8 Sep
1999
8 Sep
'99
2:54 p.m.
At 04:05 PM 9/8/99 -0500, you wrote:
You wrote...
----->Almost all drives have an 'emergency retract' which is activated if
the
Even if that's true, there should be no way that the drive will spin down
with the heads still loaded. Even if the normal parking circuit fails, the
underspeed detection should trigger the emergency retract. Check out that
emergency retract circuit closely, it wasn't put there for nothing. It's
proper operation is critical to the survival of the drive. The drives that
I worked on had a wire-OR circuit with four diodes to monitor four
conditions that would cause an emergency retract. Power fail (5V logic
power, I think) and underspeed were two of the conditions.
As a suggestion, do you have a scrap disk that you can take the platter
off of and run the hub by itself? You could run it and watch what happens
without worrying about damaging the heads. Tony is right, your voice coil
sounds like it's ok since it will load the heads. You should be able to
check it's ability to retract by pushing it further in and it should
retract to it's correct position. It should react fast enough that you
can't move it before it senses the error and returns to the correct
position. If it moves and doesn't return then you have a problem. The
drives often use + and - power supplies to supply power for seek and
retract. You may have a bad power supply or it may have a bad transistor or
whatever they use to switch the power to the voice coil. Attempting to
force it past it normal position should detect such problems.
Joe
power fails while the heads are loaded. Typically,
this consists of a
relay that drops out when the power fails that then connects a large
capacitor, or more often a NiCd battery, to the voice coil. The head is
rammed against the outer end stop, where it opens a microswitch,
disconnecting the battery.
Yes, the power board does have a set of 3 nicads on it, and they are
definitely for retracting the heads in the event of a loss of AC power.
Since this involves somewhat violent head
movements, which are not good
for the heads (or, indeed, for the positioner), many drives _don't_ use
this to retract the heads on a normal unload. Instead they use the normal
servo system. Certainly the DEC RK05 does this (I have just looked in the
manual).
I haven't checked the schematics myself definitively, but, I heard from one
repair guy that it does not use the nicads for a normal unload, just for a
power fail unload.