On 21-Sep-99 Gerhard Kroder wrote:
recently i had to configure a cisco isdn router, wich requieres 2
stopbits (at eight data and null parity bits) as communication
parameters. the only communication programm avaliable on my 80mb
deb2.1 was minicom, which didn't have 8N2 to set up. do you know
others that support 8N2, or how else can i set these parameters? i do
not want to switch to an windows evironment, where hyperterm supports
8N2
Are you finding that the comms fail if you set it at 8N1?
Usually, serial comms are robust about the number of stop bits.
In voltage terms, the line alternates between a high (MARK, M)
level and a low (SPACE, S) level.
Theoretically, the transition from one to the other is instantaneous.
In practice, it takes a little time so there is a gradient followed
by a roughly constant M or S voltage.
The ground (quiescent) state is MARK. The start of a signal packet
is signalled by a drop to SPACE level for the duration of a bit cycle.
When this transition is detected, the receiver starts its clock to
measure out bit-cycle intervals, confirming by measuring the level
(should be SPACE) after half a bit-cycle. This is the Start bit,
whose basic function is to synchronise the receiver's bit clock.
Head to tail thereafter (no theoretical time gap between them) follow
7 or 8 data bits and an optional parity bit, each at either MARK or
SPACE level, and one or more Stop bits at the MARK level. Levels are
measured at around the timed midpoints of the bit intervals, so as to
keep clear of the gradients.
Numbers of data and parity bits are determined by the comms settings,
and the receiving device simply counts bit-cycles until it has received
what it expects.
Once the data and any parity bits have been received and the first
Stop bit has been verified at MARK level, the next interesting event
will be the transition from MARK to SPACE for the Start bit of the next
packet. Therefore a second Stop bit is redundant, and as a general rule
serial comms devices will ignore it.
This is the explanation why serial comms are usually robust about the
number of stop bits.
And also why I suggest (unless you have already tried it and it doesn't
work) that you try 8N1 anyway. It is possible, however, that if a device
has been set up to insist on 2 Stop bits, then it won't start looking
for a Start bit until it has clocked at least 2 bit-cycles after
the expected time of the beginning of the first Stop bit cycle.
Hope this helps,
Ted.
E-Mail: (Ted Harding) [EMAIL PROTECTED]
Date: 21-Sep-99 Time: 13:17:39
-- XFMail --