Time Capsule power supply early death
J.J. O'Shea
Capsules bought in the Feb-May 2008 timeframe. They can be either the 1 TB or
the 500 GB model (four 1 TB, two 500 GB so far) and in all cases they were
working just fine until they suddenly, usually overnight, stopped dead. As in
when the users went to bed the green light was on and the TC was working,
when they woke up the TC was a paperweight. In a few cases (one each of the 1
TB and 500 GB models) if the TC was unplugged and left alone for a while (24
hours seems to be enough) the unit will come back to life if plugged back
in... but will then die again inside a few days, this time without being able
to be resurrected.
We got 'em 'cause the local Apple Stores (all three of them, Boca Town
Centre, Wellington Green, and Gardens) won't fix 'em. If the users had
AppleCare, they will _replace_ them, but anything on 'em is gone for good.
We did a little research, and it appears that the power supply in original
model TCs goes tits-up in 16-18 months from go. It appears to be a
reproducible problem. (No data about the new TCs, 'cause it ain't been 18
months yet, but I would suspect that as the changes weren't in the power
supply, they're gonna go tits-up at about the same rate, too...) There are
two ways of fixing the problem:
1 as the problem is due to heat-related damage to the capacitors, swap out
the caps. There are three caps which are certain to need swapping and a
fourth which may/may not be okay. See
<http://sites.google.com/site/lapastenague/a-deconstruction-of-routers-and-
modems/apple-time-capsule-repair/apple-time-capsule-repair-type2>. (No, not
my site. He just did a good job at showing how to do it and I don't see why I
should reinvent the wheel.) Be sure to add the fan mod or you're gonna be
replacing the caps in another 18 months or so, and you may then also have to
replace the hard drive, killed by heat.
2 pull the old power supply and attach an external power supply.
<http://sites.google.com/site/lapastenague/a-deconstruction-of-routers-and-
modems/apple-time-capsule-repair>
There are several other methods of doing both of the above, but that site was
handy when I was writing this.
Of course, doing either of the above will absolutely void your warranty.
While you're knocking holes in your TC, you might consider adding an eSATA
port to it and increasing the available storage.
And, oh, just so you know... Time Machine _will_ back up to network devices
other than Time Capsules _if you disable the switch that Apple put into the
system to prevent this_.
<http://forums.macrumors.com/showthread.php?t=450628> The critical part is:
open your terminal and type:
defaults write com.apple.systempreferences TMShowUnsupportedNetworkVolumes 1
and now TM will be able to back up to _any network share that is formatted
HFS+_. (Sorry, NTFS or FAT volumes need not apply, and EXT3 or EXT4 volumes
are right out.)
This means that my handbuilt NAS cage containing 3 TB of storage in a RAID 5
array can be used with TM. See
<http://www.macosxhints.com/article.php?story=20080420211034137>
--
email to oshea dot j dot j at gmail dot com.
Geoffrey S. Mendelson
> and now TM will be able to back up to _any network share that is formatted
> HFS+_. (Sorry, NTFS or FAT volumes need not apply, and EXT3 or EXT4 volumes
> are right out.)
Is that true of EXT3 and EXT4 file systems shared with AFP?
Thanks, Geoff.
--
Geoffrey S. Mendelson, Jerusalem, Israel g...@mendelson.com N3OWJ/4X1GM
J.J. O'Shea
(in article <slrnhdmai...@cable.mendelson.com>):
> J.J O'Shea wrote:
>> and now TM will be able to back up to _any network share that is formatted
>> HFS+_. (Sorry, NTFS or FAT volumes need not apply, and EXT3 or EXT4 volumes
>> are right out.)
>
> Is that true of EXT3 and EXT4 file systems shared with AFP?
>
> Thanks, Geoff.
>
>
As fair as I know, yeah. Haven't done any experimenting on it. If you can get
EXT3/4 volumes to mount in TM, then a whole bunch of NAS systems can be used
out of the box to replace TC. I doubt it though, as I suspect that the sparse
bundle that TM uses requires HFS+ to work properly.
I may set up a Linux-based NAS one weekend (not today, for sure) and test it
out to see what happens.
Jolly Roger
J.J. O'Shea <try.n...@but.see.sig> wrote:
> as the problem is due to heat-related damage to the capacitors, swap out
> the caps.
Wait... is Apple still using bad caps from the capacitor plague?!
--
Send responses to the relevant news group rather than email to me.
E-mail sent to this address may be devoured by my very hungry SPAM
filter. Due to Google's refusal to prevent spammers from posting
messages through their servers, I often ignore posts from Google
Groups. Use a real news client if you want me to see your posts.
JR
Jolly Roger
J.J. O'Shea <try.n...@but.see.sig> wrote:
> 1 as the problem is due to heat-related damage to the capacitors, swap out
> the caps. There are three caps which are certain to need swapping and a
> fourth which may/may not be okay. See
> <http://sites.google.com/site/lapastenague/a-deconstruction-of-routers-and-
> modems/apple-time-capsule-repair/apple-time-capsule-repair-type2>. (No, not
> my site. He just did a good job at showing how to do it and I don't see why I
> should reinvent the wheel.) Be sure to add the fan mod or you're gonna be
> replacing the caps in another 18 months or so, and you may then also have to
> replace the hard drive, killed by heat.
I recently found myself replacing bad caps on my mom's old iMac G5
motherboard. It took a while, but I managed to find that Mouser is one
of the better places to get capacitors:
J.J. O'Shea
(in article <jollyroger-58D09...@news.individual.net>):
> In article <hbf8t...@news2.newsguy.com>,
> J.J. O'Shea <try.n...@but.see.sig> wrote:
>
>> as the problem is due to heat-related damage to the capacitors, swap out
>> the caps.
>
> Wait... is Apple still using bad caps from the capacitor plague?!
>
>
It's not that the caps are bad, it's that the TC internals are badly designed
and generates too much heat with too bad an airflow. The caps die from
overheating, the way that hard drives dies in certain Dell machines. It's got
to the point where if we get a call about a Dell that bluescreens on startup
we carry a replacement hard drive to the call 'cause odds are good that it's
the drive gone tits-up after too much heat. Of course the customer usually
doesn't have a back up and is _really_ pissed when our guy tells them that
all their stuff is gone...
And that's why its a good idea to run the fan mod on the TC as well as the
power supply mod, as if you don't it's gonna overheat anyway, and your hard
drive will go sooner or later.
Jolly Roger
J.J. O'Shea <try.n...@but.see.sig> wrote:
> On Sun, 18 Oct 2009 11:06:13 -0400, Jolly Roger wrote
> (in article <jollyroger-58D09...@news.individual.net>):
>
> > In article <hbf8t...@news2.newsguy.com>,
> > J.J. O'Shea <try.n...@but.see.sig> wrote:
> >
> >> as the problem is due to heat-related damage to the capacitors, swap out
> >> the caps.
> >
> > Wait... is Apple still using bad caps from the capacitor plague?!
>
> It's not that the caps are bad, it's that the TC internals are badly designed
> and generates too much heat with too bad an airflow.
How sure are you about that? The capacitors with faulty electrolitic
material that caused the bad caps plague failed in the same way with
temperatures *within* specifications. Has anyone actually measured the
temperatures to see if they are within specifications?
J.J. O'Shea
(in article <jollyroger-9BA49...@news.individual.net>):
> In article <hbfbl...@news4.newsguy.com>,
> J.J. O'Shea <try.n...@but.see.sig> wrote:
>
>> On Sun, 18 Oct 2009 11:06:13 -0400, Jolly Roger wrote
>> (in article <jollyroger-58D09...@news.individual.net>):
>>
>>> In article <hbf8t...@news2.newsguy.com>,
>>> J.J. O'Shea <try.n...@but.see.sig> wrote:
>>>
>>>> as the problem is due to heat-related damage to the capacitors, swap out
>>>> the caps.
>>>
>>> Wait... is Apple still using bad caps from the capacitor plague?!
>>
>> It's not that the caps are bad, it's that the TC internals are badly
>> designed
>> and generates too much heat with too bad an airflow.
>
> How sure are you about that? The capacitors with faulty electrolitic
> material that caused the bad caps plague failed in the same way with
> temperatures *within* specifications. Has anyone actually measured the
> temperatures to see if they are within specifications?
>
>
We're sure. The caps are all good caps from reputable sources and the TCs
were reportedly running hot prior to failure. Have a look inside a TC and
you'll see what I mean about the bad design and poor airflow. This is a
design problem, and there are going to be a _lot_ of dead TCs because of it.
If the TC is kept fairly cool (next to a fan, near an a/c outlet, something
to keep the temp down) it lasts longer. If it's warm or in direct sunlight it
dies faster.
Jolly Roger
Right but my point is "running hot" isn't the most scientific
measurement, and typically these caps are rated for up to 105 degrees
Celcius (221 degrees Fahrenheit). So they can withstand quite a lot of
heat. You're saying you think the heat levels are exceeding that - but
since we have no real measurements, that's just speculation. I'm also
speculating that it's possible these caps have faulty electrolytic
material that may be causing them to fail early in high temperatures
that are within spec.