Re: ST506 and/or ST412 hard disk drives

> > From: petedunnington.plus.com
> ---snip---
> >>
> >> There was a signal on some drives that came on
the data cable.
> >> I think it may have been a write protect but I
don't recall
> >> exactly what it was. The ST506 may have used
this signal.
> >> I'm not sure if this is what he is talking
about.
> >
> > There's a "drive selected" signal, but I
think most drives have that.
> > It's there because the 34-way control cable is
daisy-chained but the
> > 20-way data cable is radial, one per drive.
>
>  Yes, that may have been it. If he is using just one
drive,
> this shouldn't be an issue.

Actually it would be if the controller depends on it, even
if you only 
have one drive. If that drive doesn't output the
'drive_selected_ signal, 
then the contorller might not enable any data receivers.
 
> 
> >
> >> I'd suspect things like step rate and number
of heads would
> >> be more important to him than anything else.
> >
> > The other important difference between the ST506
signals and the ST412
> > was that the ST506 didn't support buffered seek;
the timing of the step
> > signals had to be slow enough that the stepper
motor could keep up. The
> > ST412 was the first drive that buffered the step
signals, so they could
> > be sent rapidly, and virtually every hard drive
after that did too.
> >
> > 
> This is important because many drives that had the auto
step, were
> really slow using the fixed rate step. I had this
problem getting
 
The really strange lone (not the same interface, of course,
but similar 
in concept) is the SA4000 (14" Winchester). On that
drive, you eitehr 
have to send pulses so slowly that the the head movement it
completed  
for each one (that is, the head gets to the next track
before you send 
the next pulse) or  fast enough that you've sent all of them
before the 
heads start to move. An intermediate rate will end up with
the darn thing 
mis-stepping. This is docuemtned in the manaul and the
reason for it 
(one up/down counter with a common clock input to record the
head offset) 
is clear from the schematics.
  
-tony