On Sun, 16 Jul 2000 lpt1@cyber-wizard.com wrote:

> Well, I've got a MicroVax 3300, and I've had success using the 
> 'differential' over cat5 cabling to RS-232.
> 
> I have a pinout for the vax of:
> 
> 1 -> Ready Out
> 2 -> TX+
> 3 -> TX-
> 4 -> RX-
> 5 -> RX+
> 6 -> Ready In

This is a DEC MMJ I assume. :-)

> Normally, you tie the RX- and TX- together to make a ground. What I did is 
> use 3 pair of a Cat5 cable to make the link as follows
> 
> Vax             RJ-45           DB-25
> 
> 1               Orange          20
> 2               Green           2
> 3               green/white     7
> 4               blue/white      7
> 5               Blue            3
> 6               orange/white    6&8
> 
> Pin 7 on the DB-25 is the common tie point for RX-/TX-.

Yes. And this makes the connection *not* differential. Just plain and
simple RS-232.

> Note that this wiring scheme puts RX+ and RX- on one twisted pair, which 
> cuts down on interference. Same for TX+/TX- of course.  This let me go from 
> ~90 feet w/o errors on flat satin cable @ 9600 to 150+ Feet on Cat5 at 19200.

Twisting each signal with a ground can be benificial too, which is why you
might experience some gains.

> Also, the slower you go on RS-232, the farther you can go. The oft-quoted 
> 50 feet spec is for 9600 Baud.

True.

	Johnny

Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt@update.uu.se           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol