Is this a genuine or fake WD drive?

[lImIlAo]

Case 1

Someone bought a WD20EURS (AV-GP) on newegg (China) and found some suspicious points.

Pay special attention to the typos: Formart, Addlitional, avaliable.  If you enlarge the image, you’ll find most ｉ’s look like ｌ’s, and the printing is blurred. 

Additionally, according to the author, AIDA64 says this drive does not support streaming where AV-GP should definitely support.

However, the SN and PCBA are VALID on WD verification, and the model number displayed in AIDA64 is OK.

Both Newegg and WD (China) say the drive is genuine as long as SN is valid. <-RIDICULOUS!!!

FYI There is software that can change drive info displayed in HDTune etc.  Also, it is easy to copy SN and PCBA from genuine ones to fake ones because the packages are transparent.

Q1:

Do these typos prove the drive to be a fake one, or are they just WD’s anti-forgery measure?

Another photo that does not contain typos

Case 2

Some people posts the drives they bought from the biggest B2C website in China.  Yet the label seems to miss some signs…

…while others do not miss signs

Q2:

Are these missing-sign drives genuine or fake?

Case 3

There exist different appearance for a same model number.

Q3:

Are these appearance all genuine ones?

[lImIlAo]

All right, I’ve erased all S/N so that they will comply with privacy requirement.

[fzabkar]

The first drive, with the spelling errors, is obviously a fake, and a very bad one at that.

AFAICT, even the DCX number is wrong. AIUI, the DCX should be related to the middle three characters in the model number suffix (63S48Y0), eg …

DCX = nSxx4yy8z

One other way to check for fakes is to compare the date of manufacture on the label against the YYWW or YWW (Year / Week) date codes on the chips on the PCB.

As for the one with the different appearance, you could check its PCB number against the known number for that model. The sticker on the PCB should have a 2061-nnnnnn number, whereas the printing on the PCB itself would have the same number, but with 2060 instead of 2061.

Still another test would be a HD Tune read benchmark. The width of the access time graph would tell you the RPM. In this way you could tell if a Black drive has been substituted with a Green.

An examination of the part number of the SDRAM on the PCB would tell you if a 32MB cache model has been substituted for a 64MB cache.

Weighing the drives may reveal a difference in the number of platters and heads.

[lImIlAo]

Thank you very much for sharing so much knowledge.  I never noticed DCX number or PCB number before.

But there is still possibility for forgery.

The only thing that can be easily checked and cannot be changed is PCB number (2060-nnnnnn).  Yet identical PCB number can be used on different model, so forging sticker and R/N on the front label is still possible.  Also, since people seldom share their PCB number, it would be hard to know if a PCB number is wrong for the model.

When DCX is copied along with MDL, DCX check is nullified.

Chips and SDRAM is hard to check because they are on the inner side of the PCB.

Performance benchmark is a good idea – but only for different models with distinguished performance difference.  If a ‘new’ WD20EARS is actually forged from a repaired WD20EARS, benchmark is of no use.

Weighing is a cool idea, but only applies to models with different platter number.

Thank you anyway.

[fzabkar]

I have some additional observations in regard to the PCB part number, PCB date code, and R/N in the following thread:

http://community.wdc.com/t5/Desktop/WD20EARS-00MVWB0-and-AAM-option-Supported-and-Not-Supported/m-p/226958#M6398

AFAICT, the last 4 digits of the R/N (on the label) should be present in the PCB p/n. The rear of the PCB may also have a YWW date code.