160GB - Clicking Noise

I have every confidence in my acer contact, I aim to get an Acer repair reference for all.

Please be patient, I believe that ACER are less than impressed with WD behavior ie. the **** will hit the fan

 

We are very grateful for your efforts so far, teaker1s.

In your opinion, will longevity of the drive be prolonged if APM is disabled by setting the appropriate register to 255 (hdparm -B 255)?

Do you think it would help or hinder our cause if the details of this issue were leaked to somewhere such as The Register or Slashdot?

 

I would ask that we allow the chap from Acer time to respond, everything so far points to WD stitching anyone with this drive including computer manufacturers.
Dell got an update on firmware, but I would guess it took pressure, basically WD cannot be bothered to sort firmware.

From what I am told Acer are taking this seriously and seem less than impressed with the evidence so far, particularly after reading my conversation with wd and their reference refusal for Acer.

I think until Acer respond officially we would only harm our cause by blowing the lid on WD S***y behaviour so far.

Reference hdparm 254 or 255, if you understand you are trading shock protection v hdd life.
To recap if you are using as desktop with ac adapter on a table and don't park the heads = no great issue unless dropped, used as a netbook and jolted etc without parking heads could cause damage.

As an example on proper 120gb model I am successfully using value 200 for batt and 254 for ac


regards
Daniel

 

Re: 160GB - Clicking Noise how to check

Finding that some aspire's with160gb fail to respond to any hdparm

Firstly we need to enable right click xfce menu that is hidden on current acer setup, this link shows nicely how to turn the menu on
http://kengell.blogspot.com/2008/09/acer-aspire-one-enable-xfce4-right_11.html

now the hdd fitted to the acer is SMART capable, it means it can report back it's status and temp.
Active wireless or ethernet connection required

If you right click a couple of centimeters to the left and level with the word settings on desktop-you are presented with the xfce menu.
Select
System>add/remove software, search for smartmontools
apply and install the Smartmontools (agree to any prompts)

Smartmontools can query the hdd, all information is exactly as hdd has monitored itself.

right click again setect terminal and copy this command

Code:

sudo smartctl -a /dev/sda | grep Load_Cycle_Count

life counter for head parking will report a value eg.

now if we leave it for a period of time and issue the command again we will see it has increased.

issuing command below will tell you linux on time

Code:

uptime

or for more complete data issue command

Code:

 sudo smartctl -a /dev/sda

for total hdd power on time and load count and other information.


I have found that in linpus the load count at idle increases at idle by at least 3 per minute, now the drive is rated at 600,000 cycles.

please others can we have load count per hour, best worst scenario

 

Interesting thread. Just skimmed it but will come back and read in more detail and will be interested to see what the response from acer/wd is.

I bought the 160 linpus version from Asda as well. I had noticed the 'clicking' sound as I had the same sound from my fujitsu-siemens laptop and have recently replaced that hd. It was a fujitsu hd and I replaced it with a Hitachi which doesn't click. I also seem to remember excessive hard drive wear being an issue with ubuntu a couple of releases back.

 

Re: 160GB - Clicking Noise how to check

Drive model No., firmware revision and serial No. might be useful also. As reported by hdparm -i /dev/sda. Mine is:

Model=WDC WD1600BEVT-22ZCT0
FwRev=11.01A11
SerialNo=WD-WXC708884870

I run Windows XP on my AA1 and the head parking seemed to be non-active due to a chatty ACPIEC driver and the Wireless Zero Config service writing to disk fairly frequently. After fixing these it would seem that my Load/Cycle count (as reported by HD Tune) is not running away, with APM set to 254 via quietHDD.

I have only had the A150 for a few weeks and noticed that the Load/Cycle count reached over 600 in a very short space of time. After booting Parted Magic from a USB stick it also seems to jump 10 - 20 cycles. Under Windows XP with APM set via quietHDD the Load/Cycle count only increments after resuming from standby or hibernate.

I have to say at the moment that under XP, the Load/Cycle count does seem to be managed if APM is set to something in the region of 200 - 254 very soon after booting, otherwise the count tends to increase quite quickly.

 

WD1600BEVT-22ZCTQ
WD-WXC708633911
FIRMWARE 11.01A11 (only firmware available)

hdparm totally un responsive this one and another similar serial number

From my research this thread is the tip of an iceberg, more I delve the more I find

 

It has been a while since any update from Parimal at Acer, having emailed and yet to hear back-this week I will call his extension number and see what the situation is.

Phone call today and sounds positive as it has been passed to two teams to check and find resolution, have been assured that this is being dealt with and will be resolved.

 

Any more news from acer yet?

I used the method above to monitor my disk head parks. Let it idle for 10min and smartctl recorded 19, working out at 114 an hour.

 

Okay after spending many hours on this, I believe that the firmware across most of the WD range has a feature which has been poorly implemented and is not currently ment to user configured
I believe that WD are deceitful as the few people that know enough to complain are offered RMA. There is a configuration program from WD, which support will say does not exist when specifically asked for that will adjust and also disable the troublesome feature.

As both WD and ACER have not pro actively sorted this, WD could have added the control feature to Datalife tools and ACER could have offered the tool.

What I would like to do is a couple of days full testing, then to blow the lid off this:- as it affects everything from servers to xbox's with unofficial drive upgrades.

mostly I'm ****** off with WD lies-they know and have a solution and choose not to admit it.

Any suggestions on who to contact to get this made widely available

 

Tony Smith - Editor, RegHardware. Tony is an AA1 user and is a member of this forum. http://www.reghardware.co.uk
Barry Fox - Consumer affairs columnist at Personal Computer World magazine. mailto: [email protected]
Slashdot/Slashgear

 

Thanks, I want to sure that I don't cause anyone issue with what I have found-so I am testing and will then allow everyone access to root cause of issue and the program that wd seem less than keen to admit exist.

So far the only thing I think that may be reduced is battery life-how much I'm checking

 

From testing I have found Device Model: WDC WD1600BEVT-22ZCT0 and most of the other WD drives are defective by design and implementation of firmware.

What causes the high load count is poor implementation of power saving, the drive decides to ignore hdpharm settings over it's own defaults.

High load counts have also been reported by owners of green drives, when used in servers in a raid configuration, as you can imagine the owners of several drives running 24/7 were less than impressed-WD responded by selectively issuing wdidle3.exe and instructing them to either set idle time to 25 seconds or disable (no head parking).

Looking on WD site there is no mention of wdidle3.exe and some people who asked for it from WD support were told it did not exist

Idle park timer, this is more than just a simple time delay as using wdidle3.exe to adjust timer you would expect that you could set up to maximum of 300 (5 minutes) idle without any ill effects.

Adjusting the timer much beyond the default 8 seconds results in severe degradation of drive performance and boot messages.

So next I tried to disable the Idle timer feature in the hope that using laptop mode-tools to control hdparm would resolve the situation.

Surprisingly the head count dropped like a stone on AC hdparm 254 and on hdparm 128, I had hoped that disabling the idle timer would allow the drive to accept hdparm control-

alas the only control is do not park heads.

This is not a linux problem, this will affect any system that doesn't constantly access the hard drive.

My dealings with WD there customer support said the drive was defective and offered RMA- when explained to them the drive was built into an Aspireone and I would like to preserve the warranty and get acer to replace it any reference number was refused.

Acer I first contacted gave the standard it's a software issue and phone our national rate line and speak to a script monkey.
So I phoned Acer in January and spoke to an individual though not familiar with linux agreed to pass the information on to two groups to test-phoning back I'm told this is ongoing. Not good enough, I have one Aspireone 120gb on intrepid which is perfect and 2 120gb fitted with 160gb at the factory which are defective and I want fixed at no expense to myself.

WD -need to get their act together and either admit that the firmware cannot give power savings advertised and update the firmware like they did for 60 and 80gb models for Dell. Or recall them as defective.

Acer have had since January to get this sorted, If I can spend hours of my spare time working out the core issue, paid technicians should have got this moving sooner

 

So, would the following serve as an accurate executive summary of the problem, and if so what a realistic value for XXX?

"The default head-parking at idle strategy of some WDC 2.5" SATA drives is very aggressive. Using this default strategy, an affected device would exceed the rated 600,000 load cycle count after only XXX hours. Ordinarily, the aggressiveness of the strategy would be controlled via standard APM commands sent via the host operating system but the affected devices do not respond to APM commands other than at a level where head-parking at idle is disabled completely. This is less than optimal for a mobile device that may be susceptible to disk surface damage if the device is moved whilst the heads are unparked. Also, the device is less likely to enter reduced-power mode whilst the heads are unparked and battery consumption by the host device is therefore likely to be adversely affected."

 

Yes thats the sum total

The power saving feature they are sold on is at the expense of drive life.
Depending on usage I estimate a life span of 1.5 years based on a student using during a normal college day.
The green drives have been noted in server use to hit 200000 cycles in 4 months and are rated to 600000 before they may become unreliable and fail.

If you were to leave a netbook on downloading over night you could realistically hit 600000 cycles in less than one year.

 

Hi all, I've got this problem on my 120GB AOA150-AB. Due to Forces comittments and Comets returns policy I am unabe to return at the moment.

I have seen

Code:

sudo hdparm -B 255 /dev/sda

mentioned with values of 200, 254 and 255. What is the default setting?

Thanks

MB

 

please can you post your hard drive data as listed in bios, the genuine 120gb ones have a different hd to the wd 160 gb

 

HDD Model Number ST9120817AS
HDD Serial Number 5RE1HXR0

 

So returned the above and got a new one, is still clicking but not as often. How often is acceptable?

 

st model hard drives can be cured with hdparm 200 for battery and hdparm 254 for ac on ubuntu we used ugly fix to cure, on linpus laptop mode should work. before worrying please follow the the instructions I previously posted to test your load count