Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Parallel and Serial Ports

1 view
Skip to first unread message

bork...@zodiac.rutgers.edu

unread,
Apr 6, 1992, 2:47:02 PM4/6/92
to

I was recently sitting in the library reading 'Computer', the
magazine by the IEEE Computer Society, and in the back they had a preview
of a battery operated, portable hard drive called Flash 25. Further on the
article stated that the arddrive gets connected to the parallel port of
the computer.

I don't pretend to have mastery of computers (yet!), but this
struck me as strange because I always thought the parallel port was a one
way port. (If you remember, there were quite a few messages on this
newsgroup a while back about the virus in Iraq's defense computers, and if
it was possible to send a virus backward through a parallel printer port.
I believe the general concensus was no, that it was not possible.) I also
looked at a pin-out chart of my parallel port and (assuming the parallel
port is a standard) there were 8 data out pins and a bunch of not connected
pins. Is the parallel port non-standard, or has it changed recently? If not,
how is such a hard drive possible?

And while I'm typing this, another question/personal problem comes
to mind. When not studying and such, I work part time as a bartender at a
restaurant that has a restaurant management computer system, which takes
orders and sends food orders to the kitchen and drink orders to the bar.
The computer system also will print out a report of sales at the end of the
day to either the printer or any of the terminals. However, I was asked if
it is possible to have the report sent over a modem instead.

My basic problem is thus: The system is a completely closed system.
There is no keyboard input, making it impossible to do anything except run
the restaurant program. The host computer communicates to the seperate
order terminals using 19200bps with an RS-232 port. From what I know from
the system, it would be no problem to connect a modem to an extra port,
define it as another terminal, and call from a remote location to have
access to the system. However, if the modem is 2400bps, and there is no
way to slow down the data stream from the computer, is there anyway this
would be possible? (One idea I thought I might look into is using a serial
buffer to store the output from the computer at 19200bps and send it over
the modem at 2400bps.)

Jason
(BORK...@ZODIAC.RUTGERS.EDU)

Michael A. Covington

unread,
Apr 6, 1992, 8:47:07 PM4/6/92
to
Bidirectional (input/output) parallel ports are found on many laptops
and most if not all IBM PS/2's.

Even a unidirectional parallel port has several "printer status" lines
that can be used as input.

Parallel ports are much faster than serial ports, which is why the
external disk drive connects to the parallel port rather than the
serial port.

--
==========================================================================
Michael A. Covington, Ph.D. | mcov...@uga.cc.uga.edu | ham radio N4TMI
Artificial Intelligence Programs | U of Georgia | Athens, GA 30602 U.S.A.
==========================================================================

Bill Mayhew

unread,
Apr 6, 1992, 11:29:46 PM4/6/92
to
Four of the leads on an IBM-PC compatible printer port are
bi-directional in that you can program them as an output then
sample the level as an input to see if the signal really changed.
Products such as Laplink send data in sequences of 4-bit nybbles
taht way. Ethernet ports and portable HDUs do that too.

For question two, newer v.32 and some 2400 baud modems do
buffering and baud rate conversion. My computer at home has a GVC
v.32 modem, typical of many in the $350-400 price range. I run
data into the modem at 38.4Kbuad regardless of what is on the other
end. The modem can mediate the data flow by sending either
xon/xoff characters or using the RTS/CTS leads. Obviously, the
terminals in the bar use some sort of flow mediation. I've seen
very few terminals that can keep up with a 19.2K data stream for
more than a couple hundred characters.

--
Bill Mayhew NEOUCOM Computer Services Department
Rootstown, OH 44272-9995 USA phone: 216-325-2511
w...@uhura.neoucom.edu (140.220.1.1)

Sakari Aaltonen

unread,
Apr 7, 1992, 12:39:19 PM4/7/92
to
The ad you saw must have referred to the parallel (Centronics) port of a PC.
It is normally used for interfacing to a printer, i.e., it is NORMALLY
one-way. However, the parallel ports of most PC's are actually two-way or
bidirectional; the direction can be changed with a switch or even software.
That is why many sorts of devices can be and are connected to the parallel
port: hard disks, floppy disks, LAN adapters etc.

--
-----------------------------------------------------------------------------
Sakari Aaltonen Helsinki University of Technology
Email: sak...@cs.hut.fi
--- You can't keep a Finn down without him getting all red in the face ----

Michael Schuster

unread,
Apr 9, 1992, 6:38:16 PM4/9/92
to
In article <1992Apr7.0...@athena.cs.uga.edu> mcov...@athena.cs.uga.edu (Michael A. Covington) writes:
>
>Parallel ports are much faster than serial ports, which is why the
>external disk drive connects to the parallel port rather than the
>serial port.
>
I've noticed on the specs of the Trantor parallel port-to-SCSI adaptor that
writes (i.e. output from the parallel port) are twice as fast as reads
("input" via the parallel port). Does this mean the "input" is kludged?

--
_______________________________Mike Schuster________________________________
NY Public Access UNIX: schu...@panix.com | 70346...@CompuServe.COM
The Portal (R) System: schu...@cup.portal.com | MCI Mail,GEnie: MSCHUSTER

Adrian Godwin

unread,
Apr 10, 1992, 9:15:55 AM4/10/92
to
In article <1992Apr9.2...@panix.com> schu...@panix.com (Michael Schuster) writes:
>In article <1992Apr7.0...@athena.cs.uga.edu> mcov...@athena.cs.uga.edu (Michael A. Covington) writes:
>>
>>Parallel ports are much faster than serial ports, which is why the
>>external disk drive connects to the parallel port rather than the
>>serial port.
>>
>I've noticed on the specs of the Trantor parallel port-to-SCSI adaptor that
>writes (i.e. output from the parallel port) are twice as fast as reads
>("input" via the parallel port). Does this mean the "input" is kludged?
>

It means that reads use the 4 control-input bits as inputs, taking 2 operations
per byte. Writes use the 8 output bits in one go and are therefore twice as fast.
The remaining control-output bits are used to control the data transfer.

-adrian

f
i
l
l
e
r

--
--------------------------------------------------------------------------
Adrian Godwin (ago...@acorn.co.uk)

Jim Pieronek

unread,
Apr 10, 1992, 7:02:24 AM4/10/92
to
In article <1992Apr9.2...@panix.com> schu...@panix.com
(Michael Schuster) writes:

I've noticed on the specs of the Trantor parallel port-to-SCSI adaptor that
writes (i.e. output from the parallel port) are twice as fast as reads
("input" via the parallel port). Does this mean the "input" is kludged?

In a word, yes. If you eyeball the pinout for the printer port pinout
you will notice that there are 8 dedicated output lines, (DATA 0-7)
and 4 dedicated input lines (BUSY, PAPER OUT, etc.) So your SCSI
gizmo must be using the 8 output lines for writing and the 4 input
lines for reading. Many computers, but not all, can use the 8 data
lines bidirectionally. Some of software for the devices made for
printer ports test this when they are initialized and use it that way
if possible, otherwise they fall back to the 8 out and 4 in
configuration.

- Jim
--
J.V. Pieronek Internet: j...@ll.mit.edu
Massachusetts Institute of Technology (617) 981-3468
Lincoln Laboratory These are my opinions, not the Lab's.

0 new messages