[SOLVED] Used old PSU cables on 34TB of drives and they are no longer recognized

orangebear

Prominent
Dec 17, 2021
12
0
510
I was a dummy and swapped all cables on my new PSU except the SATA drives. All three drives no longer show in BIOS. They are shucked Western Digital Easy Drives. After some research, it seems it could be the fuses, the diodes, or I might just be screwed. Interestingly, all three HDDs provided different readings from the multimeter, which confuses my next step further. I've attached pictures with the readings. I hope somebody could provide further guidance.
 
Solution
Fuses should measure 0.0 ohms or close enough. Therefore all except the 12V fuse in the first photo are open circuit.

The 5V TVS diode in the second photo is shorted. All other diodes appear to be OK. The 0.18V or 0.19V readings are due to the low forward voltage drops of the Schottky diodes (the smaller diode in each pair).
Fuses should measure 0.0 ohms or close enough. Therefore all except the 12V fuse in the first photo are open circuit.

The 5V TVS diode in the second photo is shorted. All other diodes appear to be OK. The 0.18V or 0.19V readings are due to the low forward voltage drops of the Schottky diodes (the smaller diode in each pair).
 
  • Like
Reactions: orangebear
Solution

orangebear

Prominent
Dec 17, 2021
12
0
510
Fuses should measure 0.0 ohms or close enough. Therefore all except the 12V fuse in the first photo are open circuit.

The 5V TVS diode in the second photo is shorted. All other diodes appear to be OK. The 0.18V or 0.19V readings are due to the low forward voltage drops of the Schottky diodes (the smaller diode in each pair).

So I need to bridge all fuses except the 12V fuse in first photo and remove the 5V TVS diode in the second photo? Sorry if if that sounds like nonsense, this is my first time attempting anything like this.
 

orangebear

Prominent
Dec 17, 2021
12
0
510
Yes, but you need to be sure that there are no other problems. Of course, you also need to be absolutely sure that your current cables have the correct pinout.

Good luck.
thank you for your prompt replies. I will never in this lifetime or the next use the wrong pinouts ever again.

As for other problems: can you point me towards a resource to help me ensure there arent other issues with the boards/drives? Would it be useful to post further pictures of the boards? There was no glaring physical damage on first review.

Thanks again
 
In the future, as you've already noted but I'm going to reinforce for you, don't EVER use ANY cables with ANY power supply that the cables didn't come with. Not unless you've ordered them specifically for that unit, such as custom sleeved cables, or they came from an identical model that you know for certain has not had any pinout changes between manufacturing dates because you checked, or compatible models that you've checked and validated compatibility through the manufacturers compatibility or pinout data for.

In general, just use what comes with them. It's really a very small amount of time to change any cable. Far less time than it takes to verify compatibility through online resources or to fix a problem you created by not checking at all and reusing the old ones. You know this now, but others who come along and find your thread might not "get it" without us being very clear as to why this is a problem and the problem of course is that from manufacturer to manufacturer, or even within the same manufacturer's product lines, or EVEN in some cases within the same series but from different capacity models, the pinouts AT THE PSU might have one or more pins that are entirely different. That means you might be putting voltage straight to ground or the wrong voltage where it doesn't belong, or no voltage where you need there to be voltage. All of which might cause various different levels of "aw crap".

Thanks to fzabkar for helping this user out. While I don't always agree with the provision of some of your test methods to some users simply due to the obviousness of their inability to properly and safely carry them out, there are certainly cases where your assistance proves to be particularly valuable to some more capable users who just need that little bit of unknown information to get them over the hump. So, just wanted to say that in this case, and cases LIKE this, it is noticed and is I'm sure greatly appreciated by those you've helped even when the end result is "aw crap, it's dead Jim". :)
 
  • Like
Reactions: orangebear

orangebear

Prominent
Dec 17, 2021
12
0
510
In the future, as you've already noted but I'm going to reinforce for you, don't EVER use ANY cables with ANY power supply that the cables didn't come with. Not unless you've ordered them specifically for that unit, such as custom sleeved cables, or they came from an identical model that you know for certain has not had any pinout changes between manufacturing dates because you checked, or compatible models that you've checked and validated compatibility through the manufacturers compatibility or pinout data for.

In general, just use what comes with them. It's really a very small amount of time to change any cable. Far less time than it takes to verify compatibility through online resources or to fix a problem you created by not checking at all and reusing the old ones. You know this now, but others who come along and find your thread might not "get it" without us being very clear as to why this is a problem and the problem of course is that from manufacturer to manufacturer, or even within the same manufacturer's product lines, or EVEN in some cases within the same series but from different capacity models, the pinouts AT THE PSU might have one or more pins that are entirely different. That means you might be putting voltage straight to ground or the wrong voltage where it doesn't belong, or no voltage where you need there to be voltage. All of which might cause various different levels of "aw crap".

Thanks to fzabkar for helping this user out. While I don't always agree with the provision of some of your test methods to some users simply due to the obviousness of their inability to properly and safely carry them out, there are certainly cases where your assistance proves to be particularly valuable to some more capable users who just need that little bit of unknown information to get them over the hump. So, just wanted to say that in this case, and cases LIKE this, it is noticed and is I'm sure greatly appreciated by those you've helped even when the end result is "aw crap, it's dead Jim". :)
All fair and valid points. I hooked up the hardware late-night, went to sleep, woke up the next morning and started trying to power up and configure the server before verifying my work from the night before. Twas dumb.

I am clueless in the electrical/board repair arena but i'm willing to learn. I saw some of fzabkar's previous posts and hoped I might get a response and he is very nice to provide guidance. I told my wife about his help and she was so excited after seeing the daze i've been in since this happened.
 
No, you are not clueless. You might not have much experience, but I assure you, I've seen "clueless" hundreds of times here, and you are not that by a long shot. The fact that you ARE able to take fzabkar's information and put it to use refutes that idea. Most simply get that deer in the headlights look and then move on to some halfway version of what was recommended in the instructions they were afforded, which usually results in more "bad things" happening. Fortunately, for you, this does not seem to be the case, in this case. :)
 
  • Like
Reactions: orangebear

orangebear

Prominent
Dec 17, 2021
12
0
510
No, you are not clueless. You might not have much experience, but I assure you, I've seen "clueless" hundreds of times here, and you are not that by a long shot. The fact that you ARE able to take fzabkar's information and put it to use refutes that idea. Most simply get that deer in the headlights look and then move on to some halfway version of what was recommended in the instructions they were afforded, which usually results in more "bad things" happening. Fortunately, for you, this does not seem to be the case, in this case. :)
Thank you for the kind, encouraging words. I figured when this happened that if we are going down we are going down swinging.
🤞
 

orangebear

Prominent
Dec 17, 2021
12
0
510
I'll start with the good news: the fuses were replaced in drive 2 and 3 and both are now showing in BIOS!!! Thank you @fzabkar ! I will post dimensions and part numbers shortly for others in the same spot in the future.

The one piece of help I do still need: the tvs diode in drive 1 was replaced but it still showing a short. @fzabkar I was curious if you could tell me what else I can test on that board to make sure there aren't other problems? I have attached photos.
 
The worst case scenario, assuming that the fault is confined to the PCB, is that you will need to purchase a replacement PCB and transfer the ROM chip.

http://www.users.on.net/~fzabkar/HDD/0B40829/ROM.jpg
https://ae01.alicdn.com/kf/H4d7117a...-12TB-circuit-board-006-0B40829-PCB-board.jpg ("BIOS" = ROM)

Otherwise, you could test the resistances of every capacitor. This will tell us which of them is connected to the +5V supply. Then we can concentrate our efforts in the appropriate area.


Edit: I suspect that this part is shorted:

LX8237, Microsemi, PMIC with 5V & 12V E-Fuses and Current Monitoring for HDD and SSD, Vcore regulator (0.94V), 3.3V buck regulator, QFN-24L:
https://www.microsemi.com/document-portal/doc_download/135299-lx8237-datasheet

I have written it up here:

http://www.hddoracle.com/viewtopic.php?f=100&t=1615&p=22339#p22339

I suggest that you measure the resistance between ground (eg a screw hole) and each of +5Vin, +5Vout and +5Vout (fused).
 
Last edited:
  • Like
Reactions: orangebear
I tried to find the LX8237 and STPMIC02 (an equivalent IC), but both are unobtanium. :-(

Minimum order quantity for LX8237 = 3000, lead time = 52 weeks:
https://octopart.com/search?autosugg_idx=0&currency=USD&oq=lx8237&q=lx8237ilq-tr&specs=1

This vendor will sell you single pieces of STPMIC02, if you can work out how to navigate the web site:

https://detail.1688.com/offer/559018879658.html?spm=a261b.11197608.mof001.1.54d01a2a8kMPz9#

The package markings would indicate that the reel was produced for HGST:

https://cbu01.alicdn.com/img/ibank/2017/737/417/4720714737_1642394629.jpg

"0J45335" is a Hitachi part number. Your WD drives are actually rebadged HGST models.
 
Last edited:
  • Like
Reactions: orangebear

orangebear

Prominent
Dec 17, 2021
12
0
510
I tried to find the LX8237 and STPMIC02 (an equivalent IC), but both are unobtanium. :-(

Minimum order quantity for LX8237 = 3000, lead time = 52 weeks:
https://octopart.com/search?autosugg_idx=0&currency=USD&oq=lx8237&q=lx8237ilq-tr&specs=1

This vendor will sell you single pieces of STPMIC02, if you can work out how to navigate the web site:

https://detail.1688.com/offer/559018879658.html?spm=a261b.11197608.mof001.1.54d01a2a8kMPz9#

The package markings would indicate that the reel was produced for HGST:

https://cbu01.alicdn.com/img/ibank/2017/737/417/4720714737_1642394629.jpg

"0J45335" is a Hitachi part number. Your WD drives are actually rebadged HGST models.
Thanks for all your work. After getting two of the boards working I tabled the project for awhile. You are the reason I got my data back, thanks again