> The right thing to do is "shift down" one UDMA mode -- the problem will
> then generally go away.  Your drives almost certainly can't sustain
> more than the 16.6MB/sec UDMA mode 1 will get you anyway, and since
> we don't support IDE disconnection yet the extra bandwidth the higher
> modes get you isn't useful for putting two drives on the bus, either.

So there _is_ a disconnect feature, like with SCSI?  Hmmm....

That's what I was talking about WRT the legendary "lost interrupt" problem, 
suggesting the drive was doing a thermal recalibration.

Is it possible the problem we all have is not with electronics, but with 
protocols -- drives that want to do things the driver doesn't support and 
can't deal with, or in situations where the driver doesn't think such things 
should happen?  That too would be solved by shifting down to an older standard.

"UltraDMA" still being a relatively new standard (33 even, let alone 66), 
maybe not all of the equipment is in total agreement.  Lord knows we've seen 
that often enough with SCSI.

Of course, it's always possible there could be subtle disagreement about 
electrical specifications.  We've seen that, too.

In either case, the only trouble I had was during a heat wave.  I've had none 
since it cooled off around here.


--Jim Howard  <jiho@mail.c-zone.net>