PCbanter - View Single Post

**Paul[_32_]** · March 6th 21, 07:56 PM posted to microsoft.public.windowsxp.general

wrote:
Hi,

I have a 20 year old Gateway desktop PC (has WinXP).

Recently I experienced an odd issue using the
FDD (original with PC). Unlike problems I had encountered
in the past, such as Read Errors or Can't Access A: due to
a defective FDD in other PCs, there were NO error messages,
but a NEW FD I wrote files to did NOT "show up" in the A: directory!

I checked (Format A: /U) this FD in another PC and there were
NO bad sectors.

I replaced this 20 year old FDD with a spare. Everything is AOK
now.

Is there a FDD "integrity" test tool?

Thank You in advance, John

There have been commercial products with
specific tests for hardware before, but the
vast majority of them pay lip service to actually
testing anything. I even bought one, but was
sorely disappointed in my purchase.

The only one that came remotely close, was the
test software on a Sun computer. If you flipped
the switch on the equipment and put it in maintenance
mode, I think the software would be available
around boot time. and it could suss a keyboard
issue (like keyboard switches set to wrong locale
or something). Anyway, that's the only test suite
I ever used, that at least the tests looked like
the test writer really cared about results. Maybe
they connected actual duff equipment, just to make
sure the detections worked.

Magnetics and magnetic materials, sometimes the
heads get coated in crud. The eight inch floppies,
the media wasn't all that good, and Paul kept two
boxes of cleaning floppies and alcohol packs at
his desk, and when the usual suspects would show
up at my desk complaining about this or that,
I'd reach into my desk drawer, give them a cleaning
floppy and an alcohol pack, and send them on their
way :-)

We had a helical scan tape drive, where my fine fellow
employees were doing backups of chip designs. And
I was watching this, and going "have you people ever
checked these tapes ?". And sure enough, attempts
to read the tape yielded nothing. Inserting the
cleaner cassette and doing the ("once every 30 hours")
cleaning procedure, the drive came back to life and
cut us some good tapes. More magnets at work. You
would think these real life dramas would have some
training value, but no, you'd be wrong unfortunately.
Paul might have had a box of cleaning cassettes
for the helical scan, because you have to be ready
for these things (usual suspects).

I think it's quite possible, to break the head off
a floppy, and have it sit there with a grin on its
face. The software should be doing retries though,
if it cannot find a sync pulse or whatever. It should
"step to zero" to check for potential alignment issues
with the current seek.

The only thing a testing software could do, is
write-read-verify of the current floppy surface,
as a means of proving everything is working. I
don't think there is sufficient feedback regarding
mechanical issues (broken stepper). The presence
of a signal that makes the clock recovery on the
FDC chip work, should be all that you need to get
some sort of reaction. And if it cannot find
such a signal, you should hear the stepper moving
in and out. The floppy unit itself is completely
dumb - it is the driver that makes all the decisions
on steppers and the like. The floppy is just an
electromechanical device, no "CPU".

Paul