Power Supply mixup has killed all my hard drives!

[bigi]

Late last night (or should I say early this morning!) I noticed my tower had died. turns out the EVGA power supply has completely died. In my half asleep state I tried another module PSU which powered the board back up; I then made the mistake of connecting the hard driver power cables up to the new power supply without checking the pinouts were the same (even though the connectors where).....result was a very warm molex cable and no bootup on the pc.

I've checked the drives in another pc and two external caddies and they are dead, no power up, clicking, nothing. No Damage visible on any of the boards either.

However, reading fzabkar's guide about TVS diodes I pulled the boards and tested them - they all seem happy, but on the WD Green drive appears to have a dead zero ohm resistor on the 5v line (see first photo). The Toshiba drives are all the same board, and again the TVS diodes seem fine, but there are a couple of strange components which i'm suspecting are polyswitches - again the 5v line appears to have failed. on the Toshiba drives, these are labelled "3" (on the 5v line and appears dead) and on the 12v line they are marked either "4" or "S".

Before i short the links across and hope they work, can anyone have a look and confirm if i'm in the right ballpark? i've tested both boards in an external caddy without the drive connected and tested and they are all getting 5v and 12v in, and the 12v line is at various points around the PCB (fuses), but no sign of the 5v line after the suspect components.

Any help would be really appreciated!

Western Digital PCB

Toshiba PCB

[fzabkar]

Are you sure D3 is OK on the WD board? What is its resistance?

The "3" and "4" parts are probably fuses. I wouldn't bridge them. There appears to be no protection downstream from these fuses, so I'm expecting that there will be damage to the 5V electronics and/or the preamp. BTW, what are the resistances of diodes "72" and "XB"?

[bigi]

Thanks for the reply. for some reason checking D3 this time results in showing it short circuit - so i should be ok removing that and shorting the zero ohm resistor?

With regards to the Toshiba - only 1 fuse has gone (on the 5v line) - is there not a TVS diode protecting this circuit as well?

[fzabkar]

for some reason checking D3 this time results in showing it short circuit - so i should be ok removing that and shorting the zero ohm resistor?

That should be OK. Just power up the board on its own to be sure there are no fireworks.

With regards to the Toshiba - only 1 fuse has gone (on the 5v line) - is there not a TVS diode protecting this circuit as well?

I'm finding it hard to believe, but the "72" part looks like an RB160M-20 Schottky barrier diode, while the "XB" part appears to be a 15V, 1W Zener diode. The RB160M diode does not provide overvoltage protection. :-(

Could you upload a photo of the entire Toshiba PCB so that we could could test for further damage? That said, I suspect that the preamp (on the headstack) may be dead.

RB160M-20, ROHM, 20V, 1A, Schottky barrier diode, marking 72, SOD-123/1206:
http://www.smdmark.com/en-US/smd-model-RB160M.html

http://docs-asia.electrocomponents.com/webdocs/13e4/0900766b813e45a3.pdf (RB160M-30, marking 73)

http://rohmfs.rohm.com/en/products/databook/datasheet-nrnd/discrete/diode/schottky_barrier/rb160m-40.pdf (RB160M-40, marking 74)

https://www.buerklin.com/medias/sys_master/download/download/h72/h2d/8891259879454/technical-data-sheet-rohm-semiconductor-rb160m-60tr-en-20160314.pdf (RB160M-60, marking 76)

http://www.farnell.com/datasheets/24626.pdf (RB160M-90, marking 83)

KDZTF-15B, Rohm, 15V, 1W Zener diode, marking XB:
http://rohmfs.rohm.com/en/products/databook/datasheet-nrnd/discrete/diode/zener/kdztftr9.1b-e.pdf

[bigi]

Thanks for the reply. Typical thy 5 of my drives are the Toshiba ones . I’m assuming by headstack you mean inside the drive? I thought (read hoped) that the blown fuse was actually a poly switch that had gone past it’s breakdown voltage.

I’ve attached images of front and back of Toshiba PCB

Cheers again for the advice. If what I’m understanding from what you are saying, the drives are totally a scrap?

Attachments

[fzabkar]

@bigi, your original photo was excellent, but your latest photo was a disappointment. Is there any chance you could provide better shots of the component side of the Toshiba PCB, even if you have to break them up into smaller sections? The bottom side is OK.

I have determined how the 5V supply gets to the preamp (IC17 & Q600). Fortunately the power is switched -- it just remains to be seen whether the surge got past the switching IC. Better photos would help a lot.

[bigi]

@fzabkar apologies for the poor photos- late at night is never a good time!
Upon further investigation it appears I may be lucky-ish!
IC17 looks like part is missing. I have taken some more photos, hopefully these are better!

Attachments

[bigi]

@fzabkar, I've had chance to take a few more magnified photos in better light, as well as photos of a brand new working PCB from and identical drive for reference. Hope this helps more - you can clearly see a corner of the IC has been blown off it; I'm hoping this is where the damage ends! Please let me know if you need anything more.

(Part 1/3)

Attachments

Old PCB

Old PCB

Old PCB

Old PCB

Old PCB

Old PCB

Old PCB

Old PCB

Old PCB

Old PCB

[bigi]

Part 2/3

Attachments

New PCB

New PCB

New PCB

New PCB

New PCB

New PCB

Old PCB

Old PCB

Old PCB

[bigi]

Part 3/3

Attachments

New PCB

New PCB

New PCB

New PCB

[bigi]

....BTW, what are the resistances of diodes "72" and "XB"?

Sorry, I just saw this request!

Diode "72" fwd bias 0.85K, rev bias (above 2k diode range)

Diode "XB" fwd bias 0.776k, rev bias (above 2k diode range)

Finally, the small component next to IC7 which looks like a fuse is open.

[fzabkar]

Thanks, bigi, those are very good photos. I'm now doing some research to track down the component IDs, so I'll get back to you soon. IC17 appears to be a "high-side load switch" but I'm having a tough time identifying it.

The good thing is that you have a working PCB, so we can take measurements from it for comparison purposes.

[fzabkar]

User "lovelove" in the following thread states that, after an overvoltage, "a small black square [went] boom".

http://forum.hddguru.com/viewtopic.php?f=1&t=31612&p=255064#p255064

It appears to be the same part as your IC17 (high-side load switch). Your chip also appears to have sustained physical damage.

Here is how that area of the circuit looks to me:

Code:

           3A                .---------.
  SATA    fuse               |         |
  +5V       _           5Vin |    /    |     5Vout
   >--+--o_/ \o----+---------|---/ ----|---->to preamp
      |            |         |    |    |
      -  "72"      |         '----|--|-'
      ^  D702     .-.       Off/On|  |  IC17
      | RB160M-20 | |             |  |
      | Schottky  | |             |  |  high-side
     === diode    '-'             |  |  load switch
     GND           |              |  |
                   +--------------'  |
                   |                 |
      .------------|--.              |
      |            |  |              |
On/Off|            |  |             ===
from  |  ___     |/   |             GND
   >----|___|-+--|    |
MCU   |       |  |>   |
      |      .-.   |  |
      |      | |   |  |
      |      | |   |  |
      |      '-'   |  |
      |       |    |  |
      |       '----+  |
      '------------|--'
        Q600       |
      DTC123J      |
                   |
      digital     ===
     transistor   GND
      "E42"

The preamp appears to be powered from two supplies. One is the +5Vout from IC17 (to HDA pins 16/18), while the other appears to come from the vicinity of the motor controller IC (to HDA pin 19).

I would start by measuring the resistance between ground (eg a screw hole) and each of the 5Vin and 5Vout test points. Do this with the board off the drive. You can compare the good and bad boards.

Next I would install the PCB on the drive and perform the same measurements. You will need to locate the corresponding test points on the underside of the PCB. These measurements will tell us whether the preamp is shorted or stressed.

Attachments

Last edited by fzabkar on January 26th, 2018, 2:18, edited 1 time in total.

[fzabkar]

Datasheets

TPS27081A, Texas Instruments, 1.2V - 8V, 3A PFET High-Side Load Switch with Level Shift & Adjustable Slew Rate Control, marking AUA, SOT-6:
http://www.ti.com/lit/ds/symlink/tps27081a.pdf

DTC123J, Rohm, NPN digital transistor, 50V, 100mA, marking E42:
http://www.rohm.com/web/global/datashee ... 123jebtl-e

H5PS1G63KFR-G7C, SK hynix, 1Gbit, 64Mx16, DDR2 SDRAM, 1.8V:
https://www.skhynix.com/product/filedat ... o?seq=8138

BD91364AMUU, Rohm, step-down converter, 2.9V to 5.5V, 5A, Integrated MOSFET, VQFN20U4040M package:
http://rohmfs.rohm.com/en/products/data ... amuu-e.pdf

74LC138, 3-to-8 Line Decoder / Demultiplexer:
http://ecee.colorado.edu/~mcclurel/sn74ls138rev5.pdf

W25Q80BL, Winbond, 8M-bit 2.5V Serial Flash Memory with uniform 4KB sectors and Dual/Quad SPI:
https://www.winbond.com/resource-files/ ... 80blf1.pdf

LE25U81AQE, ON Semiconductor (Sanyo?), 8Mbit, 2.5V, serial SPI flash memory, VSON8T package:
http://www.onsemi.cn/pub_link/Collateral/ENA2199-D.PDF

EM68C16CWQD-25H, Etron Tech, 64M x 16 bit, DDR2 SDRAM, 1.8V, FBGA:
http://www.etron.com/manager/uploads/EM ... %201.5.pdf

BD3851, Rohm, op amp for shock sensor

[fzabkar]

I believe these are the preamp supply voltage test points:

Attachments

Preamp_voltage_test_points.jpg (109.8 KiB) Viewed 18562 times

[fzabkar]

I would measure the Vcore, Vin and Vout voltages on a good PCB (taking care not to short adjacent pins).

There is a second high-side load switch (TPS27081A) whose function I can't determine. I was hoping that IC17 was also a TI part, but the device markings don't seem to match anything from TI. The other thread identifies the markings on IC17 as "63AN 78" whereas yours appear to be "767N 78". This would suggest that "78" is the device code while the first four characters are perhaps batch and/or date codes.

Attachments

TPS27081A_pinout.gif (8.45 KiB) Viewed 18554 times

BD91364AMUU_pinout.gif (14.98 KiB) Viewed 18554 times

[fzabkar]

I found it. IC17 is a TPS22912C.

TPS22912C, Texas Instruments, Ultra-small, Low On Resistance Load Switch With Controlled Turn-on, Vin = 1.4V - 5.5V, 2A, 60mohm ON-resistance, marking 78:
http://www.ti.com/lit/ds/symlink/tps22913.pdf

Attachments

[fzabkar]

I'm guessing that there are 3 supplies generated by the motor controller. V1 and V2 would probably be Vio rails, possibly 1.8V and 2.5V or 3.3V.

L500 and L602 appear to constitute a Cuk converter which generates a negative supply for the preamp (Vneg), probably -5V. This could be the "unknown supply voltage" which I previously identified.

Attachments

[bigi]

@fzabkar Thank you for all hard work helping me with this! I've finally had chance to do all the tests on old and new boards, annotated pictures are attached. It seemed to make more sense on the images than trying to write them down here! A couple of the readings need some further explanation; the measurements for the +5v fused resistance were hard to determine accurately; on the old PCB the measurement stabilised slightly but was constantly increasing as i monitored it. On the new PCB, I was unable to determine a measurement at all - my auto ranging multimeter seemed to show the value was constantly going up and down. on the fused +12v test, the old PCB value kept dropping, whereas the good PCB was stable.

Attachments

VCore

Switching Circuit

Preamp Voltage Test Points

Preamp_voltage_test_points.jpg (64.37 KiB) Viewed 18402 times

HDA Connector

[fzabkar]

@bigi, could you please clarify your situation? Do you have a working spare HDD or just a good PCB?

Were your measurements on the good and bad PCBs taken when both were off the patient drive?