16 Oct
2007
16 Oct
'07
1:41 p.m.
On 10/16/07, Tony Duell <ard at p850ug1.demon.co.uk> wrote:
Right... it says "R80" on the front, but in a variety of places
(Ultrix device drivers, perhaps?) it's also called the RB80,
presumably to distinguish it from the RM80 and RA80 (same HDA, for
folks unfamiliar with it, but 3 different host interfaces).
That seems newer. Our 11/730 was ordered the day they were announced
at a DECUS, if I remember the story correctly. Our guys had a devil
of a time with the cable management, and had to escalate their calls a
couple of levels inside DEC for our "self-install" of the machine.
There was this especially tricky "Z fold" to measure. Get it right,
and the cables for the IDC disk (1 x 40 + 1 x 60 + 1 x 26), the DMF-32
(1 x 25 + 3 x 40), and the Unibus (wide, white, mylar flat cable) all
fit nicely in the tray and were properly protected when extending the
CPU. Get that Z-fold wrong, and it was a thick mess that was nearly
twice as thick as would fit. :-P
Once the cables were installed, though, it was a nice package. The
fact that it was an inexpensive (for the time) Unibus VAX, made it a
critical part of our operation for over 8 years. I learned to install
Unix (Ultrix v1.1) on that machine in 1985, and later, it was our VMS
5.0 host for relinking our product images against the VMS 5 run-time
libraries so we could support 4.x and 5.x simultaneously. I wrote a
script to stuff our object files over DECnet (via the DMF-32
DDCMP-compatible sync serial port) and pull the linked files back, so
that we could run one script on our build machine and have it crank
out both versions of the binaries into our tape-cutting directories.
That one little machine, tied to our 11/750, is how we kept our
company going for years and years. When we finally did get that 8300,
it was solely for VAXBI product development, never for day-to-day use.
To that, we trusted our VAX-11 hardware right to the very end day in
1993.
-ethan
Yes. There
were two system packages - the older 11/730-Z box
(10.5"-tall rackmount box in the middle of a 42" rack, with room for
an RB80 below, and an RL02 above), and a newer BA-11-style package
I've always known it as an 'R80' drive. It's got a somewhat SMD-like
interfacee, but different enough to give SMD hackers headaches. I am told
that the RM80 is the same drive with an external Massbus interface unit.
AFAIK, yes... the RM80 is an "R80" HDA on an RM02/RM03-like pedestal,
with a Massbus cage below. I wouldn't be surprised to learn there are
numerous board and packaging similarities )in the Massbus portion)
between an RM02/RM03 and an RM80. From experience, I know that the
RM02/RM03 HDA is a repackaged CDC 9762 (same guts, different shell),
*but* a real 9762 is SMD (and works with Emulex controllers, SI 9900s,
etc) and the RM02/RM03 does not, despite identical cabling. I was
assigned the task, once, to attempt to convert an RM02 back into a
real SMD drive - in effect, making it a 9762 with a DEC faceplate, but
it was beyond my skill level at the time. 80% or so of the boards in
the HDA card cage were absolutely identical between the two drives,
and the obivous differences didn't seem insurmountable, but it just
didn't work. Thinking back, one of the things I _didn't_ check was
the card cage backplane wiring - perhaps there was a difference there.
My 11/730 came in that type of cabinet, and was a
standard DEC
configuration. Intead of the RL02 (which, BTW, was linked to the
Integrated Drive Controller, there's no sepratate RL11 card in the
machine), there's an TSU05 front-loading 1/2" magtape. A rebadged Cipher
F880 IIRC. That does link to a separate controller card in a normal
Unibus slot.
I think I saw that once or twice in the real world - tape rather than
RL02 on top.
Yes... I didn't explicitly state it (but implied it via the device
names DQA0: and DQA1:) that the RL02 did not require a RL11. As far
as we were concerned, it was a feature, since RL02 was our
"sneaker-net" medium of choice, and at the time, an RL11 was on the
order of $600.
(though the
11/730-Z makes it easy to route the BA-11
cable, and the 11/725 does not).
There was an official DEC way to put a Unibus expansion cabinet on the
11/730. A board that went in the Unibus Out slot on the CPU backplane, 3
ribbon cables to a paenl on the bulkhead, then a screened ribbon cable
with 3 connectors on each end to a similar panel on the rack containing
the expanison box, more ribbon cables to a PCB in the Unibus In slot of
the backplane i nthe expansion box. I have this in my 11/730.