CX-85 KEYPAD››by Kevin Packard, WNYAUG››     Anyone who has ever used a›calculator must have noted how much›easier it is to enter numbers then with›a typewriter style keyboard. I am truly›convinced this is due to the difference›in key arrangement.  A numeric keypad›can be operated using one hand and›movement is limited to a small area. ›This arrangement is so much more›efficient that numeric keypads are›being incorporated as standard features›on the most keyboards.››     Programs have been written that›will redefine the ATARI keyboard so›that this arrangement is simulated.  I›find this approach cumbersome due to›the offset of the keys.  You end up›looking at your hand more often then›the input information.››     There is another approach.  You›can buy a CX-85 Keypad and plug it into›joystick port #2.  The keys are full›size and there is a bump on the #5 key›so you can always find your way back to›the home position.  Not only that but›you get an additional 4 function keys›(great for cursor controls).››     So why doesn't everyone have one? ›Cost must have been a big factor when›they first came out.   The CX-85 sold›for over $60 six years ago.  Also on›the box was the claim that it would›work with the Bookkeeper and VisiCalc. ›Software seemed to be limited.  And›maybe people just didn't know about›it.››     The first problem has been solved.›I've seen the keypads being sold for›less then $10.  You may  have to look›the advertisements over pretty well to›find it.  If you read this far you must›now know it exists and if you read on›you will read that software is less of›a problem then you might think.  If you›can't find it, write it.››     The hardware is quite interesting.›Under the 17 keys is a circuit board›that reads thee keypad.  When you press›a key it sets the output code for that›key and maintains (latches) that code›until another key is pCX-85 KEYPAD››by Kevin Packard, WNYAUG››     Anyone who has ever used a›calculator must have noted how much›easier it is to enter numbers then with›a typewriter style keyboard. I am truly›convinced this is due to the difference›in key arrangement.  A numeric keypad›can be operated using one hand and›movement is limited to a small area. ›This arrangement is so much more›efficient that numeric keypads are›being incorporated as standard features›on the most keyboards.››     Programs have been written that›will redefine the ATARI keyboard so›that this arrangement is simulated.  I›find this approach cumbersome due to›the offset of the keys.  You end up›looking at your hand more often then›the input information.››     There is another approach.  You›can buy a CX-85 Keypad and plug it into›joystick port #2.  The keys are full›size and there is a bump on the #5 key›so you can always find your way back to›the home position.  Not only that but›you get an additional 4 function keys›(great for cursor controls).››     So why doesn't everyone have one? ›Cost must have been a big factor when›they first came out.   The CX-85 sold›for over $60 six years ago.  Also on›the box was the claim that it would›work with the Bookkeeper and VisiCalc. ›Software seemed to be limited.  And›maybe people just didn't know about›it.››     The first problem has been solved.›I've seen the keypads being sold for›less then $10.  You may  have to look›the advertisements over pretty well to›find it.  If you read this far you must›now know it exists and if you read on›you will read that software is less of›a problem then you might think.  If you›can't find it, write it.››     The hardware is quite interesting.›Under the 17 keys is a circuit board›that reads the keypad.  When you press›a key it sets the output code for that›key and maintains (latches) that code›until another key is pressed.  Another›bit (or control line) is set when the›key is in the down position.››     The CX-85 connects to the computer›through a joystick port.  Which port›can be determined by the program.  The›four joystick pins and one of the›paddle ports is used for the key code. ›The trigger pin is set only when a key›is pressed.››     Two programs come with the CX-85, ›both utilise page 6 and operate during›the VBI (vertical blank interrupt). ›During the VBI the trigger is checked›to see if a key is pressed.  If the›trigger is not set the VBI routine is›exited. When it senses that the trigger›is set it then reads the 4 joystick›pins and paddle port. The paddle port›condition is added to the joystick bits›in such a way as to generate a unique›number for that key. The number of that›key is then used as an offset into a›table that contains the ATASCII code›for the character you wish the key to›represent.  The character code is then›placed to the keyboard location and the›VBI is exited. From there on the›computer acts as if a normal key has›been pressed.››     In the VBI routine there is an›auto-repeat function and a slight delay›time before reading the paddle port. ›The paddle port contains an RC circuit›that does not react to changes as›quickly as the joystick ports.  The›delay time is needed to allow the›paddle port to reach its on/off state. ›It is for this reason that the keypad›inputs need to be latched.  Once set›the keypad will not change the code›half way through the delay timing›routine.  This delay is not long in›human terms but to the computer these›milliseconds are important.››     The only difference between the›two programs is the way they are loaded›and function keys are defined different›characters.  One file is a boot›program. This program has no file name›but will boot if it is in the disk›drive when you turn the computer on. ›The routine is loaded and then you are›asked to change disks.  The program is›set up to use with VisiCalc (a spread›sheet program).  The function keys are›set up to be used like the cursor›control keys.  I have used it with›VisiCalc and Calc Magic and it makes›data entry so much easier.››     The other program is an autorun›file that will load it like any autorun›file. Once installed you can then use›it with other files.  I've used it›mostly in BASIC but I have had it›running with some machine language›programs (Speed Script is one).››     The most convenient aspect of›using the keypad with BASIC is that you›can redefine the keys under program›control. (It is possible in assembly›language but have never tried it.)  The›memory location for the keys are in the›user's guide.  The locations are not in›sequencial order so you will need a›list of the locations.  To refine the›#1 key,  POKE 1593 with the ATASCII›code you want inputted when the key is›pressed.  ››     As an example, let's say you want to draw borders around a menu display›using a screen editor.  You could›search your keyboard for the characters›for the corners and sides.  Or use the›keypad by redefining the keys so that›#7 is the upper-left hand corner, #8 is›a horizontal line, #9 is the›upper-right hand corner, etc. Now since›the keypad is in the shape of a square›it is easy to remember which key to›press for what portion of the border›you want.  If the 4 function keys are›set for cursor control, you won't have›to touch the regular keyboard to›complete the border.››     The disk also contains the source›code of the keypad handler.  The›assembly code is written using the›Atari Macro Assembler.  It is possible›to change the table and compile the new›program.  If this is not a viable way›then there is an alternative.  In the›September 1986 issue of Computer›Shopper is a program titled "CX-85›Numeric Keypad Driver Creator" by Jeff›Brenner.  This is a BASIC program that›will allow you to redefine the table›and then automatically write an›AUTORUN.SYS file to a disk.  With this›program you need to press a key on the›keypad and then press the key on the›keyboard that you want in the table. ›When you are finished press START and›the modified autorun file will be›saved.  The program does not have the›autorepeat routine, but this is a small›sacrifice for the ease of use.››     Many times while entering BASIC›programs you come across a group of›DATA statements with numeric data›(character set, machine code). The›following program changes the table to›make entry of this data easy.  It›redefines the first three function keys›as "D", "A", and "T".  The forth›function key is defined as the back›space key for mistakes, and the "." key›is changed to a ",".››10 REM CHANGE CX85 HANDLER FOR DATA›INPUT›20 POKE 1615,12›30 POKE 1611,32›40 POKE 1583,58›50 POKE 1585,63›60 POKE 1587,45›70 POKE 1589,52›100 NEW››     After the keypad handler is loaded›and you have run this program DATA›lines can be entered without touching the keyboard.  Simply type the line›number and use the function keys to›spell DATA.  After this you can start›entering in the numbers.  Press the›decimal point key for the comma to›separate numbers.  Since Atari BASIC›will place the spaces where the  are›needed there is no need to type them. ›They will be there when you list the›program. ››     The keypad can do all this and›more.  There should not be a an›application program written that›doesn't use this keypad's convenience. ›Maybe if more people owned a CX-85 more›programs would claim "CX-85›compatible".  But then if more programs›claimed this there would be more of a›reason to buy one.›