In the Western Dynax variety of these drives, there was a signal
called "funul" which was "function unload".
When a power supply was sensed to be out of spec, or any other
malfunction was detected, such as an out of range seek attempt,
funul would be pulled to kill any write current.
The heads would then have a connection made to a set of capacitors
which had sufficent energy to retract them very quickly (faster velocity
than a regular seek) back to the unload stops.
This ensured an unload when the power cable was pulled.
I had occasions where the heads were energized when the unload
happened and you could see dead sectors in a spiral all across many
tracks down to track 0, as the heads unloaded.
If you can run a verify program, and see this sort of thing, just a
sector
or two / track, you may have had this happen.
On our drives there was a few gates that could fail to cause this to
happen.
Jim
Tom Jennings wrote:
Last week, in the middle of debugging Kermit (and
making headway)
the 6070 disk dropped dead (I think the technical word for it is
'sh*t the bed').
<snip>