ABEL DMAX/16 + DM/16 on VAX

[utzoo!decvax!ucbvax!unix-wizards]

>From decvax!ittvax!swatt at Berkeley Sat Dec 26 19:17:34 1981

Forgive me if this is old hat to some, but it might save a lot of
hassle for others.  We got the ABLE DMAX/16 and DM/16 boards (this is
their old product -- 2 boards for DMAX, 1 more for the DM).  We had
lots of trouble trying to get VMS to even tolerate the boards being
there (UNIX, as usual, was prepared to accept whatever you tell it).
The first thing is to totally ignore the "standard" placement
information for this device in the DEC PDP-11 Peripherals Handbook.
The PDP folks had a lot of ideas about "vector ranking" and the like,
which if you follow them will force you to restrap all your DZ boards.
They also had "standard" addresses for the first DH and DM which were:

	dh0:	0160020
	dm0:	0170500

If you put either board at that address, VMS will go totally bonkers
and refuse to connect to the DZ's (remember, we weren't trying to get
VMS to USE the DH, just co-exist peacefully with it).  Some talks with
folks at DEC and ABLE led to the DH address

	dh0:	0170100

which worked out.

After talking some more with ABLE, they suggested DM addresses which
turned out not to be usable with their older DM board, as it is very
restricted in what it can address.  Finally they came up with

	dm0:	0170000

which DID finally work with VMS.  I am assuming that second and
following DH-DM's can go at consecutive addresses from those origins,
but I have not tried it.  BTW: the DM at 0170400 also does not work.

If you use the new ABLE SUPERMAX (single board DH-DM), you should have
less problems as the DM can be addressed anywhere in the I/O page,
unlike the previous board.  It is also cheaper ($4100 vs. $5?00) and
uses 2 less backplane slots, which cost ~ $100 each.  If you deal with
ABLE and want this board, be sure to tell them you want to run it on a
VAX under UNIX, but still not interfere with VMS; they will give you
some addresses which work.  ABLE also has a VMS driver; I am not sure
whether it will work for either product, but certainly for the
SUPERMAX.

Now for the final painful issue: the new SUPERMAX does not support full
RS-232 modem control; it lacks the CTS-RTS functions.  The old DH-11
(and the ABLE DMAX look-alike) would allow system software to detect
CTS transitions and to toggle RTS; the DZ-11 and the new SUPERMAX do
not. The 4.?bsd driver doesn't use it however.

Maybe it's a lost cause and no one uses it anyway, but I thought
electrical flow control would be much preferable to XON-XOFF for use
with Local Area Networks.

	- Alan S. Watt (decvax!ittvax!swatt)