> > Yes the jumper marked 25/33 is for running the 040 at 25 or 33 MHz with a
> > 50 or 66 MHz oscillator.)  Somebody at HP once told me that it "should"
> > work at 33.
> > 
> So you are saying that a 380 could be upgraded from a 25mhz 040 to a 33mhz
> one with the proper crystal?  Might have to try that with our main server
> machine (it has 3 SCSI controllers, 3 HP-IB controllers and 3 LAN cards so
> we cannot just move all its peripherals to our hp433 box due to lack of DIO
> slots).

Like I said, I was told that it __should__ work.  I haven't tried this.  I
haven't even gotten my machine to pass ROM diagnostics with an 040 at 25MHz.

A really quick test would be to just change the oscillator and move the
jumper.  Total cost: about $5 for the oscillator.  If this works then you
could decide whether you wanted to risk continuing to run a 25MHz part at
33MHz.  (Definitely add a heat sink, heat sink goop and maybe a fan if you
do continue to run it.)  Or if you want to buy a 33MHz part, I think they
run about $250 in small quantities.

> > If I had a ROM burner and a program that would let me read ROM contents
> > from a live system ...
> > 
> This is most likely illegal.

Ah yes.  I was merely waxing theoretically.  ;-)

> Anyway, the ROM is mapped into the physical memory address space starting at
> the low 64k or so I believe.  There is a test in hp300/mem.c which prevents
> /dev/mem accesses from going outside of RAM.  If you disable that, you should
> be able to see the ROM contents.  The ROM uses the first page of physical RAM
> as scratch space (which BSD never touches).

Sounds like fun.  I'll take a look.

Thanks

Dave

------------------------------------------------------------------------------