Subject: Re: mvme147 problems...
To: Frederick W. Koehler IV (Joe College) <koehler@valinor.chem.ohiou.edu>
From: Steve Woodford <steve@mctavish.demon.co.uk>
List: port-mvme68k
Date: 08/16/1998 21:48:39
On Sun, 16 Aug 1998, Frederick W. Koehler IV (Joe College) wrote:

> complaining that the nvram battery is dead (not surprizing).  Also,

Not unusual, now that most 147s are approaching 10yrs old :)

> anyone know of a source for a new nvram chip in the US?  

Pass. Someone should know though; they're pretty common parts.

> this application, even though they're the same chip, right?  I'd have to
> find one specific for the mvme, right? 

They're not 'mvme' specific as such, in that there's nothing which needs
to be pre-programmed to make them work with a 147. What youd need to do,
both with a new nvram, or one with a flat battery (which hopefully is all
that's wrong with yours) is to press the abort/reset buttons in the
following order to force 147bug to re-init. the nvram:

	. Press and hold abort
	. Press reset, while still holding abort
	. Release reset
	. 5 seconds later (or when 147bug prompt appears) release abort

You should now set the ethernet MAC address using the 'lsad' command. The
address is on a label on the inside of the front panel. Only the last five
digits need to be entered. Next, use the 'mm' command to set the 32-bit
word at 0xfffe0764 to zero.

> functional system w/o the nvram by setting defaults by hand, etc.  Have

Should be possible. I've done the same in the past.

> hangs (probably due to bad nvram?), if I do a iot;h, it only shows this:
>         Disk Controllers Available
> 
> LUN     Type        Address   #  dev
>  0      VME0      $00000000   0

Hopefully re-initialising the nvram solved this one. If not, try a new
nvram!

> of the 147?  I understand that the 147 can't detect newer scsi disks, if
> that's the case, how do I boot from disk, or would I always have to use
> s-records and then boot from the network?  

The 147bug prom will work perfectly with 'new' disks.

> still hangs.   Doesn't look good at this point.  After switching to
> 147-diag w/ sd, running a self test with st shows everything passing. 
> However, without a working scsi controller or tape drive, it's not much
> use.  

Yeah, ISTR Jason Thorpe having similar symptoms to this. I believe he got
a replacement nvram, which solved all the problems.

>   Here's a dumb question--can the mvme147 be booted by tape if the nvram
> is dead?  Can it be booted at all w/o the nvram working.  Also, even if
> it is working, can the machine boot (ab)  from a newer scsi drive?

Well, if the nvram is completely hosed as opposed to just a dead battery,
then it probably won't boot from anything. With a working nvram (dead
battery not withstanding), it'll boot from both scsi tape and scsi disk
(new or old). Note that 'new' scsi disks/tapes are backwards compatible
with 'old' systems by virtue of the fact that they have to be explicitly
told by the OS to switch to scsi-2 or whatever. They power-up in scsi-1
mode, which keeps the old 147bug prom happy ;-)

>   Any tips on getting this machine up or pointers to a new nvram would
> be appreciated.  Thanx...

One question, how much memory does your 147 have (4, 8 or 16Mb)? Unless
you have a scsi tape drive, you'll still have to download the srecord boot
program then netboot and install from some other server.

Cheers, Steve