21 Oct
2005
21 Oct
'05
12:02 a.m.
On 10/20/2005 at 6:23 PM Allison wrote:
Once upon a time drive need time to spin up before you
loaded the head
and read it. As drives improved that time shortend and the motors went
to brushless where relibility didn't degrade with stops and starts
(brush motors this is a relability/wear issue). So having seperate motor
enables allowed one to spin the drives and then allow a 10sec (or longer)
timeout after last access before they would stop.
I think we're saying the same thing in different ways. If you don't have a
head load signal, the head remains in contact with the media the entire
time the media's in the drive. If the spindle is rotating, that leads to
head and media wear the entire time (althought he drive needs to spin a bit
to seat the media correctly when first inserted). If the motor's turned
off, no media wear, even though the head's still in contact.
And yes, motors with brushes will fail with time. When I wrote my first
8085 diskette driver, I just turned the drive motors on at boot, having
come off an 8" floppy environment. I was admonished by a Micropolis
engineer that this was not a good idea and that the motor should be turned
on only when needed. So media was inserted into drives with the spindle at
a dead stop, leading to serious problems with seating the media. So drives
were modified to include a little spin-up when the drive door was being
closed. Not everyone did this gracefully, however, leading to wrinkled
areas around the media hub and so reinforcing rings were introduced.
The interesting thing is the chicken-and-egg nature of this. Initially,
you didn't want to spin the media all of the time on a 5.25" drive because
the brushes in the DC motor would wear. So a motor control line was
incorporated. But then, if you could stop the media from spinning, you
didn't have to have a separate head-load mechanism to guard the meda and
the head against wear while not in use.
Cheers,
Chuck