Re: Noisy Topfield TF5800
- From: Johnny B Good <jcs.computers***@xxxxxxxxxxxxxxxx>
- Date: Thu, 31 Jan 2008 18:35:35 GMT
The message <gemini.jviiz20087nyk022g.willie@xxxxxxxxxxxxxxxx>
from News Will <willie@xxxxxxxxxxxxxxxx> contains these words:
"Dave Saville" <dave@xxxxxxxxxxxxxxxxx> wrote:
On Thu, 31 Jan 2008 11:17:29 UTC, Willie <williestott@xxxxxxxxxxxxxx>
wrote:
After a year of two of virtual silence, the HD on my TF5800 has started
making a noise. You can't really hear it if you're watching something,
but it noticable throughout the day.
It's 'normal' HD reading/writing noise on a computer.
I was wondering if the HD was dying - we do record a lot and the
Topfiled is on from 9am to about midnight with the HD running.
Ta
There are PSU problems. If you have a meter get the top of and check the
voltage at the molex. The PSU is right on its limits and any slight drop
causes all sorts of disk related problems.
Thanks.
How much of a drop are we talking about then, 0.1v, 0.2v, 0.5v?
Could I power it externally?
The design brief for a PC PSU (and the TF5800 is PC technology) is that
it should maintain its various voltage rails within +/- 5% of their
nominal voltage.
Otoh, the voltage _requirements_ of the various chips and other
components used in a PC are based on a +/- 10% tolerance.
Provided the +12v is no lower than 11.4v and no higher than 12.6v and
the +5v is no lower than 4.75v and no higher than 5.25v, the PSU is
within tolerance and shouldn't cause any problem.
There may be other voltage rails present (-12v, -5v, +5vsb and 3.3v)
which might be more tricky to access. These should also be within the
+/- 5% tolerance range. However, one wouldn't expect to see problems
until the +/- 10% tolerance range of the powered devices is exceeded.
The problem with using a voltmeter to make such measurements
(particularly when a digital meter is used), is that short term dips or
excessive ripple content due to ageing capacitors will not be detected,
so it is quite possible that voltage readings within the 5% tolerance
range can still be obtained even in the presence of short term dips
below the 10% device minimum (possibly agravated by excessive ripple
content).
However, having said all that, a noisy hard drive is more likely to be
a hard disk issue (possibly related to overheat conditions) rather than
a PSU one.
I've only ever had the top cover off the one time to identify the 250GB
drive and don't recall seeing a cooling fan. I was extremely surprised
to discover that the drive was a standard 7,200rpm Spinpoint unit rather
than the 5,400rpm versions which Samsung make specially for just such
equipment (where even 5,400rpm is 'overkill' performancewise with a
suitably optimised OS to allow the simultaneous recording of two channel
streams and the playback of a third stream from the archive).
Although the measired [1] mains consumption is a mere 25 watts active,
there is still a risk of overheat damage to the hard disk if the unit is
sat on top of other heat producing equipment or its vent slots
obstructed.
The upper limit on hard drive temperature is either 55 or 60 deg
celcius and this can quite easily be approached or even exceeded if
inadequate cooling provision has been made in the design of the box.
I don't believe the TF5800 monitors the SMART data to monitor drive
temperature (possibly on account such monitoring would be disruptive to
the design function of the drive as a streaming data store), but if you
are concerned, you could always attach the drive to a PC and boot from
the drive manufacturer's hdd diagnostics disk (CD or Floppy) and check
the SMART logs. In any case, you could run the non-destructive
diagnostics tests and verify whether or not there is trouble brewing (as
well as isolate the drive from what might be the real noise source
elsewhere in the TF5800 box).
It's just possible there might even be a TAP for checking drive
temperature (or the SMART log data), I haven't looked at the TAPs list
for nearly a year so couldn't say, but it might be worth a look.
[1] Using a real watt meter, not a cheap digital unit.
--
Regards, John.
Please remove the "ohggcyht" before replying.
The address has been munged to reject Spam-bots.
.
- References:
- Noisy Topfield TF5800
- From: Willie
- Re: Noisy Topfield TF5800
- From: Dave Saville
- Re: Noisy Topfield TF5800
- From: News Will
- Noisy Topfield TF5800
- Prev by Date: Re: Noisy Topfield TF5800
- Previous by thread: Re: Noisy Topfield TF5800
- Index(es):