We review two more quiet notebook hard drives, from Hitachi and Samsung. The Hitachi 80GB 5K80 and the Samsung 40G MP0402H are both extremely quiet, low vibration drives and not badly priced. They are affordable quiet alternatives (for desktop PCs) to silent solid state hard drives. We listen, measure and describe their acoustics and performance.
December 23, 2004 by Mike Chin with Jordan Menu
| Products | Hitachi Travelstar 5K80 80GB notebook drive Samsung MP0402H 40GB notebook drive |
| Suppliers | Hitachi Global Storage Technologies Frontier PC Vancouver, BC |
| Street Prices | Hitachi: 40GB ~US$80; 80GB ~US$150 Samsung: 40GB US$70~75; 80GB US$122~150 |
To get a PC any closer to silence than the 20-22 dBA/1m noise floor imposed by the quietest desktop hard drives, we have to look at other storage options. With the best low noise fans, it is possible to achieve as much as 20~25 cubic feet per minute at a noise level of <18 dBA/1m. However, the lowest noise achieved by any desktop hard drive has remained unchanged at 20~22 dBA/1m for several years. Suspension shock mounting such a desktop drive can certainly effect a significant subjective drop in noise for a system that’s otherwise quiet, but it is very difficult to progress much below 20 dBA/1m with a desktop HDD if something approaching silence is your goal.
The very quietest storage technology is a RAM-based solid state drive, whose parts only move at the level of electrons. But their four-figure $ prices are unreachable for mere mortal PC users. A much more affordable alternative is a quiet notebook drive. As I discovered while reviewing an unusual small fanless computer, the Mappit A4F, some new notebook computers are amazingly quiet. The process of examining and using notebook hard drives as desktop HDD replacements began for me after that review; it is how this series of notebook drive reviews was born.
Our third article about quiet notebook drives involves a Hitachi Travelstar 80G, 5400 rpm drive and a Samsung 40G, 5400 rpm drive. Let’s start with a few photos.
Hitachi 3.5″ desktop drive next to review subjects Hitachi Travelstar 5K80 and Samsung MP0402H.
Another David and Goliath picture: The weight difference is typically ~600g against ~100g.
The size and weight are fairly obvious, but the thermal and power issues are not quite as obvious. You may never have thought about it before, but there’s a very good reason why notebook drives are more energy efficient than the bigger desktop drives: They have to be to work well with battery-powered notebook PCs. The difference is big.
A typical 5400 rpm 2.5″ HDD with two platters and four heads takes about 5W to spin up and no more than about 2.5W(!!!) in normal operation. A typical 7200 rpm HDD with two platters and four heads takes about 20~34W to spin up. They take 8~14W on average to seek and write normally. If there is any doubt about these numbers, please check for yourself the PDF data spec sheets on 2.5″ and 3.5″ HDDs from Seagate, Hitachi, and Samsung.
Regular readers of SPCR know that suspending a hard drive for minimal noise usually causes a rise in temperature due to the lack of heat conduction between the HDD and the chassis. With a notebook drive’s trifling power dissipation, overheating is hardly ever a problem when suspending or even enshrouding as long as a small amount of airflow is somewhere nearby.
| High Startup Power of Desktop HDDs Some readers might be surprised about the high startup power figures; I know I was. Note that this power draw only lasts a couple of seconds while the HDD motor accelerates the rotational mass of the disks from stationary to 7200 rpm. But the duration is long enough that several HDDs can stress a PC power supply enough so that it fails to boot the system even though once booted up, the same PSU would have no trouble keeping the system running. Here are some HDD startup power requirements plucked from the tech spec documents.
| |||||||||||
Back to the two notebook drives on hand, one hiccup with the comparison is that the Hitachi is an 80GB model while the Samsung is a 40G. The 80GB model uses two platters and four heads while the 40GB model uses one platter and two heads. There is a small consequence in noise: The two disk models are generally a wee bit louder. Hitachi cites the noise difference between their 2 platter and 1 platter models to be 0.3 Bels, and Samsung cites the noise difference to be 0.2 Bels. It is small but audible. A 40G Hitachi model has the same idle noise spec as the Samsung 40G, and a slightly higher seek noise spec. Please keep this in mind as we continue with the comparison.
Top views…
…bottom views: Not much to differentiate them visually.
| HDD Model | Hitachi Travelstar 5K80 | Samsung MP0402H |
| Capacity / Cache / RPM | 80GB / 8MB / 5400 | 40GB / 8MB / 5400 |
| Latency / Avg. Seek | 5.5ms / 12ms | 5.6ms / 12ms |
| Power: Startup / Typical / Low power idle | 5W / 2.5W / 0.85W | 5W / 2.5W / 0.85W |
| Max Data Transfer Rate: Media – Buffer / Buffer- Host | 450 / 100 Mb/s | 431 / 100 Mb/s |
| Operating Temp | 0 – 55°C | 0 – 55°C |
| Acoustics, Sound Power: Idle / Read-Write (typical) | 2.5 / 2.9 Bel | 2.2 / 2.4 Bel |
There’s not much to differentiate these drives aside from the capacity and the noise. What do 0.3 and 0.5 Bel differences mean in real terms? It’s difficult to convert them to the dBA/1m SPL readings we use. It’s better for us to listen and measure.
TESTING
Our focus is first on the acoustics and vibration, and then second, on the performance. Readers who familiar with the hard drive specialist web site, Storage Review, know how difficult it is to obtain truly authoritative performance information about hard drive performance; it takes many different kinds of tests in a variety of conditions. We don’t pretend to be such experts, nor do we care quite as much. Our usual point of view is: Give me the quietest drive with the smallest hit in performance.
- Our performance test procedure consisted of running each of the test samples in turn off the secondary IDE channel of an extremely quiet reference PC.
- Our acoustic test procedure consisted of running each of the test samples in an external USB 2.0 drive enclosure (cover removed) placed in a very quiet room adjacent to where the test PC was located. The test PC was SPCR’s audio recording PC, a Shuttle Zen SFF PC modified to 16 dBA/1m noise level.
- The drive / USB enclosure was placed atop a soft piece of foam to reduce vibration effects, and angled up so that its top was facing the microphone of the sound level meter when noise measurements were done. The foam greatly reduced any low frequency resonance. As mentioned already, the cover of the USB enclosure was left off.
- The microphone was positioned 3″ above the drive for the audio recordings.
- Ambient room noise during testing was 16 dBA. The room itself is 20′ x 10′ x 8′ and carpeted, but not highly damped. The doors and windows were closed, and there were no other sources of noise in the room.
Testing Equipment
SPCR’s audio recording PC:
- Shuttle Zen ST62K SFF motherboard / case / heatsink / fanless power supply
- Intel P4-2.53 (Northwood) CPU
- 512MB DDRRAM (PC3200)
- Samsung MP0402H notebook hard drive
Other Audio & Test Equipment:
- THE KP- 6M reference omni microphone
- M-Audio Tampa mic preamp
- M-Audio Firewire 410 external digital sound interface
- B&K 2203 sound level meter
Observant readers will have noticed that the test PC actually employs a Samsung notebook hard drive of the same model as the one being tested. It is an obvious giveaway about the very low acoustic output of this drive. Here is a complete listing of all the two drives in this review, along with notebook drives reviewed thus far, and our reference quiet desktop 3.5″ 7200 rpm hard drives.
| 2.5″ HDD NOISE COMPARISON | |||
| Drive Model | Idle (dBA/1m) | Seek (dBA/1m) | Subjective Notes |
| Samsung MP0402H | 17 | 18 | The acoustics of this drive are virtually identical to the Fujitsu MHT2040AT, a considerably slower 4200 rpm drive and the quietest we’ve encountered. The Samsung is extremely quiet, and there is very little if any high frequency noise to speak of. It has minimal vibration, but placing it on soft foam does reduce low freq. noise audibly. The unit used in the test PC was suspended in elastic string and mostly surrounded by soft but dense foam. Seek noise is somewhat more audible than the 1 dBA gain suggests, but very soft. |
| Hitachi Travelstar 5K80 | 19 | 20 | The Hitachi comes very close to the Samsung, but has a slightly sharper and higher pitched sound, with perhaps a touch more vibration as well. The seek noise is a touch louder too. When inside even a very quiet desktop PC, the slightly higher noise level of this drive over the Samsung may not be audible. The performance is superior, according to SiSoftware Sandra 2005, and also subjectively. |
| Toshiba MK6022GAX | 22 | – | Slightly louder than the Seagate Barracuda IV single platter 3.5″ reference hard drive. The noise signature has the broadband shhhh quality exhibited by the Samsung SP 3.5″ drives, but higher in pitch, a bit like the Seagate. A trace of whine, but not like the Seagate Momentus. Seek noise is only moderately louder than idle, perhaps by 3 dBA. Vibration is higher than any of the 4200rpm drives; similar to the Momentus. Performance seems quite speedy, as it should be with 16 MB cache and 5400rpm, but inconsistent results with all the benchmarks tried stops me from publishing results. |
| Seagate Momentus ST94811A | 24 | – | The Momentus has a terrible constant “pure” tone somewhere in the 6~10KHz range. It drops 2-3 dBA in level when the listener or the mic faces the edge of the drive because of directionality of the high frequency whine. Seek noise is substantially higher, probably 3~5 dBA. Vibration is much lower than any 3.5″ drive, but higher than either of the 4200rpm drives tried. A real disappointment, but it did perform about as fast as or faster than the Seagate Barracuda-IV. |
| Fujitsu MHT2040AT | 16 | – | The only noise maker in the Mappit A4F PC, which seemed virtually inaudible to me. The noise is not inaudible, but very low and soft, easily dismissed in the ambient noise of all but the quietest spaces. There is no high pitched whine to speak of, and the seek noise does not seem more than maybe 2 dBA higher than idle. It is the slowest performer of all the drives here. Extremely low vibration. |
| Toshiba MK4025GAS | 16 | – | This 8 MB cache 4200 RPM drive offers better performance than 2 MB cache 4200 rpm drives, and it is identical in both idle and seek noise to the Fujitsu above. Extremely low vibration. |
| Seagate Barracuda IV ST340016A | 20 | – | In idle, it remains the quietest of all 3.5″ drives. This sample is almost 2 years old, but seems unchanged in noise. There may be a touch of high frequency whine but it is very low in level, and easily obscured when mounted in a PC case. Seek is considerably higher, possibly as much as 5~6 dBA. Low vibration, but MUCH higher than any of the notebook drives. |
| Samsung SP0802N | 21.5 | – | The idle noise is a touch higher, and its seek may actually be lower than the Seagate B-IV. Similar vibration level as the B-IV, but there are reports of some samples exhibiting much higher vibration levels. This is cured by HDD decouple mounting (suspension in elastic material or placement on soft foam), which is virtually mandatory for a truly quiet PC anyway. |
Below are sound recordings in MP3 format of the two notebook drives tested here, along with the two reference 3.5″ drives. Each recording is 20 seconds long; the first 10 seconds in idle, and the second half in seek, a state effected by running Windows drive defragmentation. They are presented in order of loudness.
|
|
| HOW TO LISTEN & COMPARE These recordings were made with a high To set the volume to a realistic level (similar to the original), try playing the Nexus 92mm recording above and setting the volume so that it is barely audible. Then don’t reset the volume and play the other sound files. Of course, tone controls or other effects should all be turned off or set to neutral. For full details on how to calibrate your sound system to get the most |
Samsung MP0402H 40G 2.5″ notebook HDD (17~18 dBA/1m)
Hitachi Travelstar 5K80 80G 2.5″ notebook HDD (19~20 dBA/1m)
Seagate Barracuda IV 40G 3.5″ desktop HDD (21~24 dBA/1m)
Samsung SP80 3.5″ desktop HDD (22~25 dBA/1m)
Nexus 92mm case fan @ 5V (17 dBA/1m)PERFORMANCE
Generally, most quiet drive users report negligible if any real practical difference in perceived desktop system performance between a fast 7200 rpm drive and a slower, quieter 7200 rpm drive. We know from experience that average random access or seek time has the greatest perceivable impact on HDD related desktop performance. Here we are looking at 5400 rpm notebook drives whose stated latency is nearly 2ms slower than typical 7200 rpm drives, and whose rated average seek time is at least 3ms slower than the fastest 7200 rpm drives. You would think that the performance hit would be quite significant; we have not found this to be the case.
Samsung MP0402H notebook drives are in use at SPCR not only in the audio recording PC described above, but also a couple of other workhorse PCs. They have been employed for at least three months. In that time, we’ve rarely, if ever, noticed any significant performance penalty with any of the work or activities conducted on these PCs. These tasks include:
- Web surfing
- General Windows tasks
- Office document creation / processing (Word, Excel, Powerpoint)
- Photoshop image processing of both very large and quite small images
- Review / creation of Adobe InDesign electronic publishing files
- DVD movie playback
- DiVX / MP3 playback
- Audio file processing (sometimes very large)
SiSoftware Sandra 2005 (10.37) was used to run the file system benchmark on these notebook drives.
Samsung MP0402H: A lazy 21 ms Average Access Time was recorded.
A huge 10 ms gain over the Samsung, but the Hitachi’s overall drive index is only 1MB/s better.
A Samsung SP80 80G 7200 rpm 3.5″ desktop drive fared the same as its smaller brethren, one Index point behind the Hitachi.
This despite Access Time on par with the Hitachi. This may not be representative of all Samsung SP80s.
The Hitachi notebook drive was used as an OS drive only for a few days. The general impression was that it is a bit faster than the Samsung notebook drive. Invariably, it is only after prolonged use that we feel confident to make broad statements about HDD performance, because its impact on different applications can vary. We will report in a future follow-up how the Hitachi fares in the context of a daily use lab system.
CONCLUSIONS
The Hitachi Travelstar 5K80 and Samsung MP0402H notebook drives are exciting discoveries for silent computing.
The Samsung drive is within a decibel of the quietest hard drive we’ve ever listened to or tested. Such a close measurement makes it difficult to give much weight to the difference; slight variances in testing conditions, procedure or samples might even flip the results. Yet unlike the 4200 rpm Fujitsu that it matches for noise, the Samsung is a faster 5400 rpm drive that, at least in practical use, gives little away in performance to bigger faster desktop drives.
The Hitachi Travelstar 5K80 also turns in a great showing, with noise lower than the quietest 3.5″ hard drive, and even better performance (at least on the SiSoftware Sandra benchmark) than our particular Samsung SP80 7200 rpm desktop drive sample. A single platter 40G model Travelstar like the Samsung notebook HDD sample might well have matched it in noise; both Samsung and Hitachi say the single platter models are a little quieter. The 40G Travelstar model is one we’ll want to look at in future, in search of the best balance of high performance and super low noise.
As with most notebook drives, vibration is greatly reduced, so if the other noise sources in your PC are not at the ~20 dBA/1m level, there’s not much to be gained from decoupling suspensions. However, those who wish to plumb the depths of silence will find the small size of notebook drives extremely handy. It is possible to rig up an elastic cord suspension even in a standard 3.5″ drive bay without worrying about drive temperature.
These drives emit so little heat that temperature is rarely a concern in most desktop / tower PC setups. And quiet though they are normally mounted with adapters in a standard 3.5″ bay, it’s a real pleasure to behold the sheer absence of any lower frequency HDD vibration induced noise when they are softly mounted.
* * *
Kind of an aside: Notebook HDDs are very sturdy when installed, but a little delicate to handle. It is best to hold them by the sides rather then by their top and bottom. Even a thumb and forefinger can apply enough pressure to damage the movable actuator heads because of the thiness of the cover. (Does that sound like experience speaking?) Another thing is that you need an adapter to use them with standard IDE cables. If in doubt about how they fit, do ask someone who knows, hopefully whoever sells it to you. Because two of the pins take in 5V, if the connector is reversed, it’s probably possible to burn the drive. We use the type of adapters show in the photo below.
Another item you might consider is a pair of side extension rails so you can mount them in 3.5″ bays normally (with standard screws), but we recommend against this. It’s so easy to damp / suspend it in a 3.5″ bay!
* * *
Our thanks to Hitachi Global Storage Technologies and to Frontier PC for the notebook drive samples.
* * *
