Best of Shareware/Best of PC Windows Shareware 1.0 - Wayzata Technology (7111) (1993).iso/mac/DOS/CAD_CAM/GCPVU/GERBER.DOC

Text File | 1992-05-11 | 15KB | 428 lines

     UNDERSTANDING YOUR CAD OUTPUT: BASIC PHOTOPLOTTING PRINCIPLES

         A supporting document for GraphiCode's GC-PREVUE: Shareware PCB
         CAD output data viewer/checkplot utility

         Copyright (C) GraphiCode Inc 1991,1992 All Rights Reserved.



      What's a Photoplotter?


     A photoplotter is just what the name implies: an plotter that
     writes using light.  A plotter has to be told:

       Which tool to use.
       When to use the tool, and when not to.
       Where to go next.
       Whether to go there in a straight line or along an arc.

     For a photoplotter, "tool" means specially shaped apertures
     through which light passes to create a given shape on film.  An
     aperture can be used without movement to make a shape (a "flash")
     or with movement to make a line or an arc.

     There are two major types of photoplotters, "Vector" and "Raster"
     (or "laser").  Each handles apertures differently.

     Vector Photoplotters

     Aperture Wheels

     Traditionally, the photoplotter counterpart to a pen plotter's
     pen rack has been the aperture wheel.  The aperture wheel is a
     disk with 24 or 70 apertures arrayed radially along its
     circumference.

     When the photoplotter selects an aperture, the aperture wheel is
     rotated to place the desired aperture between the light source
     and the film.  Apertures are themselves pieces of film and can be
     made to any shape required, although in practice this is a time-
     consuming process and there is a physical limitation on size.

     Flash and Draw Apertures

     To achieve constant exposure on a vector photoplotter, apertures
     used for flashing pads are filtered differently than those used
     for drawing traces.  Therefore, Flash and Draw apertures cannot
     be used interchangeably without risk of localized over-exposure
     and under-exposure.

     Aperture Wheel Setup for Vector Plotters

     The setup of an aperture wheel is an exacting and time consuming
     process since each aperture in the wheel must be hand-mounted and
     aligned.  In order to avoid repeated setup costs, designers have
     the photoplotting vendor keep a wheel on file and are forced to
     always use that same set of apertures.  This has obvious



     drawbacks, both in terms of design flexibility and the ease of
     migration to other vendors.

     Raster (Laser) Plotters

     Aperture Lists

     Increasingly, vector photoplotters are being replaced by the
     laser photoplotter, which emulates the older style machine in a
     raster (bit-map) fashion.  While use of the term "aperture" to
     describe a pad or trace shape persists, the term "aperture wheel"
     is now being replaced by "aperture list", which implies the
     greater flexibility now available to the designer.

     There are three principle advantages with aperture lists on
     raster plotters:

       Aperture shapes can be easily generated in software, thus
       eliminating the need to design a physical wheel.
       More apertures can be defined on a list.
       Allowable apertures sizes are typically (but not always)
       greater than those imposed by the physical dimensions of an
       aperture wheel.

     Flash and Draw Apertures

     No distinction need be made between Flash and Draw aperture types
     since the light source intensity is constant.

     Speed Advantage of Laser Plotters

     Laser plotters operate much quicker than vector machines.  A
     complex plot that required hours on a vector machine can usually
     be performed in ten minutes or less on a laser photoplotter.
     This decreases turnaround time and in many markets has driven
     photoplotting costs down.

     Talking to Photoplotters

     The de facto standard for photoplotter data is the Gerber format,
     more properly known as RS-274D.  The term Gerber refers to the
     Gerber Scientific Instrument company, a pioneer and leader in
     photoplotter manufacturing.

     RS-274D is a variation on traditional Numerical Control (NC)
     machine tool languages.  It differs from traditional NC formats
     (i.e. drill data), as far as its use of tool selection codes but
     is otherwise compatible.

     RS-274D data is organized in "blocks".  A block consists of a
     combination of codes:



       Tool selection
       Setup
       Movement
       And, an End Of Block (EOB) character, which only follows a
       combination of the above codes.

     An EOB character is usually an asterisk ('*') or dollar ('$'),
     optionally followed by a carriage return and line feed.

     An RS-274D code consists of a letter D,G,M,X,Y,I or J followed by
     a numerical value.  These codes designate the following:

       * - End of Block (end of command)

       D - Select aperture, or set aperture use mode

       X - Move to X value

       Y - Move to Y value

       G - Various setup codes

       M - Various control codes

       I - Relative X location for arc center

       J - Relative Y location for arc center

     D Codes

     D codes have multiple purposes.  The first is to control the
     state of the light being on or off. Valid codes for light state
     are D01, D02, and D03.

       D01 - Light on for next move.

       D02 - Light off for next move.

       D03 - Flash (Light On, Light Off) after move (effect is limited
       to block in which appears, ie non-modal).  You can also think
       of a D03 as D02, D01, D02 series of commands linked together.

     D codes with values of 10 or greater represent the aperture's
     position on the list or wheel.  It is very important to
     understand that there is no universal "D10" or "D30".  Unlike the
     D01 , D02, and D03 counterparts which have a fixed meaning (draw,
     move, flash), D10 and higher values have aperture shapes and
     dimensions assigned to them by each individual user.  Hence, one
     job's D10 could be a 10 mil Round, when another job's D10 could
     be a 45 mil Square.

     There are two distinct ways to number an aperture list.  The
     traditional 24 aperture system started with D10 - D19, jumping
     suddenly to D70 - D71, then back to D20 - D29, ending with D72 -



     D73.  This is still a common format for output for CAD packages,
     and is still mandatory for old 24 aperture Gerber vector
     photoplotters.

     It is now common to start with D10, then increase numerically in
     steps of 1 (D10, D11, etc.) continuing up to D70 and beyond,
     rarely beyond 1000 individual apertures.

     X & Y Codes

     The X & Y values in the Gerber file determine where the aperture
     shape and dimension will be positioned and drawn.  X & Y values
     are used as coordinate pairs to determine where the light will be
     exposed, using the D codes shapes (i.e. D10) and light exposure
     status (i.e. D01, D02, D03) for drawing lines and arcs, as well
     as moving between drawing entities.

     Here are a few examples of using X & Y codes with D codes.

       D10*  { Select aperture D10}

       X1000Y1000D02* { The D02 tells us that the light will be off,
       and we move to coordinate position X1000 and Y1000}

       X2000Y3000D01* { The D01 tells us that we will draw (light on)
       to coordinate position X2000 and Y3000}

       X5500Y100D03* { The D03 tells us to move to coordinate position
       X5500 and Y100 with the light off, then flash (turn the light
       on and off)}

     G Codes

     G codes are used to configure the photoplotter.  Commonly
     implemented codes include:



       G01 - Future X,Y commands are straightline moves

       G02 - Future X,Y commands are clockwise arcs

       G03 - Future X,Y commands are counterclockwise arcs

       G04 - Ignore the rest of this block (used for Comments)

       G54 - Prepare to change apertures

       G74 - Future arcs are quadrant arcs

       G75 - Future arcs are Full 360 arcs

       G90 - Absolute data

       G91 - Incremental data



     Typically for laser photoplotters, G54 codes are rarely
     necessary.  Older vector plotter controllers may require this
     preparatory G codes for changing apertures (i.e. G54D10*).

     A common situation where G codes are mandatory for all machines
     is when the data is switching from vectors to arcs and vice
     versa.  When switching from drawing vectors (G01) to drawing arc
     (G02, G03), the controller must be informed of the change of
     mode.

     Another important case for G codes is when determining if the arc
     is a quadrant (G74) or Full 360 (G75).  Quadrant arcs never cross
     quadrant boundaries, because the center coordinate offsets (I,J
     Codes) are always unsigned (even if they are negative!).
     Therefore, it requires at least four G74 arcs to draw one
     complete circle.

     Center coordinate offsets for 360 arcs (G75) can be positive or
     negative, allowing for a single command to draw a complete
     circle.

     In either case, the center coordinates are given relative to the
     start point of the arc.  The most dramatic difference between
     Quadrant and Full 360 arcs is that a Quadrant arc with identical
     start and end points has a sweep of 0 degrees, whereas a similar
     Full 360 arc is a full circle.

     The G90 code tells the machine controller that all data following
     is absolute data.  Hence, if following X & Y data follows, the
     controller will move to the absolute value given by the X & Y
     value.



     G91 tells the machine controller that all data following is
     incremental data.  The machine will move the data by the amount
     of the X & Y value, rather than to the absolute coordinate point.

     Example:

       X1000Y1000D02*

       X3000Y3000D01*



     In absolute mode (G90), the machine will first move to coordinate
     point X1000 and Y1000 with the light off, then draw a line to
     coordinate point X3000 and Y3000 with the light on.

     In incremental mode (G91)  the machine will first move to
     coordinate point X1000 and Y1000 with the light off, then draw a
     line to coordinate point X4000 and Y4000 with the light on.  This
     was done by adding X1000 + X3000 = X4000 and Y1000 + Y3000 =
     Y4000.

     Here are some more examples of G code usage in conjunction with
     X, Y, and D code values:

       G54D10*      { Prepare to change aperture position (G54), then
       select aperture D10}

       G01X1000Y1000D02*      { Prepare to draw a vector (G01) then
       turn off the light (D02) and move to coordinate position X1000
       and Y1000}

       G90*    { This block (command) and all future commands will be
       absolute data}

       X2000Y3000D01*    Turn the light on (D01) and move to absolute
       coordinate position X2000 and Y3000}

       G91*              { The G91 command tells the controller that
       this command and all future commands that the data is
       incremental}

       X5500Y100D03*     { Turn the light off and move incrementally
       by a value of X5500 and Y100, then flash (D03) (light on and
       off)}

     M Codes

     M codes are used for machine control.  Here are the most commonly
     used:



       M00 - Full machine stop.  Commonly ignored by many plotters.

       M01 - Temporary machine stop.  Commonly ignored by many
       plotters.

       M02 - End of Plot.

     I & J Codes

     When you encounter an I & J code in a Gerber block, you have
     found an arc command.  Arc commands come in two flavors, Full 360
     or Quadrant.  The Gerber arc command is very complicated, and
     this section will only briefly describe usage of the Gerber arc.

     Full 360 arcs allow the plotter to draw a full circle (360
     degrees of arc) in one single command..

     Quadrant arcs only allow for an arc to be drawn through a maximum
     of 90 degrees of arc, never crossing a quadrant boundary.  Due to
     this restriction, I and J arc center offset codes can get away
     with never having a negative value, even if the offsets are
     negative!

     When in a Full 360 arc (G75), only one command is required to
     draw a circle.  In Quadrant mode, the same circle would require
     at leaset 4 Quadrant arcs (G74), because a circle goes through
     all four quadrants.

     Quadrant arcs will always have positively signed I and J values,
     even if the center offset is actually negative.  Full 360 arc
     center offsets can be signed positively or negatively.  A
     negative I or J is a sure indicator of Full 360 arcs.

     Modality

     It is often the case with Gerber data that when moving from one
     XY coordinate point to another XY coordinate point, the X or Y
     value will not change.  Likewise, it is likely that if the
     plotter is drawing a line with multiple segments, the segments
     will be connected and the light stays on from segment to segment.

     In both of these cases, there are redundant commands, making the
     plot data file larger than necessary.

     RS-274D allows you to omit this redundant data.  This example
     shows a box being drawn with four corners.



       Non Modal Data         Modal Data

       X0000Y0000D02*         X1000Y1000D02*

       X0000Y1000D01*         Y1000D01*

       X1000Y1000D01*         X1000*

       X1000Y0000D01*         Y0000*

       X0000Y0000D01*         X0000*

     From this example, a large amount of data has not been written,
     thus reducing the final data file size.

     Establishing the Decimal Point

     A numerical value in RS-274D data has an integer and a decimal
     part, but the decimal point ('.') is not a valid RS-274D
     character.  Thus, decimal values are written as a string of
     integers.  The implicit position of the decimal point is
     determined by three parameters:

       Number of integer digits (whole digits)
       Number of decimal digits (precision)
       Zero suppression.

     For example: In a system with integer digits=n and decimal
     digits=m (an "n,m" system), a numerical value is written using
     (n+m) digits.

     For example, in a "2,3" format the value 12.345 is written
     "12345".  In a "2,4" format, the same value is written "123450".

     Zero suppression comes in three flavors - leading, trailing and
     none.  The idea of zero suppression is to reduce data file sizes
     by eliminating unneeded 0 characters.  The simplest and most
     common form of zero suppression is leading zero suppression.

     In a "2,4" format, with no zero suppression, the value 0.0100
     would be 00 + 01000, written as "000100", but with leading zero
     suppression the same value is written as "100".

     With trailing zero suppression the same value 0.0100 would be
     written as "0001".

     How To Describe Data Formats

     Gerber data and other XY languages use a standard method for
     describing the data format.  Two examples include:



       "2,3 leading inch" or "3,3 trailing metric".  The first number
       specifies the whole digits used.  The second parameter states
       the precision.  "Leading" and "trailing" pertain to the zero
       suppression.  And the last part of the description indicates
       the units.  Refer to the above sections if these concepts seem
       unclear.