(Diode) Measurement electronic parts - WD30EFRX - oversurge?

[BenDataRecovery]

Hello everyone,

looking for a way to find the "bad parts" of the controller board. The diodes D2 and D4 both way letting through current 0 Ohm both ways. Diodes D1/D3 seems to work... 1,45 Kilo Ohm both of them and 0 Ohm other side.

Any further measurement points? Parts to be replaced - diodes cutting D2/D4? Replacing them?

Thanks for your help
Benjamin

Attachments

to describe further measurement points

the diodes as prescribed

[pepe]

no point in trying to find bad components, i guess everything is burnt.
WD's whole surge protection circuit is conceptually shit on the pcbs i checked so far (so i suppose they use the same bad circuit on all their drives).
guess the 0 ohm resistors are burnt open, if so, the circuit behind the protectors were left unprotected and got damaged by the surge.
I expect the preamp to be burnt as well, so replacing the PCB without checking the preamp is a risk.
It is your turn, Franc

[BenDataRecovery]

0 ohm resistors

Diodes!

[pepe]

please read me a bit more carefully and also check the attached pic.

there are 0 ohm resistors inseries with the TVSes, which is a conceptual mistake in this design. If they were between the TVS and the supply connector, they would function as fuses, but in this arrangement when they burn open, the whole circuit is left unprotected.
So the protection is practically useless on WD boards.

[fzabkar]

I don't see any visible damage. In particular the motor controller doesn't appear to have any burn marks, so this would suggest that the zero ohm resistors may be intact. If so, then simply removing the shorted diode(s) should be enough to get the drive working.

Otherwise you would need to transfer the "ROM" at U12, or its contents, to a replacement PCB. Some PCB suppliers (eg hdd-parts.com) include a transfer service in the price.

Catastrophic failures in Western Digital PCBs:
http://www.hddoracle.com/viewtopic.php? ... 119&p=5033

HDD TVS diode FAQ:
http://www.users.on.net/~fzabkar/HDD/TVS_diode_FAQ.html

[pepe]

let's have the OP measure the resistors and get back with the results.

[BenDataRecovery]

If so, then simply removing the shorted diode(s) should be enough to get the drive working.

That worked! Thanks for your help! And thanks for your help Pepe too.

[pepe]

great.

i just wanted to point out that the protection on these boards is really weak, they placed the resistors to the wrong place, and when diodes get shorted, the resistors burn easily, then the whole underlying circuit is exposed to the incoming overvoltage...
And they at WD have not realized the bad design in this circuit for at least 8 years ... eh

[BenDataRecovery]

i just wanted to point out that the protection on these boards is really weak, they placed the resistors to the wrong place, and when diodes get shorted, the resistors burn easily, then the whole underlying circuit is exposed to the incoming overvoltage...
And they at WD have not realized the bad design in this circuit for at least 8 years ... eh

Thanks for your help and your background information. Seems that you have good skills of measurement electronical parts and architecture?

Greetings Benjamin

[pepe]

I think it is very basic electronics, nothing complicated. Electronics start somewhere else...

[BenDataRecovery]

I think it is very basic electronics, nothing complicated. Electronics start somewhere else...

[mordros]

I don't see any visible damage. In particular the motor controller doesn't appear to have any burn marks, so this would suggest that the zero ohm resistors may be intact. If so, then simply removing the shorted diode(s) should be enough to get the drive working.

Otherwise you would need to transfer the "ROM" at U12, or its contents, to a replacement PCB. Some PCB suppliers (eg hdd-parts.com) include a transfer service in the price.

Catastrophic failures in Western Digital PCBs:
http://www.hddoracle.com/viewtopic.php? ... 119&p=5033

HDD TVS diode FAQ:
http://www.users.on.net/~fzabkar/HDD/TVS_diode_FAQ.html

Hello fzabkar!

First off, thanks for all of the time and help you put into these boards. While I only now just made an account to post, I've read through quite a few posts and your information is helpful.
Backstory: I accidentally shorted two HDDs when I plugged the wrong (modular) power SATA cable into my PSU. I thought all SATA power cables were universal with the 6-pin side that connects to the PSU, however I guess I was wrong. Seagate and EVGA apparently have different internal wirings…
I’ve already fixed one of the HDDS by identifying and removing a shorted TVS on the PCB thanks to your other posts. (Hitachi HDD)
The remaining HDD is a WD Red, 3TB drive (WD30EFRX), same as the OP. When I originally troubleshooted:
1. I found that both D3 and D1 were shorted. D2, D4 are both OK.
2. As you can see in the picture, I removed D3 by…mechanical means. After removing D3, I decided to double check D1, and noticed to my surprise that it was no longer shorted ( ??)

I decided to leave D1 intact and attempt a power-up. Zero response from the drive, same as before.

Try as hard as I might, I could not locate the famous R64 and R67 that are mentioned for other WD HDDs. Could R60 and R43 be the possible equivalents on this model PCB?

As it stands now, R60 is closed across it’s terminals. R43 is open across it’s terminals. (This is unchanged from before I removed D3, as I had checked them before modification just in case.)

What confuses me is why D1 was shorted before I removed D3, and now it appears OK. Do you have any ideas? Should I remove D1 as well as D3 (like the OP did), or, remove R43 and somehow bridge the terminals (assuming R43 is some sort of equivalent of R64 or R67, and is now blown)?

Any help would be appreciated!

Attachments

[pepe]

yes, R60 and R43 is what you are looking for.
Strange enough, it seems WD finally corrected the bad pcb layout they have been using for 10 years and decided to protect the circuit instead of the TVSes.
And why you measured D1 as shorted before removing D3: it was simply because they are wired in paralel, which means you can't really decide which one is shorted until you remove one of them.

[fzabkar]

@mordros, I would clean up the solder splash around D3.

After you sort out R43, I would also clean the HDA contacts.

Oxidisation on Western Digital PCBs:
http://www.hddoracle.com/viewtopic.php? ... 649&p=1789

If you wish to test the onboard power supplies, let me know. Otherwise I would power up the board off the drive to ensure that all is OK.

BTW, the unpopulated locations at U4 and U3 have been reserved for STEF12 and STEF05 electronic fuses. Just today I've been looking at a Seagate ST3000DM001 PCB with an unpopulated 5V e-fuse, so it looks like we might be seeing them in future.

[day1data]

Just be careful using the drive now with the TVs chip missing. I would buy a new PCB for that drive and transfer the U12 chip.

There is no protection from a second surge, eg. Preamp getting killed.

[colanco]

WD finally corrected the bad pcb layout they have been using for 10 years and decided to protect the circuit instead of the TVSes.

This design (TVS + 0 resistors) is common in the WD red series, but not in the green/blue

[pepe]

I refered to their old design, which was worth practically nothing, as they put the resistors (fuses) in series with the TVSes towards the ground, so when the tvs became shorted (or just activated by the oversurge), the fuses blew and let the tvs float with the incoming overvoltage, thus providing absolutely zero protection for the circuit beneath it .
on this pcb they put the resistors to the correct place where they can protect the circuit.

like this:
old circuit (does not protect anything after R is blown):

Code:

---------------------->to protected circuit
    |
  TVS 
    |
   R
    |
-------------------->GND


new circuit (protected circuit is completely isolated when R is blown open):

Code:

---R---------------->to protected circuit
        |
      TVS
        |
-------------------->GND


( Edit: i had to put the diagrams in code brackets because the leading spaces were removed otherwise...)

And they used the first circuit for over 10 years on all of their desktop and notebook drives (at least the models i checked).
pepe

[fzabkar]

The OP's protection circuit is even more stupid.

Code:

15 o----.
        |
14 o----+-- R --+---o
        |       |
12V     |       |
        R      TVS
        |       |
        |       |
13 o----'       R
                |
                |
               ===
               GND

[pepe]

yeah, i haven't seen that till now. i wonder if they have too many resistors and want to use them somewhere or what? ...

[fzabkar]

yeah, i haven't seen that till now. i wonder if they have too many resistors and want to use them somewhere or what? ...

Maybe they could use them here: :P