Oxidisation of Seagate & WDC PCBs

[Sergey Kubushyn]

[...]
That suicide also can happen when some old file that was not accessed for
ages is read. That attempt triggers the suicide chain.

Yes, that makes sense. However you should do surface scans on
RAIDed disks regularly, e.g. by long SMART selftests. This will
catch weak sectors early and other degradation as well.

I know but I simply didn't think all 3 drives can fail... I thought I have
enough redundancy because I put not 2 but 3 drives in that RAID1... And I
did have something like a test with regular weekly full backup that reads
all the files (not the entire disk media but at least all the files on it)
and that was that backup that triggered disk suicide.

Anyway lesson learned and I'm taking additional measures now. It was not a
very good experience loosing some of my work...

Yes, I can imagine. I have my critical stuff also on a 3 way RAID1,
but with long SMART selftests every 2 weeks and 3 different drives,
two from WD and one from Samsung. One additional advantage of the
long SMART selftest is that with smartd you will get a warning
email on every failing test, i.e. one every two weeks. For additional
warning you can also run a daily short test, e.g..

No matter what you do you can not prevent an occasional disaster One
MUST remember that "backup" in not a noun but a verb in imperative.
Indeed.

Those 4 were fine on the top of PCB. Black stuff was underneath, on those
pads contacting with springy heads pins.

Mine is fine on both sides. However there is a quite a bit of contact
area that looks and feels silver-plated to me, most notably areound
the screws and on the bottom the contacts to the head assembly.

That makes me wonder why are they silver-plated. It is definitely not the
best material longevitywise, especially for such low-level signals. It makes
me even more suspicious and adds to the conspiracy theory.

 

[Arno]

That makes me wonder why are they silver-plated. It is definitely
not the best material longevitywise, especially for such low-level
signals. It makes me even more suspicious and adds to the conspiracy
theory.

Well, maybe. However I tend to think that "never attribute to
malice what can be adequately explained by stupidity" may apply.

These contacts should be gold plated with high quality gold. It is
also possible that the HDD vibration (always present with a running
HDD) and thermal variation allows the process to creep between the
contacts and kill them. Maybe a young, inexperienced engineer was
hired to replace an older, experienced (but more expensive one)
and that person made a pretty bad judgement call due to
inexperience, wanting to save a few cents on the design.

I have to say that the last time I saw silver plating as contact
protection was in vaccuum tube equipment. Modern electronics
typically uses Gold, or Tin for low insertion cycle contacts.

I also found a statement on Wikipaedia that silver plated
copper, once the copper is exposed in a place, will rapidly
corrode all over because of some electro-chemical process.
No idea whether this is true or not.

Arno

 

[Rod Speed]

In comp.sys.ibm.pc.hardware.storage Sergey Kubushyn <[email protected]>
wrote: [...]
That suicide also can happen when some old file that was not
accessed for ages is read. That attempt triggers the suicide
chain.

Yes, that makes sense. However you should do surface scans on
RAIDed disks regularly, e.g. by long SMART selftests. This will
catch weak sectors early and other degradation as well.

I know but I simply didn't think all 3 drives can fail... I
thought I have enough redundancy because I put not 2 but 3 drives
in that RAID1... And I did have something like a test with
regular weekly full backup that reads all the files (not the
entire disk media but at least all the files on it) and that was
that backup that triggered disk suicide.

Anyway lesson learned and I'm taking additional measures now. It
was not a very good experience loosing some of my work...

Yes, I can imagine. I have my critical stuff also on a 3 way RAID1,
but with long SMART selftests every 2 weeks and 3 different drives,
two from WD and one from Samsung. One additional advantage of the
long SMART selftest is that with smartd you will get a warning
email on every failing test, i.e. one every two weeks. For
additional warning you can also run a daily short test, e.g..

No matter what you do you can not prevent an occasional disaster
One MUST remember that "backup" in not a noun but a verb in
imperative.
Indeed.

BTW, I took a look at brand new WDC WD5000YS-01MPB1 drives, right
out of the sealed bags with silica gel and all 4 of those had
their contacts already oxidized with a lot of black stuff. That
makes me very suspicious that conspiracy theory might be not all
that crazy--that oxidation seems to be pre-applied by the
manufacturer.

Urgh. These bags are airtight. No way the problem happened on your
side then. My two weeks old WD5000AADS-00S9B0 looks fine on the top
of the PCB. I think I will have a look underneath later.

Those 4 were fine on the top of PCB. Black stuff was underneath, on
those pads contacting with springy heads pins.

Mine is fine on both sides. However there is a quite a bit of contact
area that looks and feels silver-plated to me, most notably areound
the screws and on the bottom the contacts to the head assembly.

That makes me wonder why are they silver-plated. It is definitely not
the best material longevitywise, especially for such low-level signals.

Likely just some fool's reaction to the price of gold.

It makes me even more suspicious and adds to the conspiracy theory.

Nope.

 

[Sjouke Burry]

[...]
That suicide also can happen when some old file that was not accessed for
ages is read. That attempt triggers the suicide chain.
Yes, that makes sense. However you should do surface scans on
RAIDed disks regularly, e.g. by long SMART selftests. This will
catch weak sectors early and other degradation as well.
I know but I simply didn't think all 3 drives can fail... I thought I have
enough redundancy because I put not 2 but 3 drives in that RAID1... And I
did have something like a test with regular weekly full backup that reads
all the files (not the entire disk media but at least all the files on it)
and that was that backup that triggered disk suicide.
Anyway lesson learned and I'm taking additional measures now. It was not a
very good experience loosing some of my work...
Yes, I can imagine. I have my critical stuff also on a 3 way RAID1,
but with long SMART selftests every 2 weeks and 3 different drives,
two from WD and one from Samsung. One additional advantage of the
long SMART selftest is that with smartd you will get a warning
email on every failing test, i.e. one every two weeks. For additional
warning you can also run a daily short test, e.g..
No matter what you do you can not prevent an occasional disaster One
MUST remember that "backup" in not a noun but a verb in imperative. Indeed.

BTW, I took a look at brand new WDC WD5000YS-01MPB1 drives, right out of the
sealed bags with silica gel and all 4 of those had their contacts already
oxidized with a lot of black stuff. That makes me very suspicious that
conspiracy theory might be not all that crazy--that oxidation seems to be
pre-applied by the manufacturer.
Urgh. These bags are airtight. No way the problem happened on your
side then. My two weeks old WD5000AADS-00S9B0 looks fine on the top
of the PCB. I think I will have a look underneath later.
Those 4 were fine on the top of PCB. Black stuff was underneath, on those
pads contacting with springy heads pins.

Mine is fine on both sides. However there is a quite a bit of contact
area that looks and feels silver-plated to me, most notably areound
the screws and on the bottom the contacts to the head assembly.

That makes me wonder why are they silver-plated. It is definitely not the
best material longevitywise, especially for such low-level signals. It makes
me even more suspicious and adds to the conspiracy theory.

---
******************************************************************
* KSI@home KOI8 Net < > The impossible we do immediately. *
* Las Vegas NV, USA < > Miracles require 24-hour notice. *
******************************************************************

You know of course that the black silver layer is still conductive
for low level signals??

 

[Arno]

You know of course that the black silver layer is still conductive
for low level signals??

Silver Silfide is a (bad) conductor? That will help for the
R/W signal. However the lines for the moving coil go through
the same connector and they need a low resistance path.

Arno

 

[Sergey Kubushyn]

Well, maybe. However I tend to think that "never attribute to
malice what can be adequately explained by stupidity" may apply.

I agree but it looks like there is a pattern here...

These contacts should be gold plated with high quality gold. It is
also possible that the HDD vibration (always present with a running
HDD) and thermal variation allows the process to creep between the
contacts and kill them. Maybe a young, inexperienced engineer was
hired to replace an older, experienced (but more expensive one)
and that person made a pretty bad judgement call due to
inexperience, wanting to save a few cents on the design.

They did not save anything on that design. Gold plating is a common
procedure, it is everywhere, most of card-edge connectors (e.g. PCI) are
gold and they even called "gold fingers" by chinese PCB manufacturers.

Silver, on the other hand, is almost unheard of and I'm pretty sure PCB
makers would charge extra for this if they agree to do it at all. And it is
NOT that the entire board is silver-plated; there are gold-plated parts on
that same board that makes it have at least 2 different platings so it will
be more expensive than simple gold all over.

I have to say that the last time I saw silver plating as contact
protection was in vaccuum tube equipment. Modern electronics
typically uses Gold, or Tin for low insertion cycle contacts.

Yep. Silver plating was usually used in microwave equipment, HF coils etc.
where skin effect was so profound that current only ran through that silver
(that was quite thick, btw.) Silver is also used for HIGH CURRENT relay
contacts where the corrosion is removed by mechanical action of closing
contacts and burned through with high current.

If you look at low current signal relays with stated minimal current
capacity _NONE_ of them have silver contacts. It is usually gold, platinum,
rhodium, or a mix thereof.

I am all pro Occam's Razor but all this looks like deliberate effort to make
it fail after some time. It is NOT easier or cheaper to put silver there
because it is an _ADDITIONAL_ step and not so common one.

 

[Sergey Kubushyn]

[...]
That suicide also can happen when some old file that was not accessed for
ages is read. That attempt triggers the suicide chain.
Yes, that makes sense. However you should do surface scans on
RAIDed disks regularly, e.g. by long SMART selftests. This will
catch weak sectors early and other degradation as well.
I know but I simply didn't think all 3 drives can fail... I thought I have
enough redundancy because I put not 2 but 3 drives in that RAID1... And I
did have something like a test with regular weekly full backup that reads
all the files (not the entire disk media but at least all the files on it)
and that was that backup that triggered disk suicide.
Anyway lesson learned and I'm taking additional measures now. It was not a
very good experience loosing some of my work...
Yes, I can imagine. I have my critical stuff also on a 3 way RAID1,
but with long SMART selftests every 2 weeks and 3 different drives,
two from WD and one from Samsung. One additional advantage of the
long SMART selftest is that with smartd you will get a warning
email on every failing test, i.e. one every two weeks. For additional
warning you can also run a daily short test, e.g..
No matter what you do you can not prevent an occasional disaster One
MUST remember that "backup" in not a noun but a verb in imperative.
Indeed.

BTW, I took a look at brand new WDC WD5000YS-01MPB1 drives, right out of the
sealed bags with silica gel and all 4 of those had their contacts already
oxidized with a lot of black stuff. That makes me very suspicious that
conspiracy theory might be not all that crazy--that oxidation seems to be
pre-applied by the manufacturer.
Urgh. These bags are airtight. No way the problem happened on your
side then. My two weeks old WD5000AADS-00S9B0 looks fine on the top
of the PCB. I think I will have a look underneath later.
Those 4 were fine on the top of PCB. Black stuff was underneath, on those
pads contacting with springy heads pins.
Mine is fine on both sides. However there is a quite a bit of contact
area that looks and feels silver-plated to me, most notably areound
the screws and on the bottom the contacts to the head assembly.

That makes me wonder why are they silver-plated. It is definitely not the
best material longevitywise, especially for such low-level signals. It makes
me even more suspicious and adds to the conspiracy theory.

---
******************************************************************
* KSI@home KOI8 Net < > The impossible we do immediately. *
* Las Vegas NV, USA < > Miracles require 24-hour notice. *
******************************************************************

You know of course that the black silver layer is still conductive
for low level signals??

It is not. Look at low level signal relays with stated _MINIMAL_ current
capacity and think why none of them has silver contacts.

 

[Sjouke Burry]

It is not. Look at low level signal relays with stated _MINIMAL_ current
capacity and think why none of them has silver contacts.

I have done brain wave registration, eye movement detection and
skin resistance measurement with silver-chloride electrodes,
and they conducted nicely.

 

[Sergey Kubushyn]

I have done brain wave registration, eye movement detection and
skin resistance measurement with silver-chloride electrodes,
and they conducted nicely.

That is totally different application. Yes, silver sulfide is not a perfect
dielectric but it is not a good conductor either. And modern HDD heads are
magnetoRESISTIVE.

 

[454]

In comp.sys.ibm.pc.hardware.storage Sjouke Burry

Silver Silfide is a (bad) conductor?
Nope.

That will help for the R/W signal. However the lines for the moving coil
go through the same connector and they need a low resistance path.

The black silver layer conducts that fine.

 

[Arno]

I agree but it looks like there is a pattern here...

They did not save anything on that design. Gold plating is a common
procedure, it is everywhere, most of card-edge connectors (e.g. PCI) are
gold and they even called "gold fingers" by chinese PCB manufacturers.

Silver, on the other hand, is almost unheard of and I'm pretty sure PCB
makers would charge extra for this if they agree to do it at all. And it is
NOT that the entire board is silver-plated; there are gold-plated parts on
that same board that makes it have at least 2 different platings so it will
be more expensive than simple gold all over.

Good points. An exotic process would be more expensive than a
common one and two processes instead of one as well. I also happen
to know that putting gold directly on silcer is problematic, but
putting it directly on copper is fine. At least that is for galvanics
on jewelery and if I remember this correctly.

Yep. Silver plating was usually used in microwave equipment, HF coils etc.
where skin effect was so profound that current only ran through that silver
(that was quite thick, btw.) Silver is also used for HIGH CURRENT relay
contacts where the corrosion is removed by mechanical action of closing
contacts and burned through with high current.

That explains it. I have indeed seen it in power relais as well.

If you look at low current signal relays with stated minimal current
capacity _NONE_ of them have silver contacts. It is usually gold,
platinum, rhodium, or a mix thereof.

I am all pro Occam's Razor but all this looks like deliberate effort
to make it fail after some time. It is NOT easier or cheaper to
put silver there because it is an _ADDITIONAL_ step and not so
common one.

Well, it only makes the required level of stupidity larger,
because (if we have this right) they also need to mess up the
economic angle. If we assume they are competent, then indeed this
looks very much like a deliberate and rather bad design error.

Arno

 

[Arno]

I have done brain wave registration, eye movement detection and
skin resistance measurement with silver-chloride electrodes,
and they conducted nicely.

Silver cloride and silver sulfide are two different things.

Ok, finally looked it up:

Silver Sulfide (Ag2S) is black and forms when silver is
exposed to the atmosphere by a reaction with hydrogen sulfide.
As to conducticity, it seems this really messes up contact
characteristics, including formiong diode-like effects and the
like. I found an abstract of a IEEE article from 1970 online:
"Electrical Characteristics of Contacts Contaminated
with Silver Sulfide Film"
So it seems it does concuct, but not well, uniformly or even
in an ohmic fashion. Very bad. THis would explain the HDD
failures, I think. If such a noise source is found in the
signal path from/to the heads and the moving coil, I think
this can cause all sorts of problems.

Silver Chloride (AgCl)is a white crystal used as referecne
electrode, becasue it has very stable characteristics, giving
you 230mV +/-10mV against a standard hydrogen electrode. I
conclude from this that Silver Cloride conducts reasonably
well and mostly in an ohmic fashion.


Sorry, but your observation does not aplly to the discussion
at hand.

Arno

 

[454]

In comp.sys.ibm.pc.hardware.storage Sergey Kubushyn <[email protected]>
wrote:
[...]
Those 4 were fine on the top of PCB. Black stuff was underneath,
on those pads contacting with springy heads pins.

Mine is fine on both sides. However there is a quite a bit of
contact area that looks and feels silver-plated to me, most
notably areound the screws and on the bottom the contacts to the
head assembly.

That makes me wonder why are they silver-plated. It is definitely
not the best material longevitywise, especially for such low-level
signals. It makes me even more suspicious and adds to the
conspiracy theory.

Well, maybe. However I tend to think that "never attribute to
malice what can be adequately explained by stupidity" may apply.

I agree but it looks like there is a pattern here...

They did not save anything on that design. Gold plating is a common
procedure, it is everywhere, most of card-edge connectors (e.g. PCI)
are gold and they even called "gold fingers" by chinese PCB
manufacturers.

Silver, on the other hand, is almost unheard of and I'm pretty sure
PCB makers would charge extra for this if they agree to do it at
all. And it is NOT that the entire board is silver-plated; there are
gold-plated parts on that same board that makes it have at least 2
different platings so it will be more expensive than simple gold all
over.

Good points. An exotic process would be more expensive than a
common one and two processes instead of one as well. I also happen
to know that putting gold directly on silcer is problematic, but
putting it directly on copper is fine. At least that is for galvanics
on jewelery and if I remember this correctly.

Yep. Silver plating was usually used in microwave equipment, HF
coils etc. where skin effect was so profound that current only ran
through that silver (that was quite thick, btw.) Silver is also used
for HIGH CURRENT relay contacts where the corrosion is removed by
mechanical action of closing contacts and burned through with high
current.

That explains it. I have indeed seen it in power relais as well.

If you look at low current signal relays with stated minimal current
capacity _NONE_ of them have silver contacts. It is usually gold,
platinum, rhodium, or a mix thereof.

I am all pro Occam's Razor but all this looks like deliberate effort
to make it fail after some time. It is NOT easier or cheaper to
put silver there because it is an _ADDITIONAL_ step and not so
common one.

Well, it only makes the required level of stupidity larger,
because (if we have this right) they also need to mess up the
economic angle. If we assume they are competent, then indeed this
looks very much like a deliberate and rather bad design error.

Or some fool has focussed on the price of gold metal and has
lost sight of the fact that more complex pcb manufacturing
process negates any advantage by using the cheaper metal.

MUCH more likely than any conspiracy to shaft the user.

 

[Sergey Kubushyn]

Just took a brand spanking new WD5000AAKS drive out of sealed bag with
silica gel and all that stuff. The PCB is all _SILVER_ plated, no gold. And
that silver is almost totally black right out of the bag.

 

[Arno]

In sci.electronics.repair Sergey Kubushyn <[email protected]> wrote:

Just took a brand spanking new WD5000AAKS drive out of sealed bag with
silica gel and all that stuff. The PCB is all _SILVER_ plated, no gold. And
that silver is almost totally black right out of the bag.

Not good. Silver really is unsuitable for modern, low voltage,
electronics. The last WD disk I bought (a WD5000AADS, 500GB
Caviar Green) had mixed gold and silver plating and the
silver plating was completely fine, on both sides of the
PCB.

Something is fishy here.

Arno

[...]
That suicide also can happen when some old file that was not accessed for
ages is read. That attempt triggers the suicide chain.

Yes, that makes sense. However you should do surface scans on
RAIDed disks regularly, e.g. by long SMART selftests. This will
catch weak sectors early and other degradation as well.

I know but I simply didn't think all 3 drives can fail... I thought I have
enough redundancy because I put not 2 but 3 drives in that RAID1... And I
did have something like a test with regular weekly full backup that reads
all the files (not the entire disk media but at least all the files on it)
and that was that backup that triggered disk suicide.

Anyway lesson learned and I'm taking additional measures now. It was not a
very good experience loosing some of my work...

Yes, I can imagine. I have my critical stuff also on a 3 way RAID1,
but with long SMART selftests every 2 weeks and 3 different drives,
two from WD and one from Samsung. One additional advantage of the
long SMART selftest is that with smartd you will get a warning
email on every failing test, i.e. one every two weeks. For additional
warning you can also run a daily short test, e.g..

No matter what you do you can not prevent an occasional disaster One
MUST remember that "backup" in not a noun but a verb in imperative.

BTW, I took a look at brand new WDC WD5000YS-01MPB1 drives, right out of the
sealed bags with silica gel and all 4 of those had their contacts already
oxidized with a lot of black stuff. That makes me very suspicious that
conspiracy theory might be not all that crazy--that oxidation seems to be
pre-applied by the manufacturer.

Urgh. These bags are airtight. No way the problem happened on your
side then. My two weeks old WD5000AADS-00S9B0 looks fine on the top
of the PCB. I think I will have a look underneath later.

Those 4 were fine on the top of PCB. Black stuff was underneath, on those
pads contacting with springy heads pins.

 

[Sergey Kubushyn]

I just removed the PCB from a WD200EB-75CSF0 20GB drive. Tin plating
on both the PCB contacts and the mating pin array going into the HDA
assembly. No gold or silver.

However, there's one extra item that hasn't entered the discussion.
There was a sheet of foam something between the PCB and the HDA. In
this case, it was quite clean and dry, but it wouldn't take much for
most any liquid to get absorbed by the foam and rot out the board.
There was a cut-out hole around the connector area, so there was no
common points of contact, but it was close enough that any volatiles
would certainly enter the connector area.

20Gb sounds like an old drive. Those were made before this "invention."

My brand spanking new drive was purchased from Amazon 4 days ago and the
manufacturing date is something like a month back. I bought it for a router
I'm building for my recently acquired /28 CIDR IP block on Intel D510MO
Athom board and Intel dual gigabit ethernet adapter (BTW, I highly recommend
that motherboard--it is a small miracle in a huge heap of manure selling
these days.) That means it is _ABSOLUTELY_ the "latest and greatest," not
something dug out of old junk heap. And it is retail.

That means it definitely comes like this right from the manufacturer. I have
serious doubts that it got that tarnished by a mere negligence. They
supposed to have clean rooms there and those drives are packed in airtight
bags. I have a lot of silver plated parts at home in usual zip lock bags and
even without any bags and they don't get this black in years less for a mere
month...