Incorrect Molex Install has Killed Four Drives!

[ekim]

Happy to take some measurements again if they don't seem to make sense with what can be seen.

If the preamp is damaged does this make the DIY route unlikely to work (i.e. removing the diode etc.).

[fzabkar]

I would have expected R67 to be open circuit on PCB #1 (the one with the damaged motor controller). Instead your measurement shows 0 ohms.

A dead preamp means that the game is over for you.

See http://hddscan.com/doc/HDD_from_inside.html

[ekim]

I'll test PCB #1 again tonight. I do recall being surprised that it was coming back at 0 as that was the most obviously damaged board. Did some reading on the preamp so understand that if that's gone I'm at a loss.

I have four new drives due to arrive to accept any data I am able to recover. Going to see if I can get the electronics repair store to help with the diode removal and resistor soldering. I also have a spare PSU here that I can use to test the PCBs once they have been modified.

[colanco]

En el caso de los PCB # 1, las mediciones de la OP parecen ser incompatibles con la evidencia visual.

Can be correct and be an indirect measure (through preamp or VCM)

[ekim]

I'm getting the diodes removed tomorrow and the resistor soldered over.

Just want to check. I am assuming when the diode has been removed we need to add a bolb of solder to close the connection?

[jermy]

No, just remove the diode, but be careful to use a good PSU

[ekim]

Thanks! Will report back if everything works. Will be trying a PCB swap if I don't have any success.

I have a Seasonic X660 PSU which is quite high quality so happy that it should send clean power to the drives.

[ekim]

Sorry another question!

I have got a spare PSU set up and ready to test the PCBs after the modifications tomorrow. With the voltages that were suggested by fzabkar how do I test them with my multimeter?

What connections should the cables be plugged in to and what setting should I be using? I assume I just touch a screw hole with one of the probes and then the other to the places suggested to measure?

[colanco]

If you only remove the TVS and do not replace them, you do not need to repair R67 (R43). Would be useless.

If you replace TVS, then you would need to repair R67 (or R43)

[fzabkar]

If you only remove the TVS and do not replace them, you do not need to repair R67 (R43). Would be useless.

If you replace TVS, then you would need to repair R67 (or R43)

R43 and R60 are in series with the +5V and +12V supplies. They must be bridged. These resistors bridge the Vin and Vout pins of their respective unpopulated e-fuses.

[ekim]

The diode has been removed on each board and the resistors bridged.

I can't actually test the PCBs on the HDDs for another couple of days, but will report back once I have had a chance.

Attachments

Board 2

Board 3

[colanco]

If you only remove the TVS and do not replace them, you do not need to repair R67 (R43). Would be useless.

If you replace TVS, then you would need to repair R67 (or R43)

R43 and R60 are in series with the +5V and +12V supplies. They must be bridged. These resistors bridge the Vin and Vout pins of their respective unpopulated e-fuses.

Apologies for the error, what I wanted to say was "R67 (or R64)"

[ekim]

Attached the Boards 2 and 3 to the Hard Drives this evening. Both the drives were detected fine by my server and it appears that all of the data was intact.

The drives were used in an Unraid server with dual parity so I was able to recover the data on the other two drives (Boards 1 and 4). I have removed the other two drives (Boards 2 and 3) from the server now and a rebuilding the array.

Thanks for all your help! Have a good Christmas.

[fzabkar]

Attached the Boards 2 and 3 to the Hard Drives this evening. Both the drives were detected fine by my server and it appears that all of the data was intact.

Boards 2 & 3 were designed correctly while 1 & 4 were not. Congratulations on your good luck.