Bent pins

[George Macdonald]

Just did a new system this past week. This is my 6th Athlon64, 3rd with
s939 and I got a shock when I peered in the plastic window on the box of
this 3500+ (Venice) and couldn't see the CPU to check the model #. Then I
realized that the transparent clamshell containing the CPU was loose, down
in the "well" of the green preformed packing shell, under the platform
where it should cinch into the slots. The clamshell had opened and I could
see bent pins.

The tamper-proof seals on the box appeared intact, so this can only be a
packing problem at the assembly plant, unless... it was a counterfeit...
but more important, what to do?? After mulling it over, I really couldn't
face talking to NewEgg or AMD though I think I'd have had a good case for
replacement with the seals being undisturbed. Has anybody else seen
this?... just wondering if maybe their packing robots(?) were off whack
occasionally... a lot??

Anyway I opened the thing up and with some trepidation went about
straightening the bent pins - about 20 or so, with some touching their
neighbors. After trying small screwdrivers, tweezers etc., I finally fell
on the idea of using a set of feeler gauges to move them little by little
back to straight. Things worked out pretty well and after about 2 hours of
gentle coaxing, I got to the point where a .035" feeler had just a slight
amount of jiggle between the pins of every row in both directions.

It fit into the socket fine and everything seems to be working OK but it
seems to me that AMD really needs to improve their packaging procedures
and/or packing equipment. I find it hard to believe that there had been no
final inspection to be sure the CPU was sitting in its assigned place by
AMD or the vendor - the green protective packing shell has no signs of
damage and I don't see how the CPU could get from its normal position to
where it was by rough handling.

 

[Rob Stow]

Just did a new system this past week. This is my 6th Athlon64, 3rd with
s939 and I got a shock when I peered in the plastic window on the box of
this 3500+ (Venice) and couldn't see the CPU to check the model #. Then I
realized that the transparent clamshell containing the CPU was loose, down
in the "well" of the green preformed packing shell, under the platform
where it should cinch into the slots. The clamshell had opened and I could
see bent pins.

The tamper-proof seals on the box appeared intact, so this can only be a
packing problem at the assembly plant, unless... it was a counterfeit...
but more important, what to do?? After mulling it over, I really couldn't
face talking to NewEgg or AMD though I think I'd have had a good case for
replacement with the seals being undisturbed. Has anybody else seen
this?... just wondering if maybe their packing robots(?) were off whack
occasionally... a lot??

I've never seen bent pins that were not the user's fault. By now
I've handled a few thousand AMD chips and in the neighbourhood of
1000 Intels.

Anyway I opened the thing up and with some trepidation went about
straightening the bent pins - about 20 or so, with some touching their
neighbors. After trying small screwdrivers, tweezers etc., I finally fell
on the idea of using a set of feeler gauges to move them little by little
back to straight. Things worked out pretty well and after about 2 hours of
gentle coaxing, I got to the point where a .035" feeler had just a slight
amount of jiggle between the pins of every row in both directions.

Next time try a fine mechanical pencil. Discard the lead, of
course, then just slip the tip over a bent pin ...

It does occur to me that I've not had to do this for a long time
and even the finest mechanical pencil might now be too big.

It fit into the socket fine and everything seems to be working OK but it
seems to me that AMD really needs to improve their packaging procedures
and/or packing equipment. I find it hard to believe that there had been no
final inspection to be sure the CPU was sitting in its assigned place by
AMD or the vendor - the green protective packing shell has no signs of
damage and I don't see how the CPU could get from its normal position to
where it was by rough handling.

You choose not to deal with the vendor and get a replacement, and
hence the vendor has nothing to return to AMD for
repair/replacement. If they don't see the damaged goods they
can't do anything to fix whatever process caused the damage.

 

[George Macdonald]

I've never seen bent pins that were not the user's fault. By now
I've handled a few thousand AMD chips and in the neighbourhood of
1000 Intels.

What are you saying? The box was sealed, the CPU was loose in the cavity,

Next time try a fine mechanical pencil. Discard the lead, of
course, then just slip the tip over a bent pin ...

I knew about that "technique" but you can't see how far you're bending it -
worst thing is to have to bend it back after bending it too far.. and as
you note, the pins on s939 are "very fine". The feeler gauge works well
since you get alignment, you can feel if the gauge is too thick and go down
a size... and can't push too far.

It does occur to me that I've not had to do this for a long time
and even the finest mechanical pencil might now be too big.



You choose not to deal with the vendor and get a replacement, and
hence the vendor has nothing to return to AMD for
repair/replacement. If they don't see the damaged goods they
can't do anything to fix whatever process caused the damage.

The stated "vendor" policy is that CPUs are warranted through the mfr only.
I had only one CPU on hand and figured it was worth the risk to have a
working CPU in 2 hours vs. waiting for a week or more. I have reported the
"incident" to AMD Tech Supp with a description of where the CPU was: i.e.
dislodged (or never placed properly at the packing plant) from its proper
placement in the green packing shell.

According to what I read from others, there have been cases where the green
biodegradable paper/plastic composite material has broken or even
disintegrated, though that was not the case for me. AMD has work to do
here... apart from which a simple inspection would have shown that the CPU
top plate was not visible in the plastic window.

 

[keith]

What are you saying? The box was sealed, the CPU was loose in the cavity,
pins were bent. This must be the 1001st case.<shrug>

I *think* he's saying that you should have stuffed it back up the pipe
where it belonged. I certainly would have. First, had I paid the big bux,
I want the right stuff. Second, as I think I heard, the pipe doesn't
know it has a problem if everyone just grins. I'd think they'd want to
know they have a problem. Once the seals are broken, it's *your* problem.

I knew about that "technique" but you can't see how far you're bending
it - worst thing is to have to bend it back after bending it too far..
and as you note, the pins on s939 are "very fine". The feeler gauge
works well since you get alignment, you can feel if the gauge is too
thick and go down a size... and can't push too far.

....andother reson I'd ship it back, with prejudice!

The stated "vendor" policy is that CPUs are warranted through the mfr
only. I had only one CPU on hand and figured it was worth the risk to
have a working CPU in 2 hours vs. waiting for a week or more. I have
reported the "incident" to AMD Tech Supp with a description of where the
CPU was: i.e. dislodged (or never placed properly at the packing plant)
from its proper placement in the green packing shell.

If it's still in the sealed packaging, I can't see how they're going
to fault you. Tell us all who does this sort of crap, and we'll make em
pay! ;-)

According to what I read from others, there have been cases where the
green biodegradable paper/plastic composite material has broken or even
disintegrated, though that was not the case for me. AMD has work to do
here... apart from which a simple inspection would have shown that the
CPU top plate was not visible in the plastic window.

They can't do their work, unless they know they have work to do.

 

[Robert Redelmeier]

Just did a new system this past week. This is my 6th Athlon64,
3rd with s939 and I got a shock when I peered in the plastic
window on the box of this 3500+ (Venice) and couldn't see the
CPU to check the model #. Then I realized that the transparent
clamshell containing the CPU was loose, down in the "well"
of the green preformed packing shell, under the platform where
it should cinch into the slots. The clamshell had opened and
I could see bent pins.

I suspect a shipping problem. _Very_ bad things can happen.
Usually damage is evident from crushed corners/edges. But
sometimes not if it hits flat. That cannonball (HSF) packed
inside doesn't help -- absorbing it's energy will cause the
clamshell to flex.

The tamper-proof seals on the box appeared intact, so this can
only be a packing problem at the assembly plant, unless... it was
a counterfeit... but more important, what to do??

Like Keith, I'd send it back so they'd know they've got
work to do. You've got clear provinence.

If you didn't fix with mech pencil (available down to 0.3mm),
or old technical pen nibs (remove weight & wire) below that.

-- Robert

 

[Rob Stow]

What are you saying? The box was sealed, the CPU was loose in the cavity,
pins were bent. This must be the 1001st case.<shrug>

All I was saying is what /my/ experience has been. Can't comment
on yours or anybody else's.

Most of the time the "user" who bent the pins was me :-(
Butter fingers when the cpu is an inch above the socket is all it
takes.

I knew about that "technique" but you can't see how far you're bending it -
worst thing is to have to bend it back after bending it too far..

Shrug. Works for me. Lay CPU on a horizontal surface, bend
until pencil is vertical ..

and as
you note, the pins on s939 are "very fine". The feeler gauge works well
since you get alignment, you can feel if the gauge is too thick and go down
a size... and can't push too far.

Something I'll have to remember for next time.

The stated "vendor" policy is that CPUs are warranted through the mfr only.
I had only one CPU on hand and figured it was worth the risk to have a
working CPU in 2 hours vs. waiting for a week or more. I have reported the
"incident" to AMD Tech Supp with a description of where the CPU was: i.e.
dislodged (or never placed properly at the packing plant) from its proper
placement in the green packing shell.

According to what I read from others, there have been cases where the green
biodegradable paper/plastic composite material has broken or even
disintegrated, though that was not the case for me. AMD has work to do
here... apart from which a simple inspection would have shown that the CPU
top plate was not visible in the plastic window.

Could the damage have been done at the vendor's store ? The
last time I went to a "London Drugs" in Canada I saw one employee
toss boxed hard drives 20 feet to where another employee was
catching them and putting them on the shelves. I told the only
other employee in the store why I was never coming back and he
just shrugged it off. Missing a catch while restocking CPUs
could have done the kind of damage you saw.

 

[George Macdonald]

On Sun, 12 Jun 2005 20:18:35 -0400, George Macdonald wrote:


If it's still in the sealed packaging, I can't see how they're going
to fault you. Tell us all who does this sort of crap, and we'll make em
pay! ;-)

Look at www.newegg.com - plainly states that CPU is warranted through mfr.
After examining some of the boxes I've kept, I'm beginning to wonder about
the tamper-proof seals - maybe not as tamper-proof as hoped for.

They can't do their work, unless they know they have work to do.

As stated I've reported to AMD Tech Support -- what a PITA that is:
registration only worked with IE. Anyway we'll see if they get back on it.

 

[George Macdonald]

All I was saying is what /my/ experience has been. Can't comment
on yours or anybody else's.

OK I guess I *did* ask.

Shrug. Works for me. Lay CPU on a horizontal surface, bend
until pencil is vertical ..

Yeah I see that but the inside of the conical head of the pencil is err,
conical.

Could the damage have been done at the vendor's store ? The
last time I went to a "London Drugs" in Canada I saw one employee
toss boxed hard drives 20 feet to where another employee was
catching them and putting them on the shelves. I told the only
other employee in the store why I was never coming back and he
just shrugged it off. Missing a catch while restocking CPUs
could have done the kind of damage you saw.

I've examined the packing, of this and other AMD boxes, closely and don't
see how rough handling could have caused this. The green preformed shell
is undamaged and I don't see how it could distort inside the box
sufficiently for the CPU to come out of the slots it's cinched into. IMO
this is a packing plant robot error... or just possibly the tamper-proof
seals are not proofed enough for someone determined.

 

[George Macdonald]

I suspect a shipping problem. _Very_ bad things can happen.
Usually damage is evident from crushed corners/edges. But
sometimes not if it hits flat. That cannonball (HSF) packed
inside doesn't help -- absorbing it's energy will cause the
clamshell to flex.

Have you seen the current packing of a boxed Athlon64? The CPU is in a
small clear plastic clamshell with the pins embedded in foam; the closed
clamshell sits on a platform on a preformed composite plasticized paper(?)
(biodegradable) shell, where it is cinched into slots. I can't see how the
shell could distort enough, without cracking or breaking, for the clamshell
to come out of the slots.

Like Keith, I'd send it back so they'd know they've got
work to do. You've got clear provinence.

It was a toss-up - I'd have been really pissed if they'd come back and
said: "yup the seals had been umm, tampered". Others have also reported
that they have been offered a different, older core part in exchange by AMD
- I wanted this Venice core, I'd waited for it to become available retail
boxed and paid the extra $$. Fortunately all worked out.

The bottom line here is that s939 is kinda initimidating when it comes to
pin straightening but the pins are actually quite sturdy and it is still
doable.

 

[Robert Redelmeier]

Have you seen the current packing of a boxed Athlon64?

No, I have not had the pleasure.

The CPU is in a small clear plastic clamshell with the pins
embedded in foam; the closed clamshell sits on a platform on
a preformed composite plasticized paper(?) (biodegradable)
shell, where it is cinched into slots. I can't see how the
shell could distort enough, without cracking or breaking,
for the clamshell to come out of the slots.

Does this composite insert also carry the cannonball (HSF)?
If so, then when the package is dropped, inertial forces can
act in funny directions. The plasticising on the composite
can increase it's elasticity, or at least hide crease damage.

It was a toss-up - I'd have been really pissed if they'd come
back and said: "yup the seals had been umm, tampered".

But they're reputable and you _know_ you didn't tamper.

different, older core part in exchange by AMD - I wanted this
Venice core, I'd waited for it to become available retail
boxed and paid the extra $$. Fortunately all worked out.

Of course! I buy CPUs by the core they contain, not some MHz.

-- Robert

 

[Robert Redelmeier]

Yeah I see that but the inside of the conical head
of the pencil is err, conical.

Not on the pencils I've seen. There's a 5mm long essentially
straight pipe section for lead support.

IMO this is a packing plant robot error...

A very serious matter!

or just possibly the tamper-proof seals are not proofed
enough for someone determined.

Equally serious. It would mean that someone has invaded NewEgg's
supply chain from AMD. It's supposed to be reasonably secure!

-- Robert

 

[Ed]

As stated I've reported to AMD Tech Support -- what a PITA that is:
registration only worked with IE. Anyway we'll see if they get back on it.

Can you share a picture of it before/after you opened it and fixed the
chip?

Share pictures..
http://www.imageshack.us/
http://www.putfile.com/

btw, I would of sent it back for replacement. ;p
Ed

 

[George Macdonald]

Can you share a picture of it before/after you opened it and fixed the
chip?

Sorry - no, I don't have a digital camera... and I don't borrow my wife's.

 

[George Macdonald]

No, I have not had the pleasure.


Does this composite insert also carry the cannonball (HSF)?
If so, then when the package is dropped, inertial forces can
act in funny directions. The plasticising on the composite
can increase it's elasticity, or at least hide crease damage.

I guess nothing's impossible but the CPU would have to come out of its
slots, turn 90 degrees and fall down into the support cavity under its
platform, without the shell breaking. My bet is still on a robot
mishandling what is not a precision material.

 

[keith]

Look at www.newegg.com - plainly states that CPU is warranted through mfr.
After examining some of the boxes I've kept, I'm beginning to wonder about
the tamper-proof seals - maybe not as tamper-proof as hoped for.

I understad that. But if they *sold* you a defective part, it is up to
*them* to make it right. In a year, they're off the hook. The fact is
that they sold you crap. Send it back and tell them that your CC company
has been notified that the transaction is void.

As stated I've reported to AMD Tech Support -- what a PITA that is:
registration only worked with IE. Anyway we'll see if they get back on
it.

You're not listening. *FORGET* AMD. They're not ehte ones who sold you
the broken CPU. It may indeed be their problem, but it's _not_ yours.
Send the thing back and contest the CC charges as not having had
delievered what was offered for sale! ...at least I assume the
bent-pins part wasn't in the advert.

I'm 90% sure that NewEgg wouldn't bat an eyelash at your complaint.

 

[keith]

Sorry - no, I don't have a digital camera... and I don't borrow my wife's.

Come on George! It's a camera, not underware! ;-)