war story (spoiler: dying power supply takes disk drive with it)

D. Hugh Redelmeier hugh-pmF8o41NoarQT0dZR+AlfA at public.gmane.org
Tue Jul 26 05:54:39 UTC 2011


[This war story might be boring.  But some day you might have to
follow my path.]

Remember this thread?

| From: Lennart Sorensen <lsorense-1wCw9BSqJbv44Nm34jS7GywD8/FfD2ys at public.gmane.org>
| Date: Wed, 6 Oct 2010 18:01:30 -0400
| Subject: [TLUG]: Why not to buy cheap power supplies
| 
| Here is a "500W" power supply being tested at 420W load.
| 
| http://www.youtube.com/watch?v=VCwgK_fvZ6I
| 
| It was the second one they tried just in case the first one was a fluke.

One of my MythTV boxes made a funny noise when I powered it down one day 
last week.  I powered it down via GUI, nothing as crude as a mechanical 
switch.

When I tried to power it up, it wouldn't do anything.

I have a cheap power supply tester.  It said that the power supply wasn't 
supplying power any longer.

Note: this is not a cheap power supply.  It is a roughly five year old
Antec SmartPower PS-450.  But some others have problems with these
too: <http://www.badcaps.net/forum/showthread.php?t=4791>

I replaced the power supply with a CoolerMaster RP-650-PCAR that I
happened to have in inventory.

Still no joy.  When I turned power on, the CPU fan would jiggle once but 
there was no other sign of life.

When I unplugged all the peripherals, power seemed to work.

A process of elimination showed that the problem came up only when I
connected one of my Seagate 7200.11 drives.

It seems that the power supply death agony caused the 12V TVS
(<http://en.wikipedia.org/wiki/Transient_voltage_suppression_diode>) to
blow.

I was able to get the disk working again (and the data recovered) by
removing the dead TVS.  See
<http://forum.hddguru.com/help-with-disconnecting-tvs-from-seagate-barracuda-7200-t15313.html>

To get at the TVS, I needed to remove the drive's PC board.  That
required a TORX T6 tool which is moderately hard to find.  I got one
at Lowes.

I measured the resistance (DVM on 200 ohm scale) across each TVS and
found it to be close to 0 on the 12V TVS.

I poked at the 12V TVS with a fingernail and it came off.  I guess that that
means that the circuit board was slightly fried there.  But the drive
now works.

I don't know if the TVS ever protects without sacrificing itself.  I
imagine that it does.  So I should probably rescue the data and treat
the drive as junk / unreliable.

Since the power supply warranty is only 3 years, I probably cannot get
anything from Antec.  Since the power supply killed the disk drive
(and I did an operation on the drive), I don't see the drive being
covered by the Seagate warranty (the drive was purchased on the last
week that Seagate offered 5 year warranties on OEM drives).

But I'm mad.  I think that Antec should have recalled the power supply
when they knew of the bad capacitor problem.  (I haven't broken the
PS's seal so I'm not sure that this is a bad cap problem.)  Time for a
class action suit.  But I'm lazy.

Anybody had experience with hidden warranties from Antec?
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists





More information about the Legacy mailing list