fax: 886-3-6578-370
Noblesville, IN
Tel: 317-773-8323 China - Shenzhen Taiwan - Kaohsiung
Fax: 317-773-5453 Tel: 86-755-8203-2660 Tel: 886-7-213-7830
Fax: 86-755-8203-1760 Fax: 886-7-330-9305
Los Angeles
Mission Viejo, CA China - Wuhan Taiwan - Taipei
Tel: 949-462-9523 Tel: 86-27-5980-5300 Tel: 886-2-2500-6610
Fax: 949-462-9608 Fax: 86-27-5980-5118 Fax: 886-2-2508-0102

Santa Clara China - Xian Thailand - Bangkok

Santa Clara, CA Tel: 86-29-8833-7252 Tel: 66-2-694-1351
Tel: 408-961-6444 Fax: 86-29-8833-7256 Fax: 66-2-694-1350
Fax: 408-961-6445 China - Xiamen
Toronto Tel: 86-592-2388138
Mississauga, Ontario, Fax: 86-592-2388130
Canada China - Zhuhai
Tel: 905-673-0699 Tel: 86-756-3210040
Fax: 905-673-6509 Fax: 86-756-3210049

02/18/11

DS01368A-page 34  2011 Microchip Technology Inc.