VaxServer 3100 and VT340 terminal

From: Chuck McManis <cmcmanis_at_mcmanis.com>
Date: Fri Mar 30 16:51:36 2001

At 04:41 PM 3/30/2001 -0500, Chad wrote:
>I get the vt320 Ok message, but nothing past a blinking cursor after
>that.

And you've confirmed that with local echo on you can type and the terminal
will display what you type right? Good.

>What do the numbers on the console
>mean? It stalls at 3, then goes to 7, C, and then stops at 1.

They represent what the status register has in it, stopping at '1' means it
is trying to boot. You want it to stop at 3. (which is in the console
monitor). So look at the position of the rotary switch and slide switch on
the console panel. Rotary switch points to the arrow, slide switch goes
down to the side indicated with the 'dot-out-of-the-hole.'

>How do I figure out what is broken?

Work from one end out to the other. Assuming you've done all the steps in
the previous message and the follow things are true to the best of your
knowledge:
                 1) The 3400 is set to 9600 baud console speed.
                 2) On the console bulkhead the round switch (has a raised
triangle on it)
                    is set to point at the "arrow" glyph, and the slide
switch next to it
                    on the left is in the lower position
(dot-out-of-the-circle).
                 3) The VT320 is connected from the MMJ on the bulkhead to
the MMJ
                    with back to back arrows (second MMJ from the left
looking at the
                    back of the VT320)
                 4) The VT320 is set to talk 9600 baud (transmit and
receive), jump scroll
                    no flow control.
                 5) The VT320 is powered up and passes its self tests.
And it still doesn't work. Then you go into "fault finding" mode. And that
is where you start at one end and work toward the other.

A question occurs to me, did you make your own MMJ cable or is it a DEC
one? If you made it yourself did you remember to flip the wires in one end?
(All MMJ cables have a 'half-twist' which connects pins on two MMJ
connections to their opposite counterparts. This is a really cool feature
of DEC-423 ports which allows you to hook up any two pieces of gear with
the same cable and they always are wired to talk. If you made your own
cable, or had it made, this is the likely problem.

So we'll start at the VT320 end as that is the simplest to test without any
special tools.

After you've verified that your VT320 can locally echo characters then the
next step in the signal path is the serial port on the back of the
terminal. So the 6 pin connector on the MMJ has wires on one side that are
exposed. If you think of these wires as three 'pairs' with the two middle
pins as the first pair, the next two out from the center (pins 2 and 5) as
the next pair, and the outer pins as the third pair, then you want to short
the second pair. You can do this either with a very fine wire, or by
plugging the cable into a MMJ->DB25 adapter and shorting pins 2 and 3 on
that adapter. This will create a loop-back path for the terminal such that
when it is *not* in local echo mode, characters typed on the keyboard are
sent through the MMJ, down the cable, travel across your short to the other
pin, come back up the cable and back into the MMJ (now as transmitted
characters) and get displayed on your screen. If you find yourself wanting
to test a lot of DEC terminals for some reason, making an MMJ loopback
cable just for this is quite cheap and easy if you have the crimper tool.

If after the above test, you see the characters you type getting echoed on
the screen, then you know that the VT320 is working and the cable does not
have any breaks in it (you don't know if the cable is wired properly but
unfortunately.) You can test proper cable wireing with an ohmmeter or a
lightbulb/battery/two wires circuit. Check to see that pin 1 on one MMJ has
continuity with pin 6 on the other MMJ, then 2 and 5, then 3 and 4 etc.

If you want to validate the baud rate on the Vt320 you can either hook it
to a known good system (I use a laptop's serial port) or an oscilloscope.
On an oscilloscope you should type 'U' or '*' (nice pattern of 10101010 in
ascii) and measure the time between edges to verify 9600 baud timing.)

So if that works, then it gets a bit harder because it means the fault is
between the console plug on the VAX and the CPU board.

If you _do_ have an oscilloscope. Just hook your MMJ cable to the VAX and
put the scope on pins 2 and 3 (gnd) to see if you see characters being
produced. If they aren't then its quite possible that the MMJ circuit got
toasted. This happens if you have a VT220 plugged in and you unplug the
terminal from the power line rather than turning it off at the switch. It
generates a voltage spike (that looks like the current loop voltage) across
the MMJ and toasts the receiver. The fix for that problem is to replace the
differential driver chip (or the console bulk head) [Note if you or anyone
else reading this wants a 3400 console bulkhead then $10 will get you one
delivered via priority mail to any of the 48 states :-)]

At this point though my money is on the MMJ cable.
--Chuck
Received on Fri Mar 30 2001 - 16:51:36 BST

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:34:05 BST