Gavin wrote:

> Of course it's ridiculous that one should not be able to duplicate a
> simple serial cable, especially given a working example to start from,
> but the ways of serial communications on HP3000s have always involved a
> degree of black magic and at some point it's just not worth the effort
> and one is forced to give up and move on.

I was fully prepared to do just that. In fact, it was while I was looking
for pricing on the Internet for the aforementioned 40233A cable that I
stumbled across the following at
<http://hilbert.math.hr/arhive/hpux_admin/1996/0115.html> (near the bottom
of the page - extraneous stuff snipped):

<snip>

     APPENDIX A - CABLE PINOUT DIAGRAMS

     (* indicates lines typically required by modems)

                40233A
                25M-25M
       CPU               Modem
     (25-pin)          (25-pin)
     ---------------------------
     GND  1 ---------- 1  GND *
      TD  2 <--------- 3  RD  *
      RD  3 ---------> 2  TD  *
     RTS  4 <--------- 8  CD  *
     DSR  6 ---------> 20 DTR *
     GND  7 <--------> 7  GND *
      CD  8 ---------> 4  RTS
          9 <--------- 22 RI
     DTR 20 <--------- 6  DSR
      RI 22 <--------- 5  CTS

<snip/>

I realized looking at it that there was a difference between this diagram
and the documentation in the HP manual, namely a *connection* from pin 9
to pin 22. The manual showed this:

<snip>

 8 - RTS     OUTPUT    -------->   4 - RTS     INPUT
 9 - RI      INPUT                22 - RI      OUTPUT
20 - DSR     INPUT     <--------   6 - DSR     OUTPUT

<snip/>

and *didn't show a line between 9 and 22 to indicate a connection. Hmmm...

So we took a piece of wire, stuck it in pin 9 on the DTC side and pin 22
on the modem side, plugged in the cable, dialed... and *voila!* A system
prompt!

Wirt writ:

> The missing pins would have absolutely no effect on the behavior of
> the cable. The problem has to lie wholly in the wiring (or the quality
> of the wiring job).

You're mostly right, Wirt (as usual :). The problem *was* in the wiring,
but it was the quality of the *documentation* that was at fault.

Thanks all for your help (and offers of help)!

Patrick     "Sometimes surrender leads to victory."
--
Patrick Santucci
Technical Services Analyst
Seabury & Smith, Inc.