• Pardon our dust as we work on some regularly scheduled forum maintenance. You may notice some missing features during this time. Thank you for your patience!

Question Mixed PSU cables lead to Suspect blown Seagate HDD diodes

trainsim

Reputable
Jul 1, 2016
10
0
4,510
0
Hey Everyone

So i made the classic ignorance error of mixing cables between power supplies. I used cables from a corsair rmx550 with a thermaltake tough power XT 875w. The result was when I tried to initial power on, the system flashed on then off, then on then off. I fried my h110i, and 2 of my seagate barracuda's now have a no spin condition. Based on reading and testing Iam assuming that I blew the diode/s but wanted to double check iam not a complete idiot and check some things over.

So the idea Iam getting with the diodes is they create temporary shorts to ground in the event of over voltage spikes ( e.g. supply instability, initial powerup etc) however in the event of sustained over voltage they can blow, and at rest will be a constant short to ground. In this event, doing continuity tests between input pins and output pins I should be seeing total or near total continuity to ground, right?

My theory was always that the 12v and 5v got switched, in theory in this event, 5v to the 12v diode would produce nada, no shorts etc since insufficient voltage. Howeer 12v to 5v line should result in the diode 'blowing' and shorting to ground.

Testing what I assume is the 12v diode (it tracks to the 12v pins), I still get a diode reading, meanwhile the 5v diode ( marked AE and has continuity to 5v pins and common) is a total short.

So, In order to recover my data, in theory all I need to do is, desolder the 5v diode then use a truthworthy supply etc to pull the data off the drives.

Am I on the right track here or am I way off? Also from reading I know it can be a real pain to desolder surface mount components with a solder iron vs hot air station. and guess which I have :\ Can supply photos of the area Iam looking at if interested.
 

molletts

Distinguished
Jun 16, 2009
480
3
19,165
146
I can't see that removing the diodes and testing with a known-good PSU will do any more damage than has already been done.

If your soldering iron has a very fine bit, it shouldn't be too difficult to desolder the surface-mount diodes. It can help to add a little extra solder to each end first to improve the thermal interface between the iron and the joint. You may need to melt both joints first then rapidly touch the iron to each end in turn to try and get them both to be molten at the same time, while lifting the diode very gently with some fine tweezers (be very careful not to lift the PCB track or bridge solder to an adjacent part). Another trick I've used in the past has been to solder a short U-shaped piece of fairly thick, bare solid-core wire across the component (only just long enough to reach both ends), then heat the middle of the wire with the iron while lifting the component. The wire conducts the heat into both joints so that they both melt at the same time.

If you have another working drive of exactly the same model and are prepared to risk it (back it up first!), it may be possible to swap the controller board onto the dead drives. It depends on how the drive stores its calibration & configuration data - in flash memory or on reserved tracks of the disk. I've done this with a few drives in the past and it's worked, although I've only used the resulting frankendrives for long enough to recover the data and have avoided writing to them at all (use a Linux recovery environment and mount the drives read-only - Windows will write to drives even if you don't save or copy anything to them).
 
Reactions: trainsim
You can remove the shorted 5V diode with flush cutters (or fingernail clippers).

Check the associated zero-ohm resistors. If they are open circuit, then flow a blob of solder over them.



Note that for this model you must transfer the "ROM" chip from patient to donor when replacing PCBs. You must not damage the ROM, as there are only a few data recovery shops which can recover a Seagate drive with a dead ROM.
 
Last edited:
Reactions: trainsim

trainsim

Reputable
Jul 1, 2016
10
0
4,510
0
You can remove the shorted 5V diode with flush cutters (or fingernail clippers).

Check the associated zero-ohm resistors. If they are open circuit, then flow a blob of solder over them.



Note that for this model you must transfer the "ROM" chip from patient to donor when replacing PCBs. You must not damage the ROM, as there are only a few data recovery shops which can recover a drive with a dead ROM.
Thanks for that. And yes the resistors were blown as well, I ended up removing them and joining the pads with a blob of solder. Both drives were recovered. Ill also be helping a friend of mine who suffered a similar fate he has 2 seagates and 1 western digital, Ill be learning the circuit path for that as well, much obliged guys.
 

ASK THE COMMUNITY

TRENDING THREADS