Drive Gremlins continuing saga

[alfredeneumann]

Did some reading.

Appears there is no reliable way, according to several sources other
than running b/mark tests of determining the UDMA operating level of a
drive, again according to two articles I read.

Yet WD has, as part of it's DLG tools a program to both determine the
level and set it. I wonder what they mean by "setting it".

This crap machine keeps crashing hard drives from appearances. The 2nd
WD 8 GB I put in now has a bad spot. Both of the drives this machine
killed tested ok prior to putting them in.

I am guessing this might be due to mismatch between UDMA levels on the
two drives on the same IDE channel. I zeroed the drive with the bad
spot, but neither WD zeroing program nor scandisk could go to completion
as the bad spot was near the end of the drive. I reformated it and
partitioned out the bad spot so now it is 7GB. With transfers within the
same larger ok drive, I notice that those that are on widely disparate
areas of the drive take forever, regardless of their size, like 8-10
seconds for copyig a small directory. Directories that were created
around the same time transfer quickly.

I tried to transfer a 800MB directory from the larger 20GB ok drive to
the smaller drive's partition both of which are on the same IDE channel.
They are connected by an older rather crimped up 80 wire ide cable, with
no color coding for the connectors, but it is 80 wire. The system
reported it would take 70 minutes or roughly 200kb per second transfer
rate.

BIOS is set to auto and DMA L. 2 is reported, but greyed out. OS is set
to DMA. Ran WD program which reported the drives already being set to
DMA-66. When I rebooted from this UDMA detection program, I tried the
same transfer (which I had earlier aborted due to the 70 minutes) and
suddenly it was very fast, took just a minute or two. So I'm guessing
again this WD program may have done something to set the DMA levels,
even though I told it to do nothing and even though it reported the
drives already being set to DMA 66.

This whole thing is very confusing. Earlier Everest readings reported
the set rate at DMA 33 for both drives. I'm afraid to put any more slave
drives into this machine as it keeps eating them. Any fresh ideas than
what has already been said?

 

[Rod Speed]

Did some reading.

Tad radical.

Appears there is no reliable way, according to several sources
other than running b/mark tests of determining the UDMA
operating level of a drive, again according to two articles I read.

Dont believe it.

Yet WD has, as part of it's DLG tools a program to both determine
the level and set it. I wonder what they mean by "setting it".

Some drives can have that set in the drive, what they will use.

This crap machine keeps crashing hard drives from appearances.
The 2nd WD 8 GB I put in now has a bad spot. Both of the drives
this machine killed tested ok prior to putting them in.

Some drives can do that if the power is inadequate, you can end
up with bad sectors due to the power sagging while its writing.

I am guessing this might be due to mismatch between
UDMA levels on the two drives on the same IDE channel.

Why is it likely to be that ? I've run plenty with different supported
levels on the one cable without losing a single drive in the process.

I zeroed the drive with the bad spot, but neither WD zeroing program
nor scandisk could go to completion as the bad spot was near the end
of the drive. I reformated it and partitioned out the bad spot so now it is
7GB. With transfers within the same larger ok drive, I notice that those
that are on widely disparate areas of the drive take forever, regardless
of their size, like 8-10 seconds for copyig a small directory.
Directories that were created around the same time transfer quickly.

I tried to transfer a 800MB directory from the larger
20GB ok drive to the smaller drive's partition both of
which are on the same IDE channel. They are connected
by an older rather crimped up 80 wire ide cable,

That doesnt matter unless its so bad that a wire has broken.

with no color coding for the connectors, but it is 80 wire.
The system reported it would take 70 minutes or roughly
200kb per second transfer rate.

Urk. Likely that is the drive retrying on bads.

BIOS is set to auto and DMA L. 2 is reported, but greyed out.
OS is set to DMA. Ran WD program which reported the drives
already being set to DMA-66. When I rebooted from this UDMA
detection program, I tried the same transfer (which I had earlier
aborted due to the 70 minutes) and suddenly it was very fast,
took just a minute or two. So I'm guessing again this WD
program may have done something to set the DMA levels,
even though I told it to do nothing and even though it
reported the drives already being set to DMA 66.

There was some problem in that area in the era of drives that size.

Forget the detail now tho.

This whole thing is very confusing. Earlier Everest readings
reported the set rate at DMA 33 for both drives. I'm afraid
to put any more slave drives into this machine as it keeps
eating them. Any fresh ideas than what has already been said?

The eating may just be that the drives arent getting power properly
or the metal tunnels in the power connector have opened up over
time and now dont make good contact, and the bads are due
to the power coming and going intermittently while being used.

 

[alfredeneumann]

Tad radical.

haha, when all else fails, RTFM

Dont believe it.

I only believe about 10% of what I read on the net, haha.

Some drives can have that set in the drive, what they will use.


Some drives can do that if the power is inadequate, you can end
up with bad sectors due to the power sagging while its writing.

Could be. Reseated all the power connections, emoried the power
connector to make sure the metal contacts were not oxidized. Maybe I've
got a bad power outlet. No trouble with other machines though.

Why is it likely to be that ? I've run plenty with different
supported levels on the one cable without losing a single drive in the
process.

Don't know, just speculating, am very puzzled by the whole thing.

That doesnt matter unless its so bad that a wire has broken.

Well, other sources disagee on this point. One said, no bad crimps,
that makes sense since a wire can be broken internally or the insulation
shot in that one area. Also WD advises to separate the cables from the
various channels as much as possible to avoid cross-talk.

Urk. Likely that is the drive retrying on bads.

Next step is replacing the cables as soon as I can. There are so many
brands out there, it's hard to figure which is the best value; I was
unable to find any test data on the 40 vs 80 wires, or anything related
to IDE cable testing.

There was some problem in that area in the era of drives that size.

Forget the detail now tho.

Yeah, Dell support page says certain models had trouble taking the bios
DMA settings, but never said which models.

The eating may just be that the drives arent getting power properly
or the metal tunnels in the power connector have opened up over
time and now dont make good contact, and the bads are due
to the power coming and going intermittently while being used.

Well read an article from a data recovery firm on all the drives it gets
which rated WD drives as garbage and Seagate as the best consumer
brands, saying that all the consumer stuff was basically not very good
and to go with SCSI. I guess I'll have to try to test the power supply
as well. Now I gotta buy another drive. I think you might be right about
the power thing though. Thanks for the input.

 

[Rod Speed]

haha, when all else fails, RTFM

I only believe about 10% of what I read on the net, haha.


Could be. Reseated all the power connections, emoried the power
connector to make sure the metal contacts were not oxidized. Maybe
I've got a bad power outlet. No trouble with other machines though.


Don't know, just speculating, am very puzzled by the whole thing.

Well, other sources disagee on this point.
Nope.

One said, no bad crimps,

They mean bad crimps in the sense that the
connectors are crimped onto the ribbon cable.

that makes sense since a wire can be broken
internally or the insulation shot in that one area.

Its rare to see that in practice.

You certainly do see badly crimped connectors and
those that are coming apart because the ribbon is used
to yank the connector off a drive with a tight connector.

Also WD advises to separate the cables from the
various channels as much as possible to avoid cross-talk.

Crosstalk wont produce a bad drive.

And I've never seen a problem with crosstalk anyway.

Next step is replacing the cables as soon as I can. There are so
many brands out there, it's hard to figure which is the best value;

Just make sure its a flat ribbon cable that is no
longer than the legal 18". It aint rocket science.

I was unable to find any test data on the 40 vs 80
wires, or anything related to IDE cable testing.

They arent worth testing other than by replacement.

Yeah, Dell support page says certain models had trouble
taking the bios DMA settings, but never said which models.

Well read an article from a data recovery firm on all
the drives it gets which rated WD drives as garbage

Plenty agree with them there.

and Seagate as the best consumer brands,

I prefer Samsungs myself, much quieter.

saying that all the consumer stuff was basically not very good

Thats bullshit, most dont get drives failing.

and to go with SCSI.

Thats complete pig ignorant obsolete silly stuff.

I guess I'll have to try to test the power supply as well.

Again, the best test is replacement.

Now I gotta buy another drive. I think you might be right
about the power thing though. Thanks for the input.

No problem, thats what these technical groups are for.

 

[Guest]

Reseated all the power connections, emoried the power
connector to make sure the metal contacts were not oxidized.

Don't do that; it can remove the plating and lead to oxidation problems
in the future. Just use alcohol, preferrably 90% or higher purity, and
plug and unplug the connector several times or wipe the contacts with a
popsicle stick or toothpick. Allow alcohol to evaporate completely
before reappling power. But more often the problem is loose contacts,
which can be bent slightly (emphasis) to retighten. Bad contacts,
crimps, and broken wires can be detected by measuring voltages at the
circuit board and comparing them to the voltages on unused connectors.

Next step is replacing the cables as soon as I can. There are so many
brands out there, it's hard to figure which is the best value; I was
unable to find any test data on the 40 vs 80 wires, or anything related
to IDE cable testing.

With anything faster than ATA/33, an 80-wire cable is a must for
reliability, but round cables vary in quality, and many are made
incorrectly, with straight wires instead of twisted pairs (TPO). The
safest choice is a $1 generic flat 80-wire cable.

 

[Guest]

Crosstalk wont produce a bad drive.

And I've never seen a problem with crosstalk anyway.

WD once produced drives that would stop working if a cell phone was
used only inches from it. I believe a power down was required to
restore operation.

 

[Rod Speed]

(e-mail address removed) wrote

Rod Speed wrote

WD once produced drives that would stop working
if a cell phone was used only inches from it. I believe
a power down was required to restore operation.

That's not crosstalk.