Momentus XT, Momentus, Momentus Thin, MobileMax, and other notebook drives

[blairwillis]

---

tacosauce wrote:
You're sol, seagate has reported that these drives are incompatible with a macbook pro without the apple firmware.

---

@tacosauce:  Where was this?  Can you provide a link or documentation where Seagate has reported this?

FYI: I just posted the following to another related thread on this forum and copying here for those that haven't seen:

http://forums.seagate.com/stx/board/message?board.id=ata_drives&message.id=15137&jump=true

Just installed the 2.5" 500GB 7200rpm Seagate Momentus in a 15" MacBook Pro (pre-unibody) and have noticeable clicks, often accompanied by brief freezes / lag / hiccups.
 
I was very hesitant to purchase a Seagate drive, because the only hard drive failure I ever had was a Seagate, but it was many years ago and I felt that they're still around so they must be doing something right.  Plus, there isn't much choice in the 2.5" 500GB 7200rpm specs, so Seagate it was.
 
This issue is affecting many systems, both PC and Mac, OEM and upgraded drives.  I found extensive discussion on the apple forums here:
 
http://discussions.apple.com/thread.jspa?threadID=2049659&start=0&tstart=0
 
At the time of this post, that topic is currently at 128 pages and nearly 2000 posts.  Let me sum it up for you:  No fix yet.  This drive is OEM in MacBook Pros.  There is an Apple hard drive firmware upgrade which minimized the clicks for some people, but seems to cause even further system hangs or freezes.  The Apple firmware upgrade can only be applied to Apple OEM drives (that already have Apple firmware on them) rather than an upgraded drive (like the one I just installed).  I'm hesitant to disable APM, and frankly, I shouldn't have to.  Seagate?

This really is up to Seagate at this point to fix this drive.  I'm going to tough it out for a few days, but based on what I'm seeing so far (and the fact that this discussion has been "officially" ignored by Seagate), they're about to loose a customer that they had previously lost for the past decade.  And I *won't* be back in another decade, either.  I'll also be returning this drive to the retailer (Best Buy) with an earful about why they need to send back all these drives and the fact that it's being ignored by Seagate despite a very large user pool experiencing this problem across all platforms. 
 
I was so looking forward to a hefty, speedy drive in my MacBook Pro, but random system pauses and hiccups are absolutely intolerable for someone using this drive professionally.  Let's hope Seagate starts acting professionally and publicly acknowledges the problem and issues a firmware fix.

[HughR]

I don't have one of these drives.  But I'd like to understand the issues.  The discussion seems to me to be rather overheated and confused.  Please correct me if I'm wrong.

From reading some of these many posts, it seems to me:

"out of the box" these drive make clicking noises and cause machine delays in ways that people find disturbing and inconvenient.

These clicks and pauses are caused by the drive powering up (after deciding on its own to power down).

Why was the drive powered down?  APM (power management) firmware in the drive decided that the disk had been idle long enough that it was a good idea to power down.  This decision is based on so many seconds of disk inactivity.

The one cure that seems to work is to turn off APM.  There are known tools to do this (QuietHDD on MS WIndows and hdparm on OS/X and Linux).

So what is the problem?  If you don't like the APM behaviour, turn it off!

If I had the drive, I'd probably consider tuning APM rather than turning it off.  Unfortunately, I haven't seen any documentation for the meaning of each value of the APM setting (except 128 is default and 255 is off).  So a little experimentation is in order.

My guess is that the default setting is just too anxious to power down the drive.  Perhaps its time delay is just badly co-ordinated with the frequency that sync(1) runs on OS/x. 

 Seagate does not appear to have addressed the issue with a clear message and set of instructions.  Unfortunate.  Worse, Seagate support seems uninformed.  (From my unscientific sampling, this seems all too predictable.)

Please: correct me if I have the wrong impression.  I'm not trying to pick a side, I'm trying to clarify the issues.

[blairwillis]

---

HughR wrote:

I don't have one of these drives.  But I'd like to understand the issues.  The discussion seems to me to be rather overheated and confused.  Please correct me if I'm wrong.

---

Actually, I appreciate your interest in helping sort out the issues, HughR.  I'll attempt to answer your questions at least from my point of view.

>> "out of the box" these drive make clicking noises and cause machine delays in ways that people find disturbing and inconvenient.

Yes.  Such a thing could be dismissed as "subjective" so let me clarify.  I've personally owned dozens of computers, both PC and Mac, desktop and laptop, PATA, SATA, 2 TB -> 200 GB, internal, external, and everything back to cassette tapes and true "floppy" discs.  The audible emissions of these drives when installed in a proper and reasonable manner is unacceptable (but it's not quite as bad as the regular operating noise of the old growling 5.25" floppy drives!).  In between "clicks" the operating noise of the drive is appropriate and could be described as "quiet."  It's the clicking that is loud enough to be annoying in that it does take your attention.  The pause that accompanies the clicking is also noticeable and significantly enough to interrupt the computer's operation. In a professional situation, such as recording audio, this is unacceptable.  In a recreational situation, it's just annoying, and to the point that I consider it a defect in operation by reasonable standards.  (A similar product would not cause the same annoyance.)

Additionally, the frequency of this event is ridiculous.  This doesn't happen just "once in a while" but at times constantly.

>>These clicks and pauses are caused by the drive powering up (after deciding on its own to power down).

This appears to be true. 

>> Why was the drive powered down?  APM (power management) firmware in the drive decided that the disk had been idle long enough that it was a good idea to power down.  This decision is based on so many seconds of disk inactivity.

It seems the APM is not functioning properly.  The cycling does not consistently accompany periods of inactivity (or activity).  It frequently occurs during very active disk use.  The seeming "randomness" of the events is a big factor in the frustrations many users are reporting. 

>> The one cure that seems to work is to turn off APM.  There are known tools to do this (QuietHDD on MS WIndows and hdparm on OS/X and Linux).

>> So what is the problem?  If you don't like the APM behaviour, turn it off!

Today I decided to try this (using hdparm on OS X 10.6).  Since turning off APM, the clicking has gone away, as far as I can tell during several hours of active use.  

However, I don't have a good understanding of how hdparm and APM actually work.  I am concerned about the side effects this may have on my machine.  If you or someone with a better understanding of APM can educate me and others, I'd appreciate it.  Specifically, am I risking drive longevity?  Excessive temperature due to a constantly operative mechanism?  A noticeable effect on battery life? (Yes, I'm sure I could run some real-life tests on this one... I've just been plugged in all day)

>> If I had the drive, I'd probably consider tuning APM rather than turning it off.  Unfortunately, I haven't seen any documentation for the meaning of each value of the APM setting (except 128 is default and 255 is off).  So a little experimentation is in order.

I think I'd be more comfortable with this if I had a better understanding of APM.  Is it simply an activity/inactivity detector that starts and stops the drive?  I'd seen discussion that the problem may be related more to a shock or motion sensor that was meant to park the drive in to prevent motion damage.  Also, does APM change an operating system function, or is it changing an internal function on the drive itself?

 I'd really like to let my OS handle the things it's supposed to handle so I can focus more on using the machine productively.  There is a setting in Mac OS X "Energy Saver" preference panel that says "Put the hard disk(s) to sleep when possible" and I wonder if changing the APM setting has any disabling effect on this?

>> My guess is that the default setting is just too anxious to power down the drive.  Perhaps its time delay is just badly co-ordinated with the frequency that sync(1) runs on OS/x.

I don't think this is an OS specific issue, but I could be wrong.  It happens to have received a large amount of attention on the apple forums because of Apple offering the drive as a build-to-order option.

>> Seagate does not appear to have addressed the issue with a clear message and set of instructions.  Unfortunate.  Worse, Seagate support seems uninformed.  (From my unscientific sampling, this seems all too predictable.)

Agreed.  From a product performance and customer service standpoint, I think a "solution" that requires a customer to install some unix tool and tinker with an APM setting is far from ideal.  First off, communication is important.  They need to recognize that a large set of customers are experiencing a real symptom and not brush it off as a bad unit or as a normal behavior.

I took the time to research other users' experiences before beginning to post on my own.  I'd realized that this issue was widespread, and not an isolated incident.  They need to do the same.

I provided enough related info to hopefully assist others in finding the related topics (here in the Seagate forum and on the Apple forum) along with the other keywords that might help elevate this in search results.  Even a "me too" will help bring attention to the issue and perhaps motivate a solution.

Finally, I am concerned that Apple (and Dell?) have released a firmware update for their factory installed (OEM) versions of this drive (same models) that may have "fixed" this issue.  But I'm using one purchased from a retailer and installed in my MacBook Pro, and the Apple updater does not recognize the after-market installed drive as updatable.  C'mon, if this firmware update will fix the drive, can't we get it done already?  If Seagate support seems to be categorically ignoring the problem, I don't have much faith that someone is behind the scenes working on a solution.  Perception =? Reality

 

[tacosauce]

@blairwillis:  I had a conversation with a seagate tech where I was told that these drives in an OEM state are incompatible and that apple has firmware which works. So, since I don't work for seagate I wont be able to provide you with a link or what not.  I hope you get it sorted.  I no longer have the drive, instead I own a 500gb hitachi so this is likely my last post.

[fzabkar]

If the drive clicks when it is active, then AFAIK this indicates a fault in the preamp, or a weak read head, or a bad patch on the platters. A SMART diagnostic may help identify whether bad sectors are a problem. It may also show whether the drive is experiencing an excessive number of load/unload cycles.

Advanced Power Management (APM) is defined in the ATA-8 command specification.

ATA/ATAPI Command Set - 2 (ACS-2):

http://www.t13.org/Documents/UploadedDocuments/docs2009/d2015r1a-ATAATAPI_Command_Set_-_2_ACS-2.pdf

Section 7.52.6, "Enable/disable the APM feature set", has a table of APM levels.

Count ..... Level
----------------------------------------------------
00h ....... Reserved
01h ....... Minimum power consumption with Standby
02h-7Fh ... Intermediate power management levels with Standby
80h ....... Minimum power consumption without Standby
81h-FDh ... Intermediate power management levels without Standby
FEh ....... Maximum performance
FFh ....... Reserved

"Device performance may increase with increasing APM levels. Device power consumption may increase with increasing power management levels. The APM levels may contain discrete bands (e.g., a device may implement one APM method from 80h to A0h and a higher performance, higher power consumption method from level A1h to FEh). APM levels 80h and higher do not permit the device to spin down to save power."

"[Set Features] subcommand code 85h disables APM. Subcommand 85h may not be implemented on all devices that implement SET FEATURES subcommand 05h."

AFAICS, maximum performance is set using a count of 254 (FEh), while 128 (80h) is the setting for minimum power consumption without spinning down the drive.

Strangely, I couldn't find anything on APM in the Momentus 7200.4 SATA Product Manual:

http://www.seagate.com/staticfiles/support/disc/manuals/notebook/momentus/7200.4%20(Holliday)/100534376a.pdf

Page 199 of the following HitachiGST document describes how APM is implemented in the Deskstar 7K1000 model:

http://www.hitachigst.com/tech/techlib.nsf/techdocs/F8156F65CFE91CCF862573160072C729/$file/7k1000_sp.pdf

[Lutin M]

Hi, I got a Sony vaio AR41S, with raid.

I've decided to change my harddisk (2x160Go SATA - Raid0=320Go)

with 2 Momentus 7200.4 500Go with G-force, (and RAID1=500Go in Mirror)

First, the second harddrive failed after 24 hours.. I heared tic-tic of death, indicating that head are out.

After contacting my reseller, I've bought a third one, because the replacement delay were more than a week,

I've receive the new one after 48h, change the disk, (return the crashed one) create raid, rebuild mirror.. and after some reboot, the mirror is broken.

after some other RAID is ok and mirror is rebuild...

after 10 days, the raid controller said disk in error, and the disk stop working while 10 days... then it worked again...

while this time, I've received the other one (exchange after 12 days)

I've re-changed disk, rebuild mirror... but same effects with this disk.

conclusion: the momentus 7200.4 isn't compatible with RAID, and with critical application! (found on seagate website)

I've mail the problem to seagate: the answer is "this disk is not for RAID!"  (sic! raid is a sort of controller! I've never heard about disk for raid or not for raid!)

some times my Vaio make bips like MacBookPro (I canno't apply Apple firmware update...)

I've found users that have same troubles with Lenovo notebook, Dell, Sony... and only Apple made an update?

Seagate dosen't recognize the problem, and no firmware seems to be edited...

Last question: is data storage a non crtitical application? it looks like  for Seagate..."- Your disk, you'll brick-it? or it's for storage?"

Last conclusion: Seagate Momentus 7200.4 500Go is not made for data storage.

[HughR]

blairwillis: Thanks for your long and comprehensive answer.

My theory is that the click-and-delay is not a disease, it is a natural consequence of the drive powering back up.

Assuming that that is correct, and that most people don't want that, then it seems to me that they should turn it off (setting 255).

Here are some reasons that one might not want this behaviour:

• it causes delays at inconvenient times
• the noise is disturbing
• it is happening way too frequently (the drive can only handle this 20,000 (number from my memory and might be wrong) times in its lifetime)
• it causes RAID configurations to think that the drive is broken and take drastic and unfortunate remedial action
• perhaps the OS has a better idea than the disk drive about when it would be a reasonable time to power down the drive: let the OS decide.

Here are some reasons that one might want this behaviour

• if the drive happens to be powered down long enough, power is saved, heat is reduced, wear on the drive is reduced.  I don't know what "long enough" is, but I'd guess 10 minutes minimum.

• if you happen to drop the machine while the drive is powered down, it is more likely to survive.  Perhaps this advantage does not apply for the "G" model which has anti-drop features.

In some ways, the disconcerting noise is actually useful as a warning.  With other drives on a notebook, I've exprienced way too many power cycles with no noise (a silent killer!). See, for example, https://wiki.ubuntu.com/DanielHahler/Bug59695 [The stupid forum button to add a link isn't working].

 "It seems the APM is not functioning properly.
 The cycling does not consistently accompany periods of inactivity (or
activity).  It frequently occurs during very active disk use.  The
seeming "randomness" of the events is a big factor in the frustrations
many users are reporting. "

That I don't know about (because I don't have a system to test).  If true, that is a problem and a bug.  How did you reach this conclusion?  Disk activity that hits the cache may not count.  The sound happens only when activity starts again; inconveniently, there isn't a sound when the drive is being spun down (or is there?).

How frequently does OSX flush buffers to the disks (sync)?

[HughR]

fzabkar: thanks for posting useful specifications.

I also found that SMART attribute  193 Load_Cycle_Count useful (for different notebook drives).  It showed that I was having a similar problem: too many load/unloads.

I don't think that the click is a symptom of a fault.  I think that that is the way this disk works.

Calling it the click of death (as Lutin M did) is unwarranted.  I don't think that this is anything like the horrible Zip drive problems that caused this term to be coined.

[HughR]

Lutin M:

APM isn't compatible with RAID, at least not RAID with short timeouts.  Turn it off.

Now Seagate has this nasty habit of claiming (roughly) drives not certified for RAID won't be supported in RAID applications.  They have never explained why in technical terms: nothing in the specifications gives a technical reason.  I've argued elsewhere that this is unreasonable and unacceptible.

[Voicedesign]

+1 case

Lenovo X200T

Place is Russia

After one or couple "beep" MS Vista is  fully freeze.

Message Edited by Voicedesign on 10-04-2009 09:40 AM