[net.dcom] Wanted: info on EIA RS-232

newrace@hogpc.UUCP (K.LUKAS) (10/20/83)

Wanted: Info on minimum EIA RS 232-C leads needed on host
end of link. Specifically - is RING INDICATER needed on that 
interface? What are the consequences of it not being supported?
Please mail responses to me - I'll summarize to the net.
Thanks in advance.

karel lukas	{ inhp4!houxm!  or  pegasus! } hogpc!newrace
					
					     or hogpc!klio

smith@umn-cs.UUCP (10/23/83)

#R:hogpc:-27000:umn-cs:1600002:000:1026
umn-cs!smith    Oct 21 13:01:00 1983

The 'absolute minimum RS-232C' I've ever seen is three wires: transmit,
receive, and signal ground.  Will this work for you?  No way to tell without
knowing what you're hooking together.  Here are observations based on
my own experiences:

1. Everyone expects at least transmit, receive, and signal ground.

2. I've seen few systems that really require a RING signal.

3. Many (but not all) connections use Data Terminal Ready (DTR) or
   Data Set Ready (DSR) to tell if there's something connected.  Dropping
   one of these (depending on your direction) "disables" the connection.

4. Carrier Detect is usually important if you're connecting to a modem.

5. Some people use Request to Send/Clear to Send (RTS/CTS) instead of or
   in addition to DTR/DSR.

6. Any of these signals can usually be strapped permanently 'on' when
   you build the interconnecting cables.  This takes away your control
   but can make otherwise impossible things work.

Rick Smith.
U. Minnesota
[smith.umn-cs@Rand-Relay]
 [...ihnp4!umn-cs!smith]

ptw@vaxine.UUCP (P. T. Withington) (10/25/83)

This is the closest thing I've ever seen to a "universal" RS-232 cable:

   (Computer End)                                (Terminal End)
Shield  1-----------------------------------------------1
XMT     2---------------------\ /-----------------------2
			       X <=======================(makes a null modem)
RCV     3---------------------/ \-----------------------3
RTS     4----+                                     +----4
	     |                                     |
CTS     5----+                                     +----5

DSR     6----+                                     +----6
GND     7----)-------------------------------------)----7
CD      8----o----------------\ /..................o----8
			       X
DTR    20...................../ \----------------------20

If you leave out the "..." line, short 8 to 20 at the terminal end (losing the
the ability of your terminal to tell if your computer is there, which few do
anyways).  This cable can thus be implemented with a four-wire shielded wire,
or phone wire if you don't believe in shielding (which is sometimes a loss).

Someone out there is marketing an "automatic" black box, which purports to
instantly connect two RS-232 devices.  I've ordered one, but haven't seen it,
and don't remember off hand who it is.

			     't`   --Tucker (ptw@vaxine.UUCP)
			      ~

mason@utcsrgv.UUCP (Dave Mason) (10/26/83)

The absolute minimum number of leads is 2!  I have a friend with a
timesharing company, and he uses 2 wire cables for the VT100s around
the office: pins 2 & 3 and depends on the floating ground from the
power supplies to be within 3 volts or so (a safe assumption within
one building).  Never ceases to amaze me that it works fine.
 -- Dave Mason, U. Toronto CSRG,
        {cornell,watmath,ihnp4,floyd,allegra,utzoo,uw-beaver}!utcsrgv!mason
     or {decvax,linus,research}!utzoo!utcsrgv!mason   (UUCP)

dag@fluke.UUCP (David Gunderson) (10/31/83)

Be careful about floating ground systems. A fellow I know connected up an
RS-232 system like that in a building remote from the one the computer was in.
He used a direct wire connection. It turned out that the outlet his terminal
was plugged into had the ground and common reversed causing the signal lines
of the RS-232 line to be raised to about 110 volts. It caused MUCH damage in
the serial interface section of the PDP-11 that it was connected (smoke and
flame!!!).

--David Gunderson

karn@eagle.UUCP (Phil Karn) (11/02/83)

For those who need to handle dissimilar grounds between terminal and
computer, I noticed that Inmac (that company that must spend half its
profits on sending me catalogs every week) is now selling RS-232 fiber
links.  You might consider this to be a rather stretched opto-isolator.

This isn't an endorsement of Inmac; theirs just happened to be the first
ad I've seen for this kind of product.

Phil

david@varian.UUCP (11/03/83)

The  "automatic  black box" to connect 2 RS232  devices  referred  to by
vaxine!ptw is called the SmartCable from IQ Technologies, 1181 N.E.  1st
St,  Bellevue  Wa.  98005  (206)  451-0232  (that's a cute phone # :-)).
Call them to find your local distributor.

We received one a few weeks and found that it really works in most cases
(we found a few exceptions involving HP equipment that I don't remember
off hand).  I haven't taken it apart yet to examine the innards, but
it seems to work by overkill - it turns on all possible control signals
in the hope that it will hit the ones you need. It is also capable
of being a null modem (reversing pins 2 and 3) and seems to work fine
as such.  It has already saved me quite a bit of time in making quick
connections, where in the past I would have had to play with a breakout
box and jumpers.

The cost of the logic module (SC817) is $62 (there is another version
called SC817-D that contains one extra LED for debugging - it's about
$15 more and I would have ordered that instead if I had known) and consists
of a male 25 pin plug, ribbon cable connecter, 2 LED's, and a switch
(you flip the switch til both LED's are lit). For another $15 I got
the 257-Dual cable (ribbon connector to both male and female 25 pin
plugs).

	David Brown
	Varian Instruments 2700 Mitchell Dr.  Walnut Creek, Ca. 94598
	(415) 945-2199
	{ihnp4,tektronix,sytek,dual}!zehntel!varian!david
	{amd70,fortune}!varian!david
	...!decvax!sytek!zehntel!varian!david
	...!ucbvax!menlo70!sytek!zehntel!varian!david

smith@umn-cs.UUCP (11/05/83)

#R:utcsrgv:-257700:umn-cs:1600003:000:450
umn-cs!smith    Oct 29 10:40:00 1983

Concerning 'floating grounds':

  I remember a story from a few years back: a NY bank was installing a vast
number of terminals and associated datacom hardware.  Nothing worked.  Finally
a smart EE checked out the building power and discovered that the typical
office area had power coming in from two DIFFERENT mains and there was a
significant voltage drop across the alleged 'grounds'....

Rick.
[smith.umn-cs@Rand-Relay]
 [...ihnp4!umn-cs!smith]