Dirty Head or Weak Head

[shahij]

Hi mates,

There are same cases found specially on Hitachi & Seagate HDDs where in every few sectors (around 90,000 to 1,00,000sectors) drive stacks or sequential UNCs then again good and again UNCs.

According to my understanding, it's Weak Head.

In this situation, either professional Imager like DDI or DDP or other can handle to make a good image? or do you go for Head Swap?

[deftrue]

These cases require a little bit of study and examination. If head is weak, the best thing is to image all as possible first. If head is defective or damaged, sometimes trying to image will cause a headcrash, and then you'll be out of luck. How to tell damaged heads from weak heads? Well... that's the pro's job! Also, sometimes it has nothing to do with bad heads: a bad translator can cuase the drive to display similar sympthoms.

If those sectors are just defective sectors, DDI will help, since it may skip such sectors after reporting UNC and try to read them in a different manner afterwards.

Also, you may try to understand the error codes returned when trying to read such sectors, and find out what's going on.

Bottomline: Swapping heads is fine for older drives, and model dependant. Newer drives may report lots of bad sector after head swap, due to the extreme and adaptive fine tuning they have. Some adaptation may be required, if possible (uJogs, etc.).

[shahij]

Thanks for your input.

Well, I have found 2 Seagate HDDs (2.5") recently also on 2 Hitachi (1 desktop & another Laptop), while the drive sounds fine, detects normally, in both cases it's found that in every 95,000 to 1,00,000 sectors there are some sequential UNCs to all over the surface (checked randomly). Well whenever it's bad translator either the FULL DATA AREA gets inaccessible, and for seagate F3 series you all know regarding PSA. But the case isn't like any of these 2 issues. Also there is a thought on me... the distance of UNCs are not same all over the surface but close enough. Like: 95000, 96000, 100000 sectors intervals. Also, I can't believe that a drive will have so much UNCs on surface, this is something else.

Another ques: did you find ever DIRTY HEADS?

[jono-ats]

Another ques: did you find ever DIRTY HEADS?

Yes.

[shahij]

Another ques: did you find ever DIRTY HEADS?

Yes.

Do you think the behaviour because of DIRTY heads? or something else?

[deftrue]

Well whenever it's bad translator either the FULL DATA AREA gets inaccessible, and for seagate F3 series you all know regarding PSA. But the case isn't like any of these 2 issues.

That's not always that way. I've just finished a case in which translator was wrongly initialized and bad sectors were given LBA addresses. Result: User area is accessible, but there are UNC's spanned across all the user space (or a portion of it). Plus, offsets on logic addresses.

Dirty heads are common on some models on which manufacturer applied a "greasy" coating to the platters, in order to prevent headcrashes. Over time, that coating fell off, and heads became dirty. However, if you can read up to a million LBA's with no problem, and the sectors reporting UNC errors keep reporting such condition upon several read retries, I would not point towards a dirty heads issue.

[jono-ats]

Dirty heads are revealed by magnification; "weak" heads discerned by looking at dynamic waveforms on a suitable oscilloscope.

Sometimes it's hard to differentiate read issues caused by weak heads vs. dirty ones, because contamination also weakens the ability of a head to read the tracks.

Sometimes you simply have to look under the streetlight . . .