Odd Q-Bus questions 18/22bit

From: Jerome Fine <jhfine_at_idirect.com>
Date: Tue Aug 18 07:55:06 1998

>Tony Duell wrote:

> Megan Gentry wrote:
> > You should be able to do that... but the VT103 had some special
> > paddle board which plugged into the connector on the VT100
> > basic video board and which provided a connector (2x5) which
> > connects to something like an MXV11-B, and a second connector
> > which also attaches to the MXV11-B (or DLV11-J) and routes the
> > signals through to the external connector (if I remember
> > correctly - otherwise it provides a separate external
> > conenction).
>
> That paddleboard connector on the VT100 is one of the cleverest pieces of
> design I've seen from DEC. The pins are supposed to short between the 2
> sides when there's nothing in the connector, thus linking the terminal
> logic signals to the DB25 on the back. Insert a paddleboard, and you
> disconnect them, allowing the paddleboard to connect separately to the
> VT100 logic and the outside world.
>
> I have the VT100 prints if you want to try to make the paddleboard yourself.

Jerome Fine replies:

While using the paddleboard is definitely the most convenient way to
run a VT103, it is also completely unnecessary. So while I would
pay a reasonable sum to buy a paddleboard if I was going to use
a VT103 in a commercial way - or if I was into restoring a VT103,
I might even build my own paddleboard so as to run the VT103
in the normal manner, for anyone who just wishes to run a VT103
and have fun, then the use of a paddle board is not required.

The primary purpose of the paddleboard was to allow the VT100
portion of the VT103 to be directly connected to the console
port of a DLV11-J (4 serial port SLU) and for the normal DB25
connection to be used to attach a serial line printer. Internally,
one of the other ports on the DLV11-J could then be attached
to the slowest "disk drive" ever, the TU-58. On one occasion
back over 10 years ago, a friend was demonstrating how long
it took to boot RT-11. Well, after 2 minutes, we were about
to go for coffee when it finally finished.

Without the paddleboard, it is necessary to use a cable to run
from the DLV11-J to the normal DB25 at the back to enable
the use of the VT100 portion as the console device. Then, in
addition, a serial line printer must be attached to other port
via a cable which then goes to the printer. One notable aspect
of the paddleboard (if I remember correctly - if necessary either
Megan or I can check for you if you have the paddleboard) is
that the baud rate on the DLV11-J is left to float for both
the console and printer port and the paddleboard sets the
baud rate for the console port at 9600 baud and the user
via the SETUP commands is able to set the baud rate for
the serial line printer port.

As far as I know, all DEC versions of the VT103 came
with an 18 bit backplane for the QBus. If an LSI-11/03
is used, I understand it may be ill advised to make any
changes in the backplane, as well as fruitless, since this
CPU can use only a maximum of 56 KBytes of memory
in any case.

However, both the 11/23 and the 11/73 CPUs are able
to use all 4 MBytes of memory with the primary OSs
for the PDP-11. However, I have also heard that the
use of a quad CPU with a 22 bit backplane may also
risk damage to the CPU in the VT103 environment.
BUT, I have seen both the dual 11/23 (M8186) and
the dual 11/73 (M8192) used with an upgraded 22
bit backplane within the VT103 along with a full
4 MBytes of memory. In addition, there was a
DLV11-J and a third party hard disk controller,
a Sigma RQD11-B (an MFM interface - which also
had BOOT ROMs for the MSCP device) along
with a 5 1/14" hard disk drive situated under the
tube. The disk drive happened to be a Micropolis
1325 of 70 MBytes. And while I would strongly
recommend against running with a DHV11 for any
extended period of time due to the lack of a proper
power supply on the VT103 (and especially in that
case, placing the hard disk drive on an external
power source - a PC power supply does very nicely),
that can also be done for short demonstrations of
up to an hour if it seems desirable.

What is very sad is that DEC had all this software
and most of the hardware available back in 1980
and Ken Olsen, along with the senior VPs at DEC,
decided that the PC market was not where DEC
wanted to be focused - namely high volume and
low unit profits. The VT103 was, already in 1980
with the 11/23, and shortly thereafter with the 11/73,
a far better PC than the MS-DOS/Intel combination
could produce until the 486 was released in 1989.
All that was necessary back in 1980 was a decision
to compete in the PC market - a decision that IBM
also refused to make as well.

If anyone has a VT103 and needs some help with
how to set it up as well to convert to a 22 bit backplane,
likely both Megan Gentry and I are able to supply the
necessary technical information.

Sincerely yours,

Jerome Fine
RT-11/TSX-PLUS User/Addict
Year 2000 Solutions for Legacy RT-11 Applications
(Sources not always required)
Received on Tue Aug 18 1998 - 07:55:06 BST