On 5/18/2006 at 7:13 AM dwight elvey wrote:
Are you saying that it should take in what ever step
pulses I send it
and step at 8ms, regardless of what I send.
Not exactly--if you step between (according to the spec sheet) 5 and 500
usec, the drive treats it as a buffered seek request. Above that time, it
tries to step at the rate you give it.
AFAIK, only the original ST506 (and clones) required unbuffered seek. By
the time the ST412 came out (long before the PC XT), buffered seek was the
norm. It makes sense to use buffered seek as the drive electronics can
make use of the ballistic nature of the seek mechanism (Newton's first
law).
A small almost-on-topic story here. When my wife bought her PC, PC-XT
Taiwanese clones were just starting to make the appearance. The outfit she
bought it from advertised "free hard disk and controller with purchase".
It was a Xebec controller all right (nothing wrong with that), but a brand
stinkin' new ST506--which the firmware on the Xebec wouldn't drive because
of the unbuffered seek. Fortunately, it was a simple EPROM change (the
Xebec COULD perform a 3 msec seek, it just didn't have that in its DPT).
I've still got the 506, BTW, but the clone is long gone.
But yeah, the 251 is definitely a buffered seek drive and you should use
that capability if you can. Fire out your seek request with somewhere
around 100 usec pulses, the drive will count them and then perform the
seek, setting SEEK COMPLETE when done. And it will be quieter, too.
Cheers,
Chuck
That doesn't seem to be the case with my drive. I
is definitely stepping
at the rate I give it with a seek or restore command.