Question Need Help regarding identifying issue with faulty HDD

Aug 5, 2019
9
0
10
Hello everyone!
Writing this post regarding a Western Digital Green 2 TB HDD (WD20EZRX) that failed a couple of days back. The disk was purchased back on March 2013 and was used as a secondary backup device (the primary being an SSD). Just a day before the HDD failed, I noticed that it was not showing up on "This PC". I restarted my PC and it showed up normally.

Being worried, I ran a Windows Disk Check on the drive(it has only 1 partition). It completed fine. Also ran a quick test from the WD Diagnostics Utility which also completed without an errors. S.M.A.R.T. status showed OK. I accessed some files and also transferred some photos from my mobile device to the drive. Exactly the next day, while turning on my PC I noticed that Windows was taking unusually long to load.

Since this happens sometimes when a faulty drive is present, I quickly opened up "This PC" and Device Manager to find that the drive was not showing. Restarted my PC to no avail. The drive was not even showing in BIOS. I restarted the PC again and took my ears close to the drive bay and realised that the drive is not able to spin up. It spins for about 4-5 seconds on starting and stops completely and is dead silent.

Restarting the PC again causes the same thing to happen. Once after those initial five seconds, the drive doesn't try to spin up again and is virtually dead. Windows takes much longer to load on a separate SSD, while the faulty HDD is kept connected.

However, diskpart doesn't have any delay while running in Command Prompt and shows up(diskpart usually takes a lot of time to pop up when a drive with bad sectors is connected to the system). After restarting the PC once more i touched the PCB and found out that the part of the PCB below the motor controller IC is getting extremely hot(above 70 °C or so), whereas the rest of the PCB is a little above ambient temperatures.

The surface of the HDD cover is also close to ambient. I opened only the PCB by removing the torx screws. There were no apparent burn marks on any area of the PCB. Didn't check for shorts with a multimeter. I had an old WD Blue 1 TB lying around, which is still detected but has some bad sectors.

Knowing that a PCB swap with exactly the same model of HDD, having same board Rev., DCM, etc doesn't solve the issue as the firmware chip needs to be transplanted (and in some cases the firmware chips aren't even on the PCB).

On swapping the PCB anyways, from the Blue 1 TB, I noticed that the HDD was detected in BIOS as ATA(186540 MB) which is otherwise WD20EZRX(186540 MB). The drive now spinned up for a few seconds and started clicking repeatedly, probably due to firmware and board mismatch. However it didn't completely stop like before and the clicking continued.

However there was no screeching, beeping or squealing noise, neither after the PCB change, nor before. However the same area on the Blue 1 TB's PCB was still extremely hot(nearly untouchable). I turned off my PC, removed the HDD, put back the respective PCBs. Plugged back the Faulty Green 2TB. Same issue, spins for exactly 5 seconds and then dead silent.

At this point I removed it and stored it away in an ESD free bag and submitted it to a professional data recovery centre the next day. Connecting the Blue 1 TB later I found out the same area of the PCB was much cooler than it was on the Green 2 TB.

Now the Professional Data Recovery centre has charged around 700$ excluding taxes to recover the data and they won't specify me the exact issue with the HDD even after many requests. I suspect this is a PCB issue and at max a motor issue in addition to the PCB.

Need some confirmation on the matter from people who have extensive knowledge on this topic. I know what goes into Data Recovery, the costs of equipment like PCH3000, clean room costs, but I feel they are overcharging me considering the fact that PCB issues are charged around 1/3rd of what they quoted provided there are no mechanical faults or physical damage. Even a motor failure along with the PCB demands a maximum of 500$ from competitive recovery providers.

Problem is my province has only two data recovery providers and it's kind of a duopoly over here and sending it to another province where it's really competitive will involve shipping. There is also a malpractice that goes on wherein if you refuse to proceed with the job, the recovery centre cause some more physical damage internally to make it difficult for other providers to recover it. They do this even after charging a non refundable analysis fee of 15-30$.

If anyone has taken the time to go through so much stuff then thank you already. Any information that can shed light on what exactly the issue might be, will be of immense help.

P.S. : Attached a reference picture of the same PCB highlighting the area that was overheating in green.

dZkMWfN_d.jpg
 
Last edited by a moderator:
Aug 5, 2019
9
0
10
certainly not meant to be at that temperature. but your idea is really good, i have plenty of heatsinks lying around, mainly used in arduino and pi projects. will try with those. since getting the data is imperative i didn't even take any chances with DIY tricks till now, like putting in a freezer(probably doesn't work anymore in modern drives) .
 
Aug 5, 2019
9
0
10
What are the 2060-nnnnnn numbers on each of the PCBs?

Do both drives spin at 5400 RPM? What is the model number of the 1TB Blue?

Normally when a drive spins up and back down again, it is a sign that it cannot detect a servo signal, ie it may have bad heads or media.
c1ffVOr.jpg
ISzBeRx.jpg
UrUaVqp.jpg
9oMjY7v.jpg

P.S. The WD Green was returned back to me by the data recovery company yesterday. Hence the opening marks on the product label.
 
Last edited:
Aug 5, 2019
9
0
10
Continual clicking that you mentioned is usually a sign of internal mechanical failure (the heads jammed and trying to free themselves --- that's the clicking noise.).
the wd green clicks when used the the wd blue's pcb. when the green is used with it's own pcb, it spins for just around 5 seconds, and powers down, with no clicking or any other unusual high pitched noise. until i power cycle my system, the wd green remains dead. on restarting, it again spins up for the same duration (~ 5 seconds) and spins down(no other noises)

P.S : All of my data is on the Green. The Blue has been a paperweight for me since the past couple of years with many bad sectors but somehow it still works(is detected, can be accessed).
 
Aug 5, 2019
9
0
10
Using the wrong PCB is just doomed to failure. There will be firmware settings that are dependent on the particular unit in question (not the model, the physical device)

Have you put in place a backup strategy?
yes i know. even two similar models coming out back to back from the assembly line won't work with PCB swaps, let alone PCB swaps between a Blue and a Green. I am not even sure that swapping the U12 firmware chip on exactly similar boards will make the PCBs interchangeable, because it probably also depends on the DCM of the drive. I didn't attach the Blue's PCB on the Green expecting the Green to start functioning normally. I just did it to check whether it kept the HDDs behaviour the same.
 
Aug 5, 2019
9
0
10
Not to belabor the obvious point, but why was this important data not backed up?

A physically dead 1TB drive is $45, or free with warranty.
The data on it is priceless.
I never imagined an HDD might fail this way, immediately, all of a sudden, without any audible signs or warnings during usage. Drives that previously failed on me generally started degrading visibly during usage or made weird noises which allowed me to copy all the data off them before they failed. So yes big mistake on my part in this case and a lesson learned too. Will be going for a cloud backup plan if I eventually manage to recover the data, that is if the recovery company whom I had given it for analysis didn't damage or wipe my data. Most of the companies which are shady do these kinds of stuff and this company seemed quite shady because they won't tell me the exact technical issue with my drive, which I obviously deserve to know when they have quoted 750$ (without taxes). Also a 25% advance(25% of 750$) to start the recovery process (after which i would be informed how much data could actually be recovered).
 

USAFRet

Titan
Moderator
I had a 3TB WD Green go from seemingly perfect to dead in about 36 hours. No warning, no outward signs...nada.
I had a 960GB SanDisk SSD go from perfect to absolutely dead in the space of 'power off, wait 5 minutes, power up....'Hey, where's the other drive?' Dead dead dead.

Dead drive, virus, lightning strike...always act as if your data is no more than 0.25 sec away from, total disaster.
 
  • Like
Reactions: reverend95
Aug 5, 2019
9
0
10
I had a 3TB WD Green go from seemingly perfect to dead in about 36 hours. No warning, no outward signs...nada.
I had a 960GB SanDisk SSD go from perfect to absolutely dead in the space of 'power off, wait 5 minutes, power up....'Hey, where's the other drive?' Dead dead dead.

Dead drive, virus, lightning strike...always act as if your data is no more than 0.25 sec away from, total disaster.
have heard the same in case of external drives too(ik they are still the same mechanical hdds inside). after getting my data recovered, i will be going for a cloud storage plan. a single time data recovery cost is pretty much equal to what it would cost me to get 10 years of cloud backup for the amount of data i have. so yeah.
 

USAFRet

Titan
Moderator
have heard the same in case of external drives too(ik they are still the same mechanical hdds inside). after getting my data recovered, i will be going for a cloud storage plan. a single time data recovery cost is pretty much equal to what it would cost me to get 10 years of cloud backup for the amount of data i have. so yeah.
The key to data recovery is "Not in one place"
If my NAS were to die, the data still lives in the main systems or the NAS backup.
If a main system were to die, the data still lives in the NAS.
etc, etc, etc.

And cloud has its own issues.
Changing prices or terms of use.
And then trying to recover the entirety of a drive might be an issue. 600-700GB over the internet might strain some peoples data caps, or patience in waiting for it to arrive.
 
Aug 5, 2019
9
0
10
Even if the PCBs were compatible (they aren't, AFAICT), the Blue drive spins at 7200 RPM while the Green spins at 5400 RPM.

I only hope for your sake that the HDA pinouts are identical, otherwise the preamp (on the headstack) may have been damaged.
This is what the drive currently sounds like after powering up the system. Note that the first beep is from the motherboard speaker after successful POST.

https://vocaroo.com/i/s1ujlYma2PgL
 
Aug 5, 2019
9
0
10
The key to data recovery is "Not in one place"
If my NAS were to die, the data still lives in the main systems or the NAS backup.
If a main system were to die, the data still lives in the NAS.
etc, etc, etc.

And cloud has its own issues.
Changing prices or terms of use.
And then trying to recover the entirety of a drive might be an issue. 600-700GB over the internet might strain some peoples data caps, or patience in waiting for it to arrive.
except change in ToS, there are really no such contraints for me in terms of internet bandwidth/data cap. but you are right about the fact that a good backup plan means not only relying on the cloud, but also making physical backups on an individual level.
 

USAFRet

Titan
Moderator
except change in ToS, there are really no such contraints for me in terms of internet bandwidth/data cap. but you are right about the fact that a good backup plan means not only relying on the cloud, but also making physical backups on an individual level.
Right.
I just put that out there because some people do have monthly data caps.

My FiOS isn't capped. Although, I imagine if I were trying to do multi TB every single day...someone at Verizon would eventually have issue with that.
 
  • Like
Reactions: reverend95