Re: PDP-11 in a BA440 box?
Bob Armstrong wrote:[color=blue]
> I realize that this is a very off the wall question, but .... Is it
> possible, if one were the type of person who was inclined to do such a
> thing, to take a VAX 4000 in a BA440 box, remove the VAX CPU and memory
> and replace them with a PDP-11/53 and a KFQSA? The 11/53 (a 11/53+,
> actually) is the DECserver 500 variety, with the S-Box handles and
> 1.5Mb of memory.[/color]
If you're using just the Q-bus portion of the box -- basically, you
would be using the Q-bus I/O slots and not the processor slots -- you
might be able to get this configuration to work.
The BA2xx series box would be a rather better choice, as the whole
box was Q22/CD slots. The BA4xx series has a different slot set-up for
the processors, and operates the Q-bus solely as the I/O bus.
If you were feeling mildly perverse, you might look around and find
out if you can configure the PDP-11 S-box module as a co-processor, and
to configure it to run in conjunction with the VAX processor. (Some of
the Q-bus processors including a few of the MicroVAX and VAX Q-bus
processors could be configured to operate as "secondaries". I don't
know that this PDP-11 module supports such, however.)
[color=blue]
> I know the DECserver boot ROMs are a problem, but they're standard
> 27256 EPROMs that are easily reprogrammed.[/color]
I'm not familiar with the processor module used in the DECserver 500
series.
[color=blue]
> Does the KFQSA emulate a QDA/UDA50 well enough to fool the PDP-11 OS
> into working with RF7x drives?[/color]
It does a reasonably good emulation, but whether or not RSX-11M+ or
whatever PDP-11 operating system you seek to boot and run here is
sufficiently fooled is another and rather more complex matter. Things
like unexpected disk responses and unexpected disk sizes have been known
to derail versions of OpenVMS, for instance.
[color=blue]
> Are there any other problems with the arrangement that I haven't
> thought about?[/color]
I'd certainly expect to learn something with any similar
transplantation operation, and I'd tend to expect I might fry some
hardware. (I'll assume you are familiar with working inside enclosures,
and know about the hazardous and potentially fatal voltages that can be
present.)
Re: PDP-11 in a BA440 box?
Bob Armstrong wrote:[color=blue]
> Does the KFQSA emulate a QDA/UDA50 well enough to fool the PDP-11 OS
> into working with RF7x drives?[/color]
Under late 5.x RT-11, it does work. RSTS/E and maybe RSX-11 could throw
a fit if they don't like what's coming back from the KFQSA.
Each drive shows up at its own new CSR - I suspect that RSTS/E and
RSX-11 can be configured to work with varying amounts of pain (as they
have to be sysgenned to know what unit numbers go at what CSR's). Under
RT-11 the multi-controller driver works like a charm for at least a
couple of drives.
Tim.
Re: PDP-11 in a BA440 box?
In alt.sys.pdp11 Hoff Hoffman <hoff-remove-this@hp.com> wrote:[color=blue]
> Bob Armstrong wrote:[color=green]
> > I realize that this is a very off the wall question, but .... Is it
> > possible, if one were the type of person who was inclined to do such a
> > thing, to take a VAX 4000 in a BA440 box, remove the VAX CPU and memory
> > and replace them with a PDP-11/53 and a KFQSA? The 11/53 (a 11/53+,
> > actually) is the DECserver 500 variety, with the S-Box handles and
> > 1.5Mb of memory.[/color][/color]
[color=blue]
> If you're using just the Q-bus portion of the box -- basically, you
> would be using the Q-bus I/O slots and not the processor slots -- you
> might be able to get this configuration to work.[/color]
I run my 11/53 in a BA411 enclosure, after removing all the cards from
the processor slots. I have pictures, if interested:
[url]http://www.ba23.org/page0206.html[/url]
[color=blue][color=green]
> > Are there any other problems with the arrangement that I haven't
> > thought about?[/color][/color]
[color=blue]
> I'd certainly expect to learn something with any similar
> transplantation operation, and I'd tend to expect I might fry some
> hardware.[/color]
I still worrying about releasing the magic smoke. It is a strange
configuration :-)
--
Kirk Russell <kirk@ba23.org> [url]http://www.ba23.org/[/url]
Bridlewood Software Testers Guild Ottawa Ontario Canada
Re: PDP-11 in a BA440 box?
Bob Armstrong wrote:[color=blue]
> I realize that this is a very off the wall question, but .... Is it
> possible, if one were the type of person who was inclined to do such a
> thing, to take a VAX 4000 in a BA440 box, remove the VAX CPU and memory
> and replace them with a PDP-11/53 and a KFQSA? The 11/53 (a 11/53+,
> actually) is the DECserver 500 variety, with the S-Box handles and
> 1.5Mb of memory.
>
> I know the DECserver boot ROMs are a problem, but they're standard
> 27256 EPROMs that are easily reprogrammed.
>
> Does the KFQSA emulate a QDA/UDA50 well enough to fool the PDP-11 OS
> into working with RF7x drives?
>
> Are there any other problems with the arrangement that I haven't
> thought about?
>
> Thanks,
> Bob Armstrong
>[/color]
Bob, there were the BA430 cabinets that were originally used with
something like a DECsystem 3100 or DECsystem 5100. The configuration
was very much like the third Q-Bus cabinets used in configurations like
the DECserver 5X0, VAX 4000-200 and some MicroVAX 3x00 units. I believe
that chassis was the BA213. The BA430 and the BA440 were SO much
different that there was no way to interchange the backplanes even
though from the outside the chassis looked identical. These were the
units that had the two separate doors, one for the top part with the
disks and tapedrives and the bottom door covered the system boards and
their "S-Box" handles. I've converted a DECsystem and used both a
MicroVAX 3800 and a VAX 4000-200 in these. I was also considering
trying to use a DECserver 5X0 11/53 in the box but never was able to
get a bootable TK50/TK70 working to load RSX. At the time I was going
to try using a KFQSA and DSSI drives but time never quite allowed me the
chance to get it all working correctly.
Without question using a BA213 is probably the easier solution. The
processor is already proven to work in that format.
bob