Re: dummy terminal addresses

Sounds like you might have found the source for your problem - not sure if
the power outage is related in anyway but until you fix that all other
discussions might be moot.  Regarding the controller I was just wondering
about whether it acts like a remote controller or a local controller.  If
it acts like a remote controller the type of IBM controller it emulates
might be important.  However, based on your description sounds like it you
are dealing with a local controller (CTLxx, etc.).

Twinax is definitely risky as far as one problem causing the whole port to
go wrong.  With the cursor in the upper right you know that the SA is
probably not working.  To help get an understanding of what you have it
would help to know the addresses of each terminal.  Although I think fixing
your one end will go a long way to fixing your problem.  Unfortunately I
haven't dealt with TWX cable for more years than I care to admit....

So, I'd say, get that end fix and see what happens.




midrange-l-bounces+mike.crump=saint-gobain.com@xxxxxxxxxxxx wrote on
02/14/2006 11:32:49 AM:

> Mike and Mike (yeah, THATS not confusing!),
>
> answers
>
> <snip>
> 1.) What type of controller are you attached to....<snip>
>
> a 2x4 connection controller wired right into the back of my system,
> not sure of what it is called.  It is not a Perle box or anything.
> Just a standard hard connection.
>
> <snip>2.) What type of wiring is in use - twinax, twisted pair,
> twisted pair with
> a hub, etc.  In the twinax world it is possible to have the daisy chain
> affect.....twisted pair world less likely.<snip>
>
> Twinax
>
> <snip>3.) With the terminals how are they acting?  I'm not that familiar
with
> 3197's but you need to look for a System Availability (SA) indicator and
> also make sure you know where the cursor is (upper right, or upper
> left).<snip>
>
> not sure if SA indicator is on (bottom left on the operating status
> line is where it is supposed to be *see below).  cursor is upper right.
>
> <snip>The controller will use a hex addressing scheme that will
> correspond to a
> port number and a address number.  Starting with 00 (port 0, address 0)
and
> moving down the port...01, 02, 03, 04, etc.<snip>
>
> One of my first actions was to try to check the terminals to make
> sure they did not lose their addresses when the power blipped.   I
> can view on the system how the defined devices are addressed.  My
> problem is that i have 15 defined, only 4 machines.  I cannot tell
> by looking at the terminal itself which port it is defined for.
>
> <snip>It is possible that you have a problem with the port
itself.....rare but
> I've had them get zapped on occasion.<snip>
>
> If daisy chained and the port for the first machine is bad, that
> would make the rest go down?  how do i test a port?
>
> <snip>Sounds like the workstation controller may need re-setting.
> Do a WRKCFGSTS *CTL CTL*
> (assumes standard naming convention !) & see if any need
> Vary-off then Vary-on..<snip>
>
> have done this several times, nothing.
>
> i just got back from the warehouse, i was going down to fiddle with
> the tester i have.  I started at the point where the wires first
> come out of the wall to the first Workstation(the one they removed).
> Im assuming that green is good, but i did not get a green light from
> the cable marked inbound (or outbound, just to say i tested that
> too).  Do i need to have both ends of the tester connected?  i have
> no idea what it is testing... a connection to the 400?  or just a
> complete circuit tester?  tester does nothing at all on setting 1 or
> 2, but i got 2 amber lights on all tests on setting 0.
>
> As i was pulling the tester off, i noticed that one of the two pins
> was pushed in on the line marked inbound.  removing the screw casing
> revealed the pins soldering had worn and they fell to the ground.
> Problem is, im not sure if i did this while testing??? ugh, so im
> going to repair the connector.
>