Samsung single-platter 250GB drive
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
Samsung single-platter 250GB drive
I've read in the german magazin c't about a new Samsung drive, which they tested: 250GB, single platter, very quiet (among the quietest they ever tested), very fast (the fastest SATA drive they had so far: up to 103MB/s).
Sounds very promising. Tomorrow I'll post more numbers.
Sounds very promising. Tomorrow I'll post more numbers.
Just looked into the Samsung website:
HD250HJ
4.8W seek typical (<-apply a grain of salt here)
3.2W idle (<- same here)
5400 rpm
and measured:
transfer rate max.: >100MB/s (source: c't, can't recall the avg/min values)
power consumption: typical 4.low_something (measured by a good friend of mine)
Price per GByte: ~0,215€
Conclusion:
I have to get maybe two of these.
HD250HJ
4.8W seek typical (<-apply a grain of salt here)
3.2W idle (<- same here)
5400 rpm
and measured:
transfer rate max.: >100MB/s (source: c't, can't recall the avg/min values)
power consumption: typical 4.low_something (measured by a good friend of mine)
Price per GByte: ~0,215€
Conclusion:
I have to get maybe two of these.
Are you sure that it is 5400rpm? Different sources claim it being 7200rpm; looking at Samsungs site, all other S250 (and S166) series HDDs are 7200rpm ones too.
And one thing from their site, what I can't understand (my english is poor, I know, but anyway):
Seek (typical) 4.8 W
Read/Write (typical) 3.0 W
Idle (typical) 3.2 W
How can R/W power be less than Idle power???
And one thing from their site, what I can't understand (my english is poor, I know, but anyway):
Seek (typical) 4.8 W
Read/Write (typical) 3.0 W
Idle (typical) 3.2 W
How can R/W power be less than Idle power???
It was just one source I read which said 5400rpm. Maybe it's up to 7200rpm with a throttle to 5400rpm?
More data:
Read min/avg/max: ~54/~78/~103 MB/s
Noise idle/rw/seek: 0.2/0.3/0.4 sone (measured in free air, 50cm distance if I remember it right)
It looks like two of them will live in my P180 soon and replace these ugly IDE drives
In that test they also wrote that this drive is the first consumer drive to be faster than PATA allows, by the way.
The thing about idle power is really strange imho. Maybe they mixed something up here. But as I said before, a friend of mine measured between 4.0 and 4.3 during normal usage (web browsing, gaming, office, Windows Vista )
More data:
Read min/avg/max: ~54/~78/~103 MB/s
Noise idle/rw/seek: 0.2/0.3/0.4 sone (measured in free air, 50cm distance if I remember it right)
It looks like two of them will live in my P180 soon and replace these ugly IDE drives
In that test they also wrote that this drive is the first consumer drive to be faster than PATA allows, by the way.
The thing about idle power is really strange imho. Maybe they mixed something up here. But as I said before, a friend of mine measured between 4.0 and 4.3 during normal usage (web browsing, gaming, office, Windows Vista )
-
- *Lifetime Patron*
- Posts: 618
- Joined: Mon May 23, 2005 7:05 am
- Location: State College, PA
from the Samsung site spec sheet it's a 5400rpm drive, though the HD251+2HJ models are 7200rpm, one with double the rated power draw, 1 with the same as the HD250HJ. Confusing or what? Going through Samsung's HDD webpage with all the differet product codes made my head hurt
-
- Posts: 2049
- Joined: Thu Dec 15, 2005 11:06 am
- Location: Klamath Falls, OR
The same spec sheet's average latency sez this is a 7200RPM drive. Sez it loud and clear!mattthemuppet wrote:from the Samsung site spec sheet it's a 5400rpm drive...
-
- *Lifetime Patron*
- Posts: 618
- Joined: Mon May 23, 2005 7:05 am
- Location: State College, PA
-
- Posts: 2049
- Joined: Thu Dec 15, 2005 11:06 am
- Location: Klamath Falls, OR
3 places offer this on-line in the U.S. of A. All call it a 7200RPM drive. Lowest price so far just under $90 but none of the "regular retailers" are on that short list.
edit: more info
"Target application" is "laptops PCs and SFF PCs" (explaining the low power, but why aren't all HDDs low power??). Also model HD162HJ, 160G capacity, has the identical media-to-cache spec of 158MB/sec, meaning the same linear bit density per track. Apparently this smaller model is for disk media that cannot support that bit density toward the center of the drive. Noise is identical, specs identical except no "5400RPM" typo, none available yet for sale here in the U.S. of A.
edit: more info
"Target application" is "laptops PCs and SFF PCs" (explaining the low power, but why aren't all HDDs low power??). Also model HD162HJ, 160G capacity, has the identical media-to-cache spec of 158MB/sec, meaning the same linear bit density per track. Apparently this smaller model is for disk media that cannot support that bit density toward the center of the drive. Noise is identical, specs identical except no "5400RPM" typo, none available yet for sale here in the U.S. of A.
Please dont tell me we have another "WD Green Power" episode here - but in reverse this time.
Samsungs own spec sheet says 5400rpm, but it also says 4.17ms seek time (as Felger says 7200rpm), 8.9ms seek is also a standard 7200 rpm measurement, but then we come to the power usage which is more 5400 rpm than it is 7200rpm, then comes the noise which is more 7200rpm than 5400rpm, then we look at the people who are selling it, they ALL say 7200rpm.
And OCUK have them in stock.
http://www.overclockers.co.uk/showprodu ... =HD-051-SA
Who to believe....... only a review will tell, but I expect Samsung got it wrong by accident unlike WD who love to lie to their customers on purpose bastards.
Andy
Samsungs own spec sheet says 5400rpm, but it also says 4.17ms seek time (as Felger says 7200rpm), 8.9ms seek is also a standard 7200 rpm measurement, but then we come to the power usage which is more 5400 rpm than it is 7200rpm, then comes the noise which is more 7200rpm than 5400rpm, then we look at the people who are selling it, they ALL say 7200rpm.
And OCUK have them in stock.
http://www.overclockers.co.uk/showprodu ... =HD-051-SA
Who to believe....... only a review will tell, but I expect Samsung got it wrong by accident unlike WD who love to lie to their customers on purpose bastards.
Andy
The spec sheet now says:
But since there's F1 series now, why look at the lower density S250 ? S250 does have very low power consumption though.7,200 RPM class. Actual speed can be different a little.
Has anyone got hold of the HD250HJ yet? I'm seriously looking at this as a replacement for my 3yr old Maxtor DM10 200Gb.
I've been looking at the F1 and GP's too but they're still very expensive here.
My last hard drive purchase was a WD5000AAKS a few months ago, and it's been very disappointing - seeks not too bad with AAM on, but it's not a quiet drive. I'm starting to wonder if any 3.5" hard drive can be considered truly quiet?!
I've been looking at the F1 and GP's too but they're still very expensive here.
My last hard drive purchase was a WD5000AAKS a few months ago, and it's been very disappointing - seeks not too bad with AAM on, but it's not a quiet drive. I'm starting to wonder if any 3.5" hard drive can be considered truly quiet?!
-
- Posts: 2049
- Joined: Thu Dec 15, 2005 11:06 am
- Location: Klamath Falls, OR
Depends on how you mount it. I put mine in a DIY rubber box with 3/16" side thickness, along with a spot-cooling fan. The air intake and exhaust are both oriented toward the rear of the case. Truly quiet. Honest.AndrewD wrote:I'm starting to wonder if any 3.5" hard drive can be considered truly quiet?!
I recently got six of them, of which I put two into our server. The raid 0+1 now puts out up to 205 MB/s . The other two disks were put into my P180 as a replacement for a Samsung 120 GB drive and a WD 200 GB drive (these were bought before the silencing virus hit me, so no exact names here). Now I can only hear occasional seek noise over two tuned-down Fanders and a S12II-430. The drives are not specially suspended, I'm just using the P180's lower drive bay with the silicone thingies.AndrewD wrote:Has anyone got hold of the HD250HJ yet? I'm seriously looking at this as a replacement for my 3yr old Maxtor DM10 200Gb.
Edit: of course I put four drives in the server to have a raid 0+1 And the "thingies" in the P180 are of course made of silicone.
Last edited by mexell on Mon Jan 14, 2008 1:19 am, edited 1 time in total.
It sounds interesting. Would you mind taking a picture of your DIY rubber box? I can provide web space for public if you need them.Felger Carbon wrote: Depends on how you mount it. I put mine in a DIY rubber box with 3/16" side thickness, along with a spot-cooling fan. The air intake and exhaust are both oriented toward the rear of the case. Truly quiet. Honest.
-
- Posts: 2049
- Joined: Thu Dec 15, 2005 11:06 am
- Location: Klamath Falls, OR
You'll find the first, and most complete, explication of my box complete with CAD drawing and pics about halfway down page 7 on the "show your suspension" sticky, which is now in the disk drive forum.loimlo wrote:It sounds interesting. Would you mind taking a picture of your DIY rubber box?
Edit2: Here's the basic idea:
Later models have gotten smaller and I now use a 1-HDD version with a 60mm spot-cooling fan. I took a cardboard "proof-of-cooling" prototype, added 3/32" McMasters rubber, and am using it right now as I type this in my #1 computer. You can see pics of this version in my Nexus & Logisys review, posted recently in the Users Review forum.
I've published pics of other versions here in SPCR from time to time. I haven't kept this thing a secret!
Edit: Since page 7 of the sticky is currently unavailable, here's excerpts and pics from an old posting:
For a long while now, I've been using a big rubber box holding two HDDs, closed on 5 sides with an opening on the rear, so any noise gets directed toward the back of the computer case. This is written up, with pics, on page 7 of the "show your HDD suspension" sticky (now in this forum). The rubber provides outstanding damping for seeks and vibration in general. I was suspending these boxes using Stretch Magic, and on one computer I still am. Here's 3 of the larger boxes, and a newcomer:
The newcomer is fabricated using cardboard and superglue, and is a test fixture for cooling. It allows adjustment of the air channel height above the HDD, which is mounted upside-down. Turns out 4" wide by 2/3" high is about optimum (the 4" is fixed by the HDD width). This box uses an 80mm GW NCB fan; it has the interesting property that if you listen from the rear, the motor bearing noise goes away! At 6.6V, the fan turns 1160RPM and the temp rise above room ambient is 14C. At 5V, the fan turns 800RPM (9.7dBA quieter calculated) and the temp rise is 17C, so 3C higher temp saves almost 10dBA fan noise.
This is based on a Hitachi chinese-designed 7K160 SATAII HDD that idles at a specified 5.9W.
I just built a second test box using a quiet 60mm fan, a SilenX from SVC.com that's specified at 1500RPM +/- 15%, at 12V. (Why SilenX? Do you know anybody else who makes a low-RPM 60mm fan?) I measured 1940RPM. No problem, because the fan starts at 6V (specified 8V). At ~9V (four .71-.72V diodes in series with the fan header) the fan turns 1420RPM with a temp rise of 15C. At 12V I measured 14C, so I'm continuing the testing, with 7V my next test. (With the fan turned off, the temp rise was 25C above room ambient.)
Here's the initial design:
When I actually built the box, I decided to use thinner cardboard, and to apply 3/32" rubber to the cardboard to block (attenuate) the noise:
The difference between the 80mm and 60mm fan version? The small fan costs $10 more and the box is 0.8" shorter. If space is an important consideration, the extra cost is worth it!
Thanks mexell - sounds like the 250HJ is a promising option. The 500GB GP's seem to have just arrived in Aust, so it looks like it's between these 2 until the F1 320GB is available. I'm just not sure if 1 platter @ 7200rpm vs 2 platters @ 5400rpm is going to be quieter .
Felger - I usually put my drives on foam at the bottom of the case. The DM10 doesn't vibrate much, so even a thin piece of foam is enough, but it has an annoying higher frequency whine. An enclosure might help, but since the drive is getting old, I'd rather replace it with a quieter one first.
Felger - I usually put my drives on foam at the bottom of the case. The DM10 doesn't vibrate much, so even a thin piece of foam is enough, but it has an annoying higher frequency whine. An enclosure might help, but since the drive is getting old, I'd rather replace it with a quieter one first.
-
- Posts: 2049
- Joined: Thu Dec 15, 2005 11:06 am
- Location: Klamath Falls, OR
This drive is now in stock at Newegg. I just placed an order.
-
- Posts: 2049
- Joined: Thu Dec 15, 2005 11:06 am
- Location: Klamath Falls, OR
My HD250HJ just walked in the door. It sez "(250GB/7200rpm/8M)" in largish, dark print. It sez "product of China" in small, light print and in a different place "SAE Magnetics Ltd/China" also in small light print.
I also bought a ~$20 toy Scythe "wireless" infrared thermometer, which does not work wireless but does work when plugged into USB using the enclosed cable. It does come with a CR2032 battery which may be dead. The accuracy spec is +/- 2C which is why I call it a toy. Dunno what the angle of measurement it (the narrower the better, but on this toy probably pretty wide). I plan to use this for close measurement of the motor section of the HD250HJ and another HDD to see if the low power spec seems reasonable, so the angle isn't critical for this measurement. (I don't know where the SMART temp sensor is located, and if it's on the PC board I won't trust the reading, which is why the infrared thermometer).
The HDD was from Newegg and the toy from Jabtech.
I also bought a ~$20 toy Scythe "wireless" infrared thermometer, which does not work wireless but does work when plugged into USB using the enclosed cable. It does come with a CR2032 battery which may be dead. The accuracy spec is +/- 2C which is why I call it a toy. Dunno what the angle of measurement it (the narrower the better, but on this toy probably pretty wide). I plan to use this for close measurement of the motor section of the HD250HJ and another HDD to see if the low power spec seems reasonable, so the angle isn't critical for this measurement. (I don't know where the SMART temp sensor is located, and if it's on the PC board I won't trust the reading, which is why the infrared thermometer).
The HDD was from Newegg and the toy from Jabtech.
-
- Posts: 2198
- Joined: Thu Feb 10, 2005 11:20 am
- Location: TN, USA
FWIW, Samsung and Western Digital should have a 320GB single platter drive out this year (as in any day now).
HD322HJ
http://wdc.com/en/company/releases/Pres ... B19F6C62FF}
SPCR doesn't like the URL on the WDC site so you'll have to copy/paste or choose from the list of press releases.
HD322HJ
http://wdc.com/en/company/releases/Pres ... B19F6C62FF}
SPCR doesn't like the URL on the WDC site so you'll have to copy/paste or choose from the list of press releases.
-
- Posts: 2049
- Joined: Thu Dec 15, 2005 11:06 am
- Location: Klamath Falls, OR
Here's the label:
The HDD comes in the familiar transparent clamshell containing the HDD, installation instructions, and a package containing 4 mounting screws:
In the background there's an unconnected HDD used as an "ambient" target for the thermometer, then the HD250HJ soft-mounted outside the case and connected as SATA drive #2, and the toy thermometer.
In idle, the new HDD measures 14C higher than the "ambient" drive. When I copy a partition from my #1 SATA, a Hitachi 7K160 to the Samsung 3810MB copy in 1:31 (91 sec). When copying back to the 7K160, it took 1:38, using Ghost. This copying procedure "touches" the source disk once (read), and the destination disk twice (write, then read to verify). These numbers indicate the new Samsung is 25% faster than the 7K160, which is fine with me. I didn't buy the new drive for speed.
During the 90+ second copy, the Samsung drive gradually increased in temperature by ~1C in almost a minute, and then showed no further increase in temperature. Some SPCR folk have commented on the unusual HD250HJ power spec while seeking and writing being slightly less than when idle. I seem to detect very slightly more power. Pretty close to spec.
My 7K160, when not cooled by a spot fan, runs 25C higher than ambient. The 7K160 SATA is specified to idle at 6W. The Samsung idles 14C higher than ambient, which seems to confirm that its idle power consumption is indeed slightly over 3W.
It appears to me that a suspended HD250HJ would not need spot cooling, meaning the front panel air intake could be sealed to prevent noise from escaping the case innards thru the bezel.
In my "User Review" of the Nexus Frizzbee, I reported that I could hear - barely - the seeks of the 7K160. That was thru the bezel at a distance of 2 feet or more. When I was doing the 90+ second partition copies, the Samsung was outside the case, as the photo above shows. I put my ear within 9 inches of the HDD. I could hear nothing. Quiet day, no traffic, double glazed windows, nothing separating my ear and the Samsung but 9 inches of air.
I don't claim to be an expert evaluator of HDD noise. I'm just saying what I didn't hear. Perhaps some of you will share your opinions of the noisiness of the new HD250HJ. I was mainly interested in the power consumption, and I've proved to myself that the consumption is indeed very low and close to spec.
I've shown this mounting gadget before: a rectangle of lightweight styrene with four 1/4" foam pads on the top and four more on the bottom, providing 1/2" of soft foam HDD support.
edit: after solving two equations in two unknowns, the Samsung takes 28 seconds to handle 3810MB once, the Hitachi 35 seconds. So the new Samsung is 25% faster than the Hitachi, not ~10%. It seems I'm out of practice with my estimates...
The HDD comes in the familiar transparent clamshell containing the HDD, installation instructions, and a package containing 4 mounting screws:
In the background there's an unconnected HDD used as an "ambient" target for the thermometer, then the HD250HJ soft-mounted outside the case and connected as SATA drive #2, and the toy thermometer.
In idle, the new HDD measures 14C higher than the "ambient" drive. When I copy a partition from my #1 SATA, a Hitachi 7K160 to the Samsung 3810MB copy in 1:31 (91 sec). When copying back to the 7K160, it took 1:38, using Ghost. This copying procedure "touches" the source disk once (read), and the destination disk twice (write, then read to verify). These numbers indicate the new Samsung is 25% faster than the 7K160, which is fine with me. I didn't buy the new drive for speed.
During the 90+ second copy, the Samsung drive gradually increased in temperature by ~1C in almost a minute, and then showed no further increase in temperature. Some SPCR folk have commented on the unusual HD250HJ power spec while seeking and writing being slightly less than when idle. I seem to detect very slightly more power. Pretty close to spec.
My 7K160, when not cooled by a spot fan, runs 25C higher than ambient. The 7K160 SATA is specified to idle at 6W. The Samsung idles 14C higher than ambient, which seems to confirm that its idle power consumption is indeed slightly over 3W.
It appears to me that a suspended HD250HJ would not need spot cooling, meaning the front panel air intake could be sealed to prevent noise from escaping the case innards thru the bezel.
In my "User Review" of the Nexus Frizzbee, I reported that I could hear - barely - the seeks of the 7K160. That was thru the bezel at a distance of 2 feet or more. When I was doing the 90+ second partition copies, the Samsung was outside the case, as the photo above shows. I put my ear within 9 inches of the HDD. I could hear nothing. Quiet day, no traffic, double glazed windows, nothing separating my ear and the Samsung but 9 inches of air.
I don't claim to be an expert evaluator of HDD noise. I'm just saying what I didn't hear. Perhaps some of you will share your opinions of the noisiness of the new HD250HJ. I was mainly interested in the power consumption, and I've proved to myself that the consumption is indeed very low and close to spec.
I've shown this mounting gadget before: a rectangle of lightweight styrene with four 1/4" foam pads on the top and four more on the bottom, providing 1/2" of soft foam HDD support.
edit: after solving two equations in two unknowns, the Samsung takes 28 seconds to handle 3810MB once, the Hitachi 35 seconds. So the new Samsung is 25% faster than the Hitachi, not ~10%. It seems I'm out of practice with my estimates...
Great thread, given that single platter F1s or 320GB WDs are yet to come this drive looks interesting. Is this HD250HJ really the currrent holy grail of the silent pc enthusiast, with low noise/temps and yet good performance?
"When I copy a partition from my #1 SATA, a Hitachi 7K160 to the Samsung 3810MB copy in 1:31 (91 sec). When copying back to the 7K160, it took 1:38, using Ghost. This copying procedure "touches" the source disk once (read), and the destination disk twice (write, then read to verify)."
This means about 17% quicker than the Hitachi, or 91MB/s (average?) STR, if I got things right. Is a 5400rpm 250GB platter hdd really that fast? On paper it lists close to F1 transfer rates (158 vs 175 MB/s), but that's only on paper... Could someone post some benchmark numbers? Even better a hdtach or hdtune graph.
And if it is 7200rpm instead how can it get such amazing wattage? Either way looks uncanny.
"When I copy a partition from my #1 SATA, a Hitachi 7K160 to the Samsung 3810MB copy in 1:31 (91 sec). When copying back to the 7K160, it took 1:38, using Ghost. This copying procedure "touches" the source disk once (read), and the destination disk twice (write, then read to verify)."
This means about 17% quicker than the Hitachi, or 91MB/s (average?) STR, if I got things right. Is a 5400rpm 250GB platter hdd really that fast? On paper it lists close to F1 transfer rates (158 vs 175 MB/s), but that's only on paper... Could someone post some benchmark numbers? Even better a hdtach or hdtune graph.
And if it is 7200rpm instead how can it get such amazing wattage? Either way looks uncanny.