See the BOTTOM of this page for a solution for Windows, including Windows Vista and Windows 7.
Some back story:
I recently purchased a new hard drive. When there was no disk activity, I would hear the hard drive click every few seconds. I’ve heard drives click before, and it was never a good sign.
I didn’t think a new drive could be going bad so quickly, so I googled it to see if that particular drive had a clicking problem.
It turned out a lot of drives have the “clicking problem”. They have over-aggressive power management settings that try to park the drive heads over and over. Every time there is a click, the drive’s “Load_Cycle_Count” goes up by 1. This can be viewed in the drive’s SMART information.
My “brand new” drive had over 10 times as many load cycles after a few days usage compared to a Desktop drive I have that was a year old!
This is seen as a problem by many, as hard drives only have so many load cycles before MTBF.
Besides the thought of losing my drive quicker than I should, it was mostly the annoying clicks that I wanted to get rid of!
Here are some links I found regarding the issue:
It seems that the problem happens with any drive that boasts “power savings”, such as Laptop/Notebook hard drives and the Western Digital “Green” line of Desktop drives.
The issue isn’t specific to any OS since the “problem” is set in the hard drive’s firmware.
There are a few tools you can use to set the drive’s power management settings, but the change is temporary.
hdparm for Windows: https://sites.google.com/site/disablehddapm/
“hdparm -B 254 hda” to disable power management
“hdparm -M hda” check acoustic settings
“hdparm -M 254 hda” set acoustics to minimal (fastest performance) – this DOES stick on the drive
“hdparm -i hda” display drive info (such as Power Management level)
smartctl for Windows: http://smartmontools.sourceforge.net/download.html
“smartctl -A hda” display SMART info (check the number for “Load_Cycle_Count”)
Notebook Hardware Control (NHC): http://www.pbus-167.com/
NHC will sit in your system tray and will try to keep the hard drive’s Advanced Power Management set to 254/disabled.
Even if you disable the “APM” with hdparm, if you reboot or your system resumes from sleep/hibernation, the drive will be right back to its default setting.
I just found another program that tries to correct the issue. Unlike NHC, *all* it does is try to control the hard drive.
It’s called quietHDD: http://sites.google.com/site/quiethdd/
quietHDD quickly loads, sets the drive’s power management, and uses hardly any memory (under 1 Meg). It even re-sets the APM mode after your system resumes from sleep. It can even start without a tray icon!
Of course, with programs like NHC and quietHDD, they require Admin-rights to the drive, so that means you can’t run them on startup under Vista or Windows 7.
There is a solution for that; Microsoft provides a way to easily run any EXE as a service.
I don’t want to get into all the details now, but it involves downloading and installing a Resource Kit, running some commands, editing the registry, etc.
To make it simple, I have some pre-setup stuff here.
**** Solution (for Windows):
This is basically a quick and dirty hack using someone else’s files. So I may remove them later.
1) Download quietHDD from here: http://sites.google.com/site/quiethdd/
2) Extract “quiteHDD.exe” and move it to C:\Windows\System32\
3) Download http://xenomorph.net/data/hdd_click/quietHDD_Setup.zip
(it has a quietHDD config file for 254/255 for both acoustic and power management, and the Microsoft service-loader).
4) Extract the zip file and move the contents to C:\Windows\System32\
5) Download http://xenomorph.net/data/hdd_click/quietHDD_Service.zip
(registry file to set up the service).
6) Run the REG file in that zip to add the service, then Reboot.
On my T43 (32-bit, Windows 7), it seems to work perfectly. When the system reboots, quietHDD is loaded as a service, power management is disabled, and the hard drive no longer clicks.
I just got some information from Fujitsu USA:
“The reason for the clicking is due to the drive parking its heads on a ramp. We do this not as much for power consumption but for purposes of getting the heads in a position where they are less likely to be damaged during an impact.
All disk drive makers think pretty much the same way on this topic. Power management is a secondary benefit.
As for early disk failure due to continuous unloading on the ramp, We have studied the result of this function and understand that there is no wear on the actuator and minimal wear on the ramp itself. However, any wear is far outweighed by the reduction of drive failures due to head disk impact when the heads are left flying over the media.”
I still don’t know if a high Load_Cycle_Count would trigger the “drive failure imminent” message that SMART does if certain numbers it keeps track of get too high.
I’ve seen the drive clicking referred to as a “laptop drive killer” by some, and then called completely harmless by others. Now I’ve seen the clicking said to be actually good for the drive.
Something else I’m not sure about – supposedly these drives are designed to park their heads after so many seconds (27.5 seconds). This reduces chance of head crash if the drive is dropped.
The reason I purchased the specific Fujitsu drive is because it supports “IDLE_IMMEDIATE with UNLOAD” – a feature required for my ThinkPad’s “Active Protection” system. Basically, when my system’s motion sensor detects too much movement, it tells the drive to park its heads immediately (brace for impact).
So would I really need the drive parking itself every 27.5 seconds when the computer itself will tell it to park if needed?
It’s all so confusing.
I got another email from Fujitsu USA:
The guy explained ramp-related failures account for something like 1% or less of hard drives.
But “excessive shock” is the #1 killer.
“Excessive Shock damage causes the heads to slap into the disk, if the heads are not up the ramp, which is usually catastrophic. If the head is on the ramp (as we suggest) and the system/drive receives a shock then no damage can happen to the head or media.”
Regarding the 600,000 Load_Cycle_Count:
“The 600,000 is the typical spec of most drive ramps. We actually test to 1,000,000 load/unloads over various temp ranges and conditions. It is possible for the drive to exceed 600K in a year (however unlikely) if you use the drive nonstop, 24 hours a day and access the drive every 40 seconds continuously for that whole year. Now that scenario is not only unrealistic but it is so far beyond the design spec of the drive its not funny. The reality is a drive typically unloads 40 times an hour if the drive (not the system) is under constant use. Based on a typical 8 hours of drive use per day that’s 336 unloads per day. If you do that every day for a year that’s 122,640 unloads per year. That means you will reach the spec of 600K in 4.89 years.
We track Load/unload values in SMART. We do not have a threshold for load/unload however. Which means we do not raise any warnings to the system for load/unload. There are SMART values of much higher significance that we do track and have a warning threshold for.”
So, even if you hit 600,000 Load_Cycles, SMART won’t start to nag or warn.
Fujitsu says their drives are tested to 1,000,000 Load_Cycles, and even then, Ramp failures are super-rare.
Part of the email included attachments labeled “Fujitsu Property & Confidential”, so I guessing they went a little above and beyond to give me some info on it.
In summary, what it sounds like, even though the clicking is annoying, an excessive Load_Cycle_Count may not be something to worry about.