Samsung Spinpoint T Series: Successor to a Quiet Legacy

Want to talk about one of the articles in SPCR? Here's the forum for you.
Devonavar
SPCR Reviewer
Posts: 1850
Joined: Sun Sep 21, 2003 11:23 am
Location: Vancouver, BC, Canada

Samsung Spinpoint T Series: Successor to a Quiet Legacy

Post by Devonavar » Sun Sep 03, 2006 5:57 pm


rpsgc
Friend of SPCR
Posts: 1630
Joined: Tue Oct 05, 2004 1:59 am
Location: Portugal

Post by rpsgc » Mon Sep 04, 2006 12:12 pm

Glad to see Samsung back on the game. Too bad they didn't do that with the P120 :lol:

qviri
Posts: 2465
Joined: Tue May 24, 2005 8:22 pm
Location: Berlin
Contact:

Post by qviri » Mon Sep 04, 2006 12:35 pm

I must say I prefer the cut-away shape of the P120, though. :? Makes it look less boxy and rectangular and in general boring.

Glad to hear there's two good options for high capacity now... :D

Would Samsung not have an advantage since all T-series are quiet, as opposed to older WDs with the same product number as the one reviewed by SPCR but apparently louder?

alfred
Posts: 127
Joined: Sun Jul 02, 2006 8:32 pm

Post by alfred » Mon Sep 04, 2006 1:40 pm

Many thanks for this great review, Devon!

As I described in _this_ topic (several posts), my WD3200KS, manufactured June, 2006, has higher-than-expected seek noises, even after two months of daily use. It sits in a Scythe enclosure, lying on soft foam without any direct contact to the case, and this case has damped side panels too; nevertheless I still clearly hear its seek noises during normal WinXP use provided a quiet ambient noise level, not even talking about defrag; but we clearly couldn't describe it as defective. To the contrary, my HD401LJ is exceptionally low on vibrations. Another very good example of sample variance effects : my Samsung T133 is way better than my WD-SE16, while your T133 and WD5000KS drives seem to be almost on a par. Obviously no one can say for sure cause we won't be buying 30+ identical drives in order to calculate average values, but I guess your WD5000KS was one of the best considering sample variance, while your HD400LJ seems quite bad on vibrations; my WD3200KS probably is below-par for this model, provided it can be compared to the WD5000KS.

Mats
Posts: 3044
Joined: Fri Dec 26, 2003 6:54 am
Location: Sweden

Post by Mats » Mon Sep 04, 2006 5:01 pm

This model showed up last year with two capacities available, 300 and 400 GB. Samsung have added a 320 and 250 GB models lately. At first I though that maybe the 320 GB possibly have two platters? I was wrong though, according to the specs. But I still think it's a bit strange that they introduce this capacity, especially since the bigger 400 GB already was there (and not like when WD's biggest disk were 320 GB).

About the noise of the latest WD disks. So far I've read more positive comments about the 500 GB version than for the one with 320 GB. Yes I know, there's no logic (no pun intended) in that. :?

alfred
Posts: 127
Joined: Sun Jul 02, 2006 8:32 pm

Post by alfred » Tue Sep 05, 2006 1:47 am

Yes Mats I'm starting to believe the current WD3200KS really has noisier seeks than the WD5000KS.

Back to the Samsung T133 topic -- I guess Samsung is planning to stop manufacturing P120 series as I don't see any other good reason to include 250, 300 & 320 GB in the T133 series. Nevertheless the 300 & 320 GB capacities are somewhat dull considering 133 GB platters; marketing rules probably.

specular
Posts: 13
Joined: Mon Jun 05, 2006 2:25 pm

Post by specular » Tue Sep 05, 2006 3:42 pm

Do any current Samsung drives have what Seagate calls an "offline scan"? Basically, some Seagate drives will, when they've been idling for a while, start scanning the drive "to ensure quality", thereby making very audible seek noises. It's really irritating.

If the T series doesn't do this, that would be great. Thanks in advance.

andyb
Patron of SPCR
Posts: 3307
Joined: Wed Dec 15, 2004 12:00 pm
Location: Essex, England

Post by andyb » Tue Sep 05, 2006 4:44 pm

The Samsung T-Series dont have an "offline scan" feature, and as such stay remarkably quiet when idle.

My 400GB T series is the quietest of my 4 HDD's, my drive's vibration is quite low (due to sample varience no doubt), and this drive is the best non-laptop drive I have ever heard.


Andy

JazzJackRabbit
Posts: 1386
Joined: Fri Jun 18, 2004 6:53 pm

Post by JazzJackRabbit » Tue Sep 05, 2006 5:25 pm

Good to see samsung putting another quiet hard drive out there. I'm not really bothered by seeks, but after listening to recordings the idle noise seems noticably lower than WD500 and only slightly louder than P120/BIV.

BTW, newegg still has it for only $140 shipped.

spolitta
Posts: 317
Joined: Thu Oct 13, 2005 5:44 pm
Location: Earth

Post by spolitta » Tue Sep 05, 2006 7:57 pm

Does anybody know how the overall speed is compared to SE16?

alfred
Posts: 127
Joined: Sun Jul 02, 2006 8:32 pm

Post by alfred » Tue Sep 05, 2006 11:32 pm

spolitta, I ran a few benchmarks I described in the topic I linked in my previous post; you'll find results here. These can easily be compared to benchmarks of the WD SE16 that has been widely reviewed online. The HD400/401LJ Samsungs perform well and are on a par with most other current 7200rpm drives. I'd say a +/-3% gap in performance territory shouldn't change how much you like your computer, while a +/-3dB gap in acoustic behavior surely will -- unless you've got 35dBA/1m fans or PSU obviously :wink:

In order to minimize hdd's noise when it's meant to stay idle, I configure each partition as this in Windows XP : disable indexing service; disable hard disk optimizing when idle (in tweakUI); disable System Restore; configure pagefile according to your software using habits.

Mats
Posts: 3044
Joined: Fri Dec 26, 2003 6:54 am
Location: Sweden

Post by Mats » Wed Sep 06, 2006 5:34 am

alfred wrote:disable hard disk optimizing when idle (in tweakUI)
I didn't know that you could do that, do you have any more info, or a link? Is this function called like that in Windows, or what should I search for in the registry? Thanks.

specular
Posts: 13
Joined: Mon Jun 05, 2006 2:25 pm

Post by specular » Wed Sep 06, 2006 9:12 am

alfred, the Seagate issue is a fairly well known problem and occurs even when you turn all those Windows services off. Believe me, I've tried all those things. The scan is built in to the drive's firmware.

alfred
Posts: 127
Joined: Sun Jul 02, 2006 8:32 pm

Post by alfred » Wed Sep 06, 2006 11:29 am

Mats : http://www.microsoft.com/windowsxp/down ... rtoys.mspx
This old free tool still is very useful.
Image

specular -- Yes I know this issue, I have it too on two WD drives. I had a similar thing on some old IBM/Hitachi drives, too; way louder than on 2005/2006 WD drives.

Devonavar
SPCR Reviewer
Posts: 1850
Joined: Sun Sep 21, 2003 11:23 am
Location: Vancouver, BC, Canada

Post by Devonavar » Wed Sep 06, 2006 12:47 pm

JazzJackRabbit wrote:I'm not really bothered by seeks, but after listening to recordings the idle noise seems noticably lower than WD500 and only slightly louder than P120/BIV.
Wow. There's definitely some subjective differences in our hearing and / or our sound systems. My listening identifies the B-IV as the quietest by far, followed by the T Series, P120 and WD500 all at the same level. The only differences I hear between these three are differences in noise quality; I can't hear any difference in volume.

whiic
Posts: 575
Joined: Wed Sep 06, 2006 11:48 pm
Location: Finland

"platter" and "bowl"

Post by whiic » Thu Sep 07, 2006 12:29 am

Thanks for a nice review of T-series Samsungs.

I do not work for Samsung, but I do have an idea what Samsung might mean about "bowl" and "platter" type designs. Probably it refers to the shape of the base casting of the HDD. Classic marketing images of drives without top covers are probably the best way to visualize the difference.

P80/P120 "platter" type base casting:
http://www.computeronlineshop.net/image ... 216684.jpg
The base casting is a flat piece of metal. Where as base casting is flat, the top cover is bowl shaped cast metal piece.

T133 "bowl" type base casting:
http://www.computeronlineshop.net/image ... 234035.jpg
The base casting has side walls and looking from any of the sides, all internal components would be hidden behind it. The top cover is a thin metal plate.

But I do agree they make it sound more confusing than what it actually is. (Considering the drives with flat base have a bowl shaped cover and the ones with a bowl shaped base casting have a flat cover. So either one could be the "bowl" or the "platter". But most likely they refer to the shape of base casting and nothing else.)

____

Also the "The T133-Series moved from the traditional platter-type design to the bowl type design to improve shock & vibrational performance" sounds a bit dubious as the main motivation has probably been to reduce manufacturing cost. No doubt is a simpler construction (adopted by all other manufacturers a long time ago) cheaper to manufacture.

But as long as making it simpler doesn't make things worse, I don't object to changes. All that was lost was just the unique appearance of Samsungs.

Also the rubber feet that are present in P-series under the casing were removed from T133. Probably also to reduce manufacturing cost. Those rubber feet are useful when mounting the HDD by it's bottom screws. (Not many cases support it, but it's usually possible to mount a drive on the bottom of the computer case. Some drilling of extra holes is required though.)

Shining Arcanine
Friend of SPCR
Posts: 502
Joined: Sat Oct 23, 2004 2:02 pm

Post by Shining Arcanine » Sat Sep 09, 2006 6:45 pm

Mike, I do not think that your statement that "The noise measurements place the Spinpoint T slightly better than either the Western Digital or Samsung's previous drives" is entirely accurate, because the WD5000KS is a 4 platter drive, while the HD400LJ is a 3 platter drive, which means that the WD5000KS by nature has 1/3rd more platter surface area than the HD400LJ, which due to friction with the air should translate into roughly 1/3rd greater noise generation, making a noise comparsion between the WD5000KS and the HD400LJ, an apples to oranges comparsion.

From an acoustical standpoint, an apples to apples comparsion would place the HD400LJ against either the WD3200KS (assuming that Western Digital is not yet using 160GB platters for WD3200KS) or the WD4000KS (assuming that Western Digital is using 160GB platters for the WD4000KS). Mathematically speaking, since the HP400LJ has 3 platters and that doubling it would result in 6 platters and a 3 decibel increase in noise, adding another platter should result in a one decibel increase in noise. Acoustically speaking, such a consideration would place Western Digital's previous drives as being slightly better than Samsung's T series.

Devonavar
SPCR Reviewer
Posts: 1850
Joined: Sun Sep 21, 2003 11:23 am
Location: Vancouver, BC, Canada

Post by Devonavar » Sat Sep 09, 2006 7:06 pm

I think you're misreading the article. The sentence you quote is not a general statement about WD and Samsung drives in general, but to the specific drives that we've tested in the past. I would draw you attention to this paragraph:
Our test drives were compared against our reference drives, the Seagate Barracuda IV and Samsung Spinpoint P80, which are profiled in our methodology article. To get a good idea of where the drives in this review stand, it is important to read the methodology article thoroughly. It was also compared against our current low-noise champ: A 500 GB Western Digital WD5000KS. A 250 GB Spinpoint P120 was also included in the comparison.
Thus, I refer to the Western Digital, and "Samsung's previous drives" refer to the P80 and P120 samples that we've seen. I claim that the measurements place it slightly better because we measured those drives at 21 dBA@1m, and the Samsung T at 20 dBA@1m. I never claimed to be making an apples to oranges to pomegranates comparison. I am directly comparing numbers that I measured, and reporting how they differ. I then take pains to point out that the tiny difference that I measured was not audible in practice.

I might add that a 1 dBA difference is probably not significant — it's too close to the margin of error. I don't think it's signficant.

I might also add that your reasoning seems suspect. By your logic, a four platter drive should be 12 dBA louder than a one platter drive, since it has four times as many platters. We know this isn't the case... The relationship between noise and the number of platters is not linear, and even if it was, you wouldn't be able to apply that assumption to noise measurements, which are logrithmic.

Shining Arcanine
Friend of SPCR
Posts: 502
Joined: Sat Oct 23, 2004 2:02 pm

Post by Shining Arcanine » Sun Sep 10, 2006 5:09 am

My logic said that doubling the platter count increased the noise by 3dB. So a four platter drive should be 6dB louder than a 1 platter drive. Conversely, cutting it in half should make the drive 3dB quieter. My logic also states that it would take a 16 platter drive to increase noise production by 12 dB.

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Sun Sep 10, 2006 7:13 am

Shining Arcanine wrote:My logic said that doubling the platter count increased the noise by 3dB.
If doubling the platters also doubled the number of motors and bearings, 3 dB would be the correct figure. But the number of motors remains one, and neither does the number of bearings increase. So, from an engineering standpoint, your logic is faulty. Sorry! :lol:

Tibors
Patron of SPCR
Posts: 2674
Joined: Sun Jul 04, 2004 6:07 am
Location: Houten, The Netherlands, Europe

Post by Tibors » Sun Sep 10, 2006 7:21 am

And from the average silencers standpoint the number of platters is irrelevant. He just cares about capacity and noise.

Shining Arcanine
Friend of SPCR
Posts: 502
Joined: Sat Oct 23, 2004 2:02 pm

Post by Shining Arcanine » Sun Sep 10, 2006 1:45 pm

Felger Carbon wrote:
Shining Arcanine wrote:My logic said that doubling the platter count increased the noise by 3dB.
If doubling the platters also doubled the number of motors and bearings, 3 dB would be the correct figure. But the number of motors remains one, and neither does the number of bearings increase. So, from an engineering standpoint, your logic is faulty. Sorry! :lol:
The amount of surface area increases by a factor of two when the platters are doubled. The motor primarily generates vibrations (due to the FDB) so when a drive is suspended, the motor noise is insignificant compared to the noise generated by the motion of the platters.
Tibors wrote:And from the average silencers standpoint the number of platters is irrelevant. He just cares about capacity and noise.
The number of platters is directly proportional to the amount of noise generated by hard drives with fluid dynamic bearings in any given hard drive family, making platter count an extremely important consideration when selecting drives within any given family that uses fluid dynamic bearings. Platters do not have frictionless surfaces, there are many elevations on their surfaces and when they spin, these elevations collide with the air, generating noise. In a typical 3.5" 7200 RPM drive, the outer edges of each platter move at approximately 75 mph.

If SPCR really disregards the platter count as a factor in noise generation, would SPCR be so kind to review mutiple drives with different platter counts from a single family (e.g. the Western Digital Caviar family) and determine the acoustical, vibration and energy properties of each one? If the hypothesis that the number of platters is irrevelent is correct, there would be no acoustical difference between the drives. If the hypothesis that the number of platters is directly proportional to the amount of noise generated by hard drives is correct, then there would be measureable differences between various platter counts, with a three decibel difference between two and four platter drives.

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Sun Sep 10, 2006 3:24 pm

You are placing the onus on SPCR to test your (rather unusual) theory of disk noise vs platters. Since this is your theory, isn't it up to you to provide some factual basis for your theory?

If SPCR's personnel and resources were at the beck and call of every nut with a weird theory, SPCR would never get anything useful done.

The ball is firmly in your court! :D

Devonavar
SPCR Reviewer
Posts: 1850
Joined: Sun Sep 21, 2003 11:23 am
Location: Vancouver, BC, Canada

Post by Devonavar » Mon Sep 11, 2006 2:26 am

Shining Arcanine wrote:If SPCR really disregards the platter count as a factor in noise generation,
We don't. But, we also don't buy your theory that noise is directly proportional to the number of platters. Proportional in some way, yes, directly, not.
Shining Arcanine wrote:would SPCR be so kind to review mutiple drives with different platter counts from a single family (e.g. the Western Digital Caviar family) and determine the acoustical, vibration and energy properties of each one?
I'd love to. You try convincing PR people that you want to review one of every model, not just their flagship.
Shining Arcainine wrote:If the hypothesis that the number of platters is irrevelent is correct, there would be no acoustical difference between the drives. If the hypothesis that the number of platters is directly proportional to the amount of noise generated by hard drives is correct, then there would be measureable differences between various platter counts, with a three decibel difference between two and four platter drives.
See these two reviews:

Seagate 7200.9 500 GB (A four platter drive)
Seagate 7200.9 160 GB (Same family, one platter)

The difference in noise level at idle is 3 dBA@1m (24 vs. 21 dBA@1m)... half of what your theory predicts. The primary acoustic difference in the presence of a high pitched whine in the 500 GB version, not air turbulence.

I would be very, very surprised if you can find a line of hard drives where the noise level jumps 3 dBA@1m between the single and double platter versions. The jump from two to four platters is more plausible to me, but I still think you'd only be able to find specific instances, not a general rule.

***

I should also add that SPCR reviews hard drives individually; we do not consider one drive a review of the entire line. Yes, it would be nice to test every drive exhaustively, but we are at the mercy of time and what drive manufacturers are willing to send us.

I see little point in trying to level the playing field using number of platters as a criteria. What matters is how loud a drive is, not how loud it is per platter. If anything, noise per gigabyte is a more valid metric, but even then I don't it as important for everyone, since not everyone cares about whether they have a 50 or a 500 GB drive. The only criterion that matters to SPCR is noise. I don't care if it has one platter or ten, so long as it's quiet.

Shining Arcanine
Friend of SPCR
Posts: 502
Joined: Sat Oct 23, 2004 2:02 pm

Post by Shining Arcanine » Mon Sep 11, 2006 3:02 am

Devonavar wrote:See these two reviews:

Seagate 7200.9 500 GB (A four platter drive)
Seagate 7200.9 160 GB (Same family, one platter)

The difference in noise level at idle is 3 dBA@1m (24 vs. 21 dBA@1m)... half of what your theory predicts. The primary acoustic difference in the presence of a high pitched whine in the 500 GB version, not air turbulence.
This is precisely why I did not mention Seagate. This principle might have applied to the original Barracuba IV, but the more recent ones are relatively noisy, such that much more noise is coming from the motor and bearings than I would expect in a Western Digital or Samsung hard drive.

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Mon Sep 11, 2006 4:45 am

If anything, noise per gigabyte is a more valid metric
I thought about that, but the problem with that is that very large drives (ie 750GB etc) will have artifically good dB/GB figures, although in real life they are probably noisier than many SPCR's are willing to tolerate (ie 30dB(A)+ at all times); so SPCR can deal with that by warning people about the increased noise and heat of very large drives, so that ultimately they can decide whether they are willing to make the trade-off.

More platters = more noise is a useful rule of thumb, but like all such rules there are exceptions to it and I don't think an effort to formulate a quantitative version of the rule will help SPCR do what it does best, ie alerting people to quiet components.

whiic
Posts: 575
Joined: Wed Sep 06, 2006 11:48 pm
Location: Finland

Post by whiic » Tue Sep 12, 2006 4:16 am

Shining Arcanine wrote:The amount of surface area increases by a factor of two when the platters are doubled. The motor primarily generates vibrations (due to the FDB) so when a drive is suspended, the motor noise is insignificant compared to the noise generated by the motion of the platters.

The number of platters is directly proportional to the amount of noise generated by hard drives with fluid dynamic bearings in any given hard drive family
HDD noise (even with FDB motor) is more than just air turbulence. It's bearing whine/grind (no longer present when there's no metal/ceramic ball bearings), motor coil whine (changes of current in the coil causes it's core to change shape repeatedly and produce high frequency audible noise), whine produced by electorics on the PCB, etc.

What you're saying is that wind noise is the only form of noise FDB drives produce, yet when performing subjective evaluation, we might find that some drives (even with FDBs) produce high-pitch noises and pure tones. These cannot be produced just by air turbulence which is quite broadband.
Shining Arcanine wrote:If SPCR really disregards the platter count as a factor in noise generation, would SPCR be so kind to review mutiple drives with different platter counts from a single family (e.g. the Western Digital Caviar family) and determine the acoustical, vibration and energy properties of each one?
Do you consider WD Caviar a single family of drives? No way. WDs are the least predictable drives when it comes on drive "families".
1) you take two drives manufactured the same month, like WD5000 and WD3200/2500 or WD3200/2500 and WD800 they're not the same family! They don't even look the same.
2) you take two drives of same name (like WD1200BB) but manufactured different months. One might be 1 platter, one might be 2 platter, and the oldest probably 3 platters. Also the base casting might change during the several year of manufacturing and I can guarantee there's nothing common with todays WD400 and the WD400 when it was introduced very long time ago.

WD is the least predictable when it comes to telling the number of platters used without actually opening the top cover and counting them visually. Maybe using a scale could help... but the differences in weight isn't that big and the weight of an "empty" HDD without platters vary between different families (and "WD800" for example isn't a "family" but several families).

Seagates aren't predictable either. Their "families" also consists of several data densities and even capacity points have different samples ranging from clipped head to short-stroked.

The product family concept has become more vague lately with even Hitachi combining different densities in their "family" T7K500. T7K500 250GB for example is closet to T7K250 250GB than T7K500 320GB.
Shining Arcanine wrote:If the hypothesis that the number of platters is irrevelent is correct, there would be no acoustical difference between the drives.
There is a relation. But the relation isn't necessarily as simple as direct propotionality (as you suggested). If we deny your theory of direct propotionality, you say the consequence of denying it would mean there's no reliation at all. That's BS.
Shining Arcanine wrote:If the hypothesis that the number of platters is directly proportional to the amount of noise generated by hard drives is correct, then there would be measureable differences between various platter counts, with a three decibel difference between two and four platter drives.
Yes... IF.

toto
Posts: 15
Joined: Fri Aug 11, 2006 2:11 pm
Location: New York

Post by toto » Wed Sep 13, 2006 6:39 am

A few weeks prior to this review, I purchased the HD300LD (300GB PATA version of this drive) to replace an 80GB Spinpoint P80 (same as the one used for comparison).

Out of the box, my perception was that the P80 was much quieter in seek mode, but using the SilentDrive AAM (software) Tool on the quietest setting brought the HD300LD in line with the review's conclusions about the HD400LJ.

FYI, both the 80P and HD300LD drives were enclosed in the same Zalman ZM-2HC2 Heatpipe HDD Cooler which was mounted in the bottom of the same insulated mid-tower case. With AAM enabled, it's an extremely quiet drive.

Michael Sandstrom
Posts: 606
Joined: Wed Oct 13, 2004 4:03 pm
Location: Albany, GA USA

Post by Michael Sandstrom » Wed Sep 13, 2006 4:26 pm

For several days I have been running a Samsung HD400LJ with a Nidec motor purchased from Newegg. The new drive replaced a Samsung SP1614C with a JVC motor. I have been very pleased by the noise and heat characteristics of both Samsung drives. I can notice no acoustic difference between the two drives. Both are extremely quiet. Also, DTemp reports identical 37 degree temperature without a dedicated fan.

My system is a good test bed for comparison. The case is an Antec 3700BQE which sits on my desk two feet from my head. Other than the HD (which rests on Sorbothane) the sources of noise are a 92mm Nexus, a 120mm Nexus and a 80MM Japanese Panaflo. All fans are running at 5 volts.

Thank you SPCR and fellow forum members for always providing valuable guidance.

Cehtna
Posts: 3
Joined: Wed Sep 15, 2004 11:48 pm

HD501LJ Samsung Spinpoint T166 500GB S-ATA II 16MB NCQ NG?

Post by Cehtna » Thu Feb 08, 2007 8:38 am

What about this harddrive:

- HD501LJ Samsung Spinpoint T166 500GB S-ATA II 16MB cache, 7200RPM, NCQ, Noise Guard?

Does this have the same noise characteristics as the T133 reviewed here?
The T166 500GB is cheaper per GB!

Which would you pick:
- 500GB T166 Samsung? /
- 400GB T133S Samsung? / http://www.silentpcreview.com/article657-page1.html
- 500GB SE16 Western Digital? / http://www.silentpcreview.com/article617-page1.html

Post Reply