On Fri, 4 Feb 2005 15:12:35 -0800 (PST), Vintage Computer Festival
<vcf_at_siconic.com> wrote:
> Well, as far as I can tell, the DSD and MXV11 are pretty much drop in
> replacements for the equivalent DEC controllers. They work on the same
> address vector and same interrupt, and are supposed to be 100% compatible.
Yes... there are no funny drivers, no funny treatments, etc...
*except*... sometimes, some of these cards provide boot code, and I
don't know without looking them up, which ones do and do not. Once
the OS is coming up, though, you are right... they are 100%
compatible.
> Is it possible my bus isn't terminated properly? I've got a M9400-YA at
> as the last device in the bus, which is supposed to provide "120-ohm
> terminators with refresh & floppy boot".
I suppose it's possible... ah... refresh and floppy boot... you should
definitely check your floppy controller for boot code... none of the
DEC floppy controllers come with boot code,
so things can be a little different when you have 3rd-party cards...
one has to know the caveats when going off the 100% DEC path. Back in
the day, many people did just that, but they picked certain cards for
price and performance (mostly price) if they weren't prepared to pay
what DEC was asking for certain things. Consultants who were widely
versed in such things used to get lots of money for that kind of
knowledge and flexibility.
> I have considered going with 100% DEC but I can't find an RXV21.
Fair enough.
> Besides, all the hardware I'm playing with worked at one point, and worked well.
There's no reason to expect that you are having anything more than a
conflict or a corrupt floppy or something simple. Unfortunately, it
doesn't sound as if you have the gear set up to confirm what's on the
floppies you have - only the unit you are testing in the first place.
> The only changes I've made is putting the DSD controller in place of the
> MXV21, and replacing a clock/calendar/terminator board with the M9400.
> Maybe I should put the clock/calendar back in and see if that fixes
> things...
Perhaps... or maybe there's a way to disable the boot code on the
M9400 (don't have my books in front of me). In any case, you should
confirm that you have one copy of boot ROMs in the memory map, or zero
(and type in a bootstrap via ODT)
> Well, I always do get stuck with the stupid problems. There's got to be
> something higher level going on here. Are there any general diagnostic
> type things I can do? Any code I can type in through the ODT console to
> test out various things?
No type-in diagnostics as such... one can verify that low memory is
working (possibly a good quick test), and do a few, simple things, but
in the case of Qbus machines, paper-tape diagnostics were falling from
favor by 1976, and it was just assumed that if swapping cards didn't
fix a problem, swapping the entire box probably would. I am pretty
sure that once floppy-based systems started pushing paper tape systems
out of the low end, the diagnostics tended to be delivered on floppy
and required a machine that could at least boot a simple OS or program
loader framework like XXDP.
One could _write_ some test programs, but they would take a long time
to type in and verify - an autoloader like we wrote in FORTRAN to
force-feed programs via ODT would be much handier for this. When
interfaces were simpler to boot (RK11, etc.), toggle-in boots and
paper tape diagnostics were small enough to be viable. Programmed-I/O
devices like the RX11 need a lot more code to make work, but it could
be done.
-ethan
Received on Fri Feb 04 2005 - 17:40:47 GMT
This archive was generated by hypermail 2.3.0
: Fri Oct 10 2014 - 23:37:35 BST