problem with new WD drives (patterned UNC or *S07)

[drc]

I've had a handful of newer western digital drives come through with this problem. The ones I can think of off the top of my head have been 2500BEVS-UST, 3200BEKT, and 3200AAJS-VWA. Essentially for the first half (or so) of the drive there are regularly repeating sections of UNC errors (sometimes scanning it in MHDD actually produces strange results with *S07 showing up in the scan) and at some point around the halfway mark the drive clears up completely and is fine. I've posted about this before and people suggested head troubles, but I'm not sure I'm convinced. Anyone else running into this? PMs are welcome.

Attachments

[drc]

No, this came straight out of an apple computer.

[BlackST]

I vote for a head becoming weak... when the drive switches to THAT head, the show begins. But can also be a scratch on one zone, or FW related. Try ERASING with MHDD. If problem is persistent, or the sectors from X become slow. the drive need to be serviced. Maybe a selfscan can solve. One thing is sure : these drives are crap.

[drc]

With a head becoming weak would you not expect to see this issue across the entire drive? Total LBA on this one is ~488M and after ~180M or so there are no problems at all. I'm watching the scan go past here and the UNC errors are happening with exactly 1125 or 2250 good sectors in between.

Also, am recovering data first. After that is finished I can play with erasing and such.

[BlackST]

No, data is "distributed" between the heads and platters.

[drc]

I understand that. Ordinarily if a head has failed or is weak I would expect to see something like 2 out of every 8 rows in MHDD come up with errors over the course of the whole drive, instead of just in the first third which is sort of what is happening here. In this case it is not even 2 full rows, just one UNC or AMNF every 1125 sectors or so.

[pcimage]

I've seen similar things before, including drives that are virtually unreadable for the first ~66%, with the remaining 37% perfect.

PCB swap has solved it, after reprogramming ROM of course.

[HDD Spaz]

I may be wrong but can a portition of the head become weak? The size of a head is much larger than a MHDD block. Maybe a portion of the head is failing/demagetised or dirty?

I have no evidence to support this as this is something just popped in my head. Please feel free is rubbish this theory if its wrong.

[drc]

I've seen similar things before, including drives that are virtually unreadable for the first ~66%, with the remaining 37% perfect.

PCB swap has solved it, after reprogramming ROM of course.

Interesting.

[pepe]

it can be a tiny ~radial scratch that makes one sector in each track bad on one or more heads. Usually such problems develop when the drive suffers some physical shock...

pepe

[drc]

it can be a tiny ~radial scratch that makes one sector in each track bad on one or more heads. Usually such problems develop when the drive suffers some physical shock...

pepe

Conceptually this seems reasonable as well, though it doesn't account for why it happens specifically to the new WD drives unless they made some change to be more susceptible to this particular kind of damage.

[BlackST]

Conceptually they are crap drives built having in mind probably FIRST the low cost and users do the rest handling them like we know.

[fayn]

But this does not explain why a PCB swap HELPS!

We used to have the very same problem with older WD drives as well.

Does anybody have an idea WHAT makes the difference between (causes the problem on) the original PCB and (does not cause the problem) the other (donor, reprogrammed) one? I believe it's not in the EEPROM -- there's nowhere to "hide" any more data...

[drc]

Well, in the case of the tiny radial scratch, a PCB swap wouldn't help.