home *** CD-ROM | disk | FTP | other *** search
- TITLE: how to un-blink the cursor...
-
- It appears as though most monochrome display cards (and graphics such as
- HERCULES and "compatibles" use the same basic layout. This is roughly how it
- works, for those who have no schematics: The CURSOR output (pin 19 of the
- 6845) goes into a 74LS174. There it is delayed since the cursor location as
- output by the controller is normally at the location of the last character.
- The cursor is delayed by one character position so that it is displayed just
- beyond the last character. This is done in two flip-flops in the '174. The
- cursor signal enters on D4, the Q4 output is fed into D5. The Q5 output is
- the delayed cursor signal. It is fed into a 7464, usually on pin 9. The '64
- basically consists of four AND gates feeding a NOR gate. Pins 9 and 10 are
- the inputs of a 2-input AND, and pin 10 is the clock pulse providing for the
- blink rate. To force this long story to an abrupt end, simply cut the trace
- leading to pin 10 of the 7464 and the cursor will quit blinking. You might
- want to do it right and pull pin 10 up to 5V via a 2.2k resistor after you
- have verified that it indeed works.
-
- There appears to be only one 7464 on the board, so you can't miss. Don't be
- confused by the board layout - the 64 may be clear on the other end of the
- board as seen from the 174! If your display card doesn't have a 7464, then
- you will ahve to do some tracing. A schematic would be very helpful at this
- point. In all the display cards I have seen thus far, the cursor signal is
- delayed in a 74LS174 using two sections of the chip. The delayed cursor
- signal then enters an AND or NAND gate at one point. If it is a 2-input
- gate, all you should have to do is to CAREFULLY cut the trace leading to the
- other input. It is important to be able to reverse this procedure in case
- you cut the wrong trace. Beware, though, since traces may pass under chips
- and emerge at a different point on the other side, or even switch circuit
- board sides. The (N)AND gate may even be located at the other end of the
- board.
-
- In tracing through a few more cards, I found an unnamed taiwanese color
- graphics card where pin 19 of the 6845 goes to one section of a 74LS273 and
- from there to a 74LS00; cutting the other input to the '00 stopped the
- blinking.
-
- Here is some more informastion, based on the IBM manuals:
- 1. MONOCHROME DISPLAY ADAPTER: cut trace going to pin 9 of U3 (LS08)
- 2. COLOR/GRAPHICS MONITOR ADAPTER: cut trace going to pin 12 of U49 (LS02)
-
- One problem with turning the cursor blink off this way is that if your
- cursor is something other than an underline (like, a block), you cannot see
- the character immediately below the cursor if you move it around the screen!
-
- You may use the following code fragment to control the cursor under
- DEBUG or include it in programs:
-
- MOV AH,1
- MOV CX,cccc
- INT 10H
-
- Here are a few values for "cccc" for you to play with:
- (INT 10, subfunction 1 is explained in detail in various manuals and books
- but there no mention is made of bits 5 and 6 of register CH.)
-
- cccc = 600BH - slow blinking block
- cccc = 400BH - fast blinking block
- cccc = 000BH - steady block
- cccc = 0B0BH - steady underline
-
- As you can see, bits 6 and 5 of the CH register control blink:
-
- BIT 6 BIT 5
- 0 0 no blink
- 0 1 no cursor display
- 1 0 fast blink
- 1 1 slow blink
-
-
- I hope this has helped some people who, like myself, are suffering from
- blinking cursors. I might add that cutting traces on circuit boards makes
- your warranty vanish in a puff of greasy black smoke. Fooling with
- electronic equipment not knowing what you're doing may make the equipment
- malfunction horribly, causing much grief and emptiness of the wallet, so you
- should only attempt to unblink your cursor if you are quite sure you know
- what you're doing... If all else fails ask a friend to do it for you. Just
- don't blame ME for breaking your computer, ok?
-
- May all your cursors be steady ones!