At 06:26 AM 8/10/00 -0400, Jon Lindgren wrote:
>Last night, while reviving my 4000/200, I was trying to give my VAX a nice
>disk.  I've got a CMD qbus scsi controller which seems to work dandy -
>except for one item.  It doesn't seem to like larger disks.  I started
>with an RZ55, although I either need jumper wisdom or the drive is dead
>(didn't spin up, couldn't see the drive anywhere on the bus).

This is the 'spin up on command' problem. If you can't jumper the drive to 
start spinning on power up you're hosed for now. Sometimes a 'sho dev' will 
start up the drives.

>   Trying with
>a spare 9GB seagate (an older, FH drive), it was probed no problem, but
>gave media errors (the disk _does_ work).  I tried another 9GB drive of
>exactly the same model - no go, it failed in exactly the same manner.

This is a firmware problem on the CMD. It is limited to drives of 8GB or 
less. There is a firmware update that can fix this, or you can use the code 
pages on the Seagate to make it look smaller than it is. (the latter is 
more reliable)

>I got it working beautifully with an _old_ 300MB HP drive.  It makes more
>noise than a small honda, but it does seem to work okay.

:-) I have an RZ23 that sounds like the concorde on a take off roll

>Is the CMD controller just picky?  All of these drives have worked in
>sparcs, i386 and alpha machines no problem.  The 9GB drive was pulled from
>a working RAID array.  I figured maybe the CMD controller was just too
>picky about timing, maybe geometry (although it shouldn't
>matter) etc... but backwards compatibility with SCSI is one thing I've
>never had a problem with.

No, its just a bit older generation than most. Of all the SCSI controllers 
I've tried on Q-bus machines the CMD has given me the least problems. There 
is also the issue that if you're CMD wasn't told to "autosize" the drives 
then it might have stored some drive geometry in its NVRAM that is bogus.

--Chuck