WD Caviar WD5000AAKB prevents computer booting


#1

Hi there.  My computer (HP Vectra VL400 with 6515WU mother board)  contains a WD Caviar WD5000AAKB which has been isolated as the reason it won’t boot up.  Nothing happens at all.  When the drive is removed from the system the motherboard will boot up again.  I have tried the drive in another computer (Dell Dimension 4100) and get the same result.

I believe the HDD must have a fault on the power side of the drives PCB which is an E222034.  Has anyone any ideas to where the problem may lie?  All the components look OK.  

I have read at  http://community.wdc.com/t5/My-Book-for-PC/please-help-i-need-a-logic-board/m-p/12438 that is the pcb needs replacing I need to replace the EEPROM chip to the new PCB but not sure where it is - the thread suggests it is U12 but it looks like there is no chip there.  

Here are the pictures of my PCB

http://picasaweb.google.co.uk/lh/photo/5pwt8eOTc1y3qd5ujN6Aww?feat=directlink

http://picasaweb.google.co.uk/lh/photo/8Dn_1yMwTgNZkH_6Gu-k7A?feat=directlink

Many thanks


#2

The “adaptive” data are internal to the Marvell MCU (big “M”). This means that a DIY board swap is probably not feasible. :frowning:

However, if you can provide a detailed photo of the area around the Molex power connector, there may be a shorted TVS protection diode that can be removed.


#3

Hi fzabkar

http://picasaweb.google.co.uk/lh/photo/WdgR9tnqNfSh5K6PPc-pBA?feat=directlink

Looks to me like D3 may be burnt out.  Is it safe to remove this and operate it without or should I fit a new diode?  Any idea why this might short out?

Thanks again!


#4

D3 and D4 appear to be the 5V and 12V TVS diodes, respectively. Resistors R67 and R64 are the associated zero-ohm links. These resistors normally behave like fuses. Yours appear to be intact. If they were open, then your power supply would not have shut down.

Rather than removing components on suspicion, I would use a digital multimeter (US$10) to measure the resistances of each component (D3, D4,R67, R64) on the 200 ohms scale. The shorted diode will test close to zero ohms. You can then remove it by snipping its pins with flush cutters. Just remember that your drive will no longer have overvoltage protection on the affected supply rail. Hopefully the diode will have successfully contained the damage.

For continued protection, you can replace the 5V and 12V diodes with an SMAJ5.0A or SMBJ12A, respectively. Both are available from Farnell,Mouser, Digikey.

Best of luck.


#5

Al the components (D4, R64 and R67) are 0.4 ohms apart from D3 which is measuring 65 Kohms both ways.   So does this indicate that D4 has shorted?  Should the resistance of D3 be so high?  Without a circuit diagram it’s hard to know .  

The sites you mention all seem to have UK versions so I will see if I can get the replacement diodes!  

Thanks again for your help.


#6

D4 is definitely shorted. If you remove it, you should be good to go, provided that your PSU is OK, and assuming there is no other damage. I wouldn’t worry about D3, but you can test it on the diode or “continuity” range of your DMM. However this will measure the resistance of every device connected to the same supply rail.


#7

Many thanks fzabkar - I will give this a go!  :-)


#8

Many thanks again.  It worked and the defective drive is now in an external enclosure and I can rescue my data.  I may well fit a new diode and give it a second chance!

Is it likely that the Power Supply (COLORSit Silent Power M-400U) in the computer is unreliable?  This is a second computer and is seldom used and the Power Supply is only a year old as the original PS went bad and killed the original motherboard.  Could it be that the diode damaged when the original PS went bad and was doomed to fail eventually?

I know it’s probably impossible to say - I’m just wondering it it’s safe to use the computer with this Power Supply.  

Thanks again for your top advice!  :-)


#9

All you can say for sure is that your drive experienced an overvoltage on its 12V rail. That overvoltage had to come by way of the PSU. I don’t know anything about COLORS-IT, but it would appear that they rebrand generic PSUs. In your case I would watch the voltage fluctuations using a hardware monitor utility such as SpeedFan, or whichever software was supplied with your motherboard.

Here are just a few that Google found:

http://www.almico.com/speedfan.php
http://www.cpuid.com/softwares/hwmonitor.html
http://www.hmonitor.com/
http://openhardwaremonitor.org/

In particular, watch the 12V rail when doing CPU intensive tasks. In my experience, PSUs usually regulate by sensing a weighted average of the +5V and +12V rails. If the 5V supply drops under load, then the 12V rail rises. You will hear the fan speed increase when this happens.

You could use CPUBurn or Prime95 to torture test your CPU.

http://freshmeat.net/projects/cpuburn/
http://en.wikipedia.org/wiki/Prime95


#10

Thanks.  If I get that computer up and running again I will run the monitoring software - I think I’ll install it on my main computer as it looks very useful.

Thanks again.