LinHES Forums http://forums.linhes.org/ |
|
Noisy Hard drive Seagate ST750640AS http://forums.linhes.org/viewtopic.php?f=2&t=16998 |
Page 1 of 1 |
Author: | Algenon [ Wed Oct 17, 2007 1:15 am ] |
Post subject: | Noisy Hard drive Seagate ST750640AS |
I have recently upgraded from R5A30.2 to R5F27 in order to utilise the larger SATA drives, specifically I wanted to get 750GB capacity without having to resort to LVM. I purchased a Seagate ST3750640 SATA drive and the first thing I noticed was how much more disk access R5F27 seems to require compared with R5A30.2 for live TV. Its not too much of a problem until you simultaneously record and playback (then the hard drive head is thrashing about a lot). I put the drive in a Smart Drive 2002C enclosure to try to quieten it down (and cool it down). Having read the forum and in particular http://mysettopbox.tv/phpBB2/viewtopic. ... hard+drive I am wondering if I should have gone for a Western Digital drive instead and before I change it out, does anyone else have a similar experience with Seagate drives and noise? Algenon |
Author: | alien [ Wed Oct 17, 2007 2:53 am ] |
Post subject: | |
http://www.silentpcreview.com/article29-page1.html is a good place to start. The problem is that as HDs get bigger, it is harder to keep them quiet. I have a Western Digital Caviar SE16 WD5000KS now (2nd on the list). It is noticeably louder than the Samsung P120 SP2004C (used to top the list, but is no longer available) that it replaced. |
Author: | Algenon [ Wed Oct 17, 2007 10:54 am ] |
Post subject: | |
So its not that Seagate drives or SATA drives are necessarily noisy, its the size of them? Would I be better off with say two 350GB's, or is it possible to move the live TV recordings to another drive? On R5A30.2, the live TV used to be on /cache, but that seems to have gone now on R5F27. Algenon. |
Author: | tjc [ Wed Oct 17, 2007 5:38 pm ] |
Post subject: | |
Well the Seagate drives are simply noisier too. It used to be the opposite. When I bought the two 250Gb drives in my KM production box the Seagate drives were substantially quieter than the equivalent WD drives. When I added the 500Gb backup drive, WD had switched over to fluid dynamic bearings and theirs was the quietest available. It's still a bit louder than the older 250Gb Seagates, but much less so than equivalent newer 500Gb Seagates. |
Author: | grante [ Wed Oct 17, 2007 7:32 pm ] |
Post subject: | |
alien wrote: The problem is that as HDs get bigger, it is harder to keep
them quiet. Why is that? It's not like the platters are getting more massive or there are more of them. I can see why a 3.5" 7200RPM drive might be louder than a 2.5" 5400RPM disk (less mass spinning at a lower speed). But why would a drive with a higher areal density be noisier than a drive with a lower areal density if they're otherwise mechanically similar? |
Author: | tjc [ Wed Oct 17, 2007 8:34 pm ] |
Post subject: | |
Sounds like a question fro Silent PC Review http://www.silentpcreview.com/article29-page1.html Which refers to this white paper. http://www.seagate.com/support/kb/disc/tp/acoustics.html Other things being equal (number of platters, rotational speed) the biggest difference I'd expect from a higher density drive is the need for shorter sharper track to track seeks meaning higher radial acceleration and deceleration of the head. |
Author: | grante [ Wed Oct 17, 2007 10:43 pm ] |
Post subject: | |
tjc wrote: Sounds like a question fro Silent PC Review
http://www.silentpcreview.com/article29-page1.html Which refers to this white paper. http://www.seagate.com/support/kb/disc/tp/acoustics.html Other things being equal (number of platters, rotational speed) the biggest difference I'd expect from a higher density drive is the need for shorter sharper track to track seeks meaning higher radial acceleration and deceleration of the head. I don't understand why higher accelerations would be required if inter-track distances were reduced. If track-to-track seek times are maintained, wouldn't shortening the distance to be moved in a fixed amount of time reduce the required acceleration? Reductions in R/W head mass should also reduce the amount of noise generated (though head mass is probably negligible compared to the rest of the R/W arm assembly). From what I've gathered from those links, as drives have gotten larger over the past few years, they've generally gotten quieter also -- due mainly to a switch to fluid dynamic bearings, but also to increased attention to acoustic design. Neither link seems to support the claim that more bits means louder. Anecdotally, I've noticed that the last handful of drives drives I've bought (which were at least 10X the capacity of the ones they replaced) were far quieter than the old ones. In my experience (and according the references) bigger (physically) and faster means louder. |
Author: | alien [ Thu Oct 18, 2007 1:18 am ] |
Post subject: | |
The technology is definitely getting better, so an old small drive is typically louder than a big new drive (remember what HDs sounded like when 40M was big????). However, I think there is also a tendency for "cutting edge" drives to be loud. Right now (and this will change), reasonably priced 750+ Gig HDs tend to be loud. Drives <=500G can be found that are not loud. Two years ago, quiet 500G drives couldn't be found. I am generalizing here. Don't go by drive size, silent claims on the box or even how it sounds in the store. If you have your mythtv box next to your home entertainment system, you probably have a different standard for "quiet" then typical desktop users. When selecting a quite drive, first check a reputative reviewer such as silentpc review. It wouldn't hurt to check a few sources. There is nothing worse than installing a new drive in your living room only to find out it sounds like a tractor (as happened to me with my 160G WD several years ago). |
Author: | tjc [ Thu Oct 18, 2007 7:11 pm ] |
Post subject: | |
Odd as it seems shorter moves can be more energetic and thus noisier than longer ones. It has to do with the rate at which you're putting energy into the system (most actuators have a top speed at which they coast during longer moves), and the vibrational frequencies generated. (Human hearing is far more sensitive to some frequencies than others.) Don't get me wrong, in general drives have gotten quieter over the years, but there are gains and losses within the general trend. Seagate drives have gotten noisier in the range from the very quiet .6 family through the rather noisy .10 family because they stopped focusing on it. (Vague recollection says that white paper is from ~2002 when they were focused on it.) |
Author: | Algenon [ Fri Oct 19, 2007 11:34 am ] |
Post subject: | |
Its an interesting debate gentlemen. Is there a way to limit the thrashing about that the head does by mounting the live tv images on another, say, IDE drive? I don't know where the live tv files are stored now in R5F27. Algenon |
Author: | Human [ Fri Oct 19, 2007 12:15 pm ] |
Post subject: | |
I think Western Digital publishes noise specifications for all their drives. It seems to me that if a drive vendor does not publish noise specifications, it's likely that low noise is not a priority for them. Their 1TB Caviar GP drive averages between 25 and 27dB in various modes, and their enterprise-class RE2 500GB SATA drive averages between 28 and 33dB (up to 34dB for the 750GB model). When choosing a drive, you need to find the intersection of noise, reliability, and cost that works for you. |
Author: | grante [ Fri Oct 19, 2007 12:52 pm ] |
Post subject: | |
Algenon wrote: Its an interesting debate gentlemen. Is there a way to limit the thrashing about that the head does by mounting the live tv images on another, say, IDE drive? If you mean separate from the normal recordings, then no (see below). The best ways to limit head "thrashing" is to first tell the driver code you care more about noise than about speed using hdparm. I presume KM does that. The next thing to do is to make sure you've got plenty of RAM so that swapping isn't contributing to the problem. KM only requires about 256MB to avoid swapping, but 512MB or more will allow more disk blocks to be cached (avoiding disk accesses). Quote: I don't know where the live tv files are stored now in
R5F27. They're stored the same place that normal recordings are. Live TV is "just another recording" that you happen to be watching as it's being recorded. It's possible that using a second drive for / and /usr would somewhat reduce head movement some of the time. But, that second drive is going to add niose and heat, so I doubt there's a net gain. |
Author: | ik632 [ Mon Oct 22, 2007 11:03 am ] |
Post subject: | |
I was running two Hitachi 200gb pata drives and those are known for the periodic cat whine that they make. My wife started getting a kick out of it because the two drives would "meow" at different times and our cat started meowing back ![]() I have one of the 320gb Seagate drives in my box now and it's very quiet. |
Author: | Algenon [ Mon Oct 22, 2007 12:20 pm ] |
Post subject: | |
I have no problems with drives whining (I don't have a cat, but my daughters have a hamster!). hdparm has many options, which ones do I choose. I note that DMA is enabled already/not applicable to SATA drives. When I run hdparm /dev/sda -v, I get: Code: alan@mythtv2:~$ hdparm -v /dev/sda
/dev/sda: IO_support = 0 (default 16-bit) readonly = 0 (off) readahead = 256 (on) geometry = 25665/255/63, sectors = 1465149168, start = 0 alan@mythtv2:~$ Are there any recommended hdparm settings for SATA? Algenon. |
Page 1 of 1 | All times are UTC - 6 hours |
Powered by phpBB® Forum Software © phpBB Group http://www.phpbb.com/ |