RE: TI-H: Clocking the link port
[Prev][Next][Index][Thread]
RE: TI-H: Clocking the link port
Yes it's true that the calc uses reverse logic for linking (0v = 1, 5v = 0).
This is basically how the TI-TI protocol works:
(active denotes a voltage of 0v ("on") )
Sending calc | Receiving calc
--------------------------+---------------------------
| Waits until a wire is active,
| if both are active, jumps to
| Error Handler
|
Sends the first bit: |
1 : Red wire is active, |
White is inactive |
0 : White wire is active, |
Red wire is inactive |
|
| Activates the wire that is
| inactive and records the
| bit it received (both wires
| are now active)
|
Waits until both wires |
are active, then makes |
both inactive (but one is |
still active from the |
receiving calc) |
|
| Waits for only one to be
| active (the one it
| activated), then makes both
| inactive.
|
Waits until both wires |
are inactive, then |
repeats 7 more times (for |
a byte) |
Everytime either calc waits, it is in a loop. Usually it checks for the stop
condition twice per loop. Everytime the loop completes, it decrements de
(which is initially set to $FFFF). When de is 0, it jumps to the Error
Handler (this is a timeout).
At 09:57 PM 9/2/97 -0800, Jan Zumwalt wrote:
>Let me add another question too. I heard that the calc uses Ov for a mark
>and 5v for a zero. Is this true?
>
>>The calc to calc protocol is asynchronous, as are all the homemade
>>link
>>cables. The official GraphLink converts the TI protocol to 9600 baud.
>>--
>>Brian Leech
>>butvis@mindspring.com
>>ICQ UIN: 1355611
>
>Okay, so now I'm lost. How do the calcs know when a bit is ready to be
>read from the link. Maybe I should start at the begining, can someone
>explain how linking works from a hardware perspective?
--
Brian Leech
butvis@mindspring.com
ICQ UIN: 1355611
References: