Hi All, I'm posting a follow-up to the discussion I started a few days ago  regarding slow parity sync speeds using my Adaptec ASR 7805 HBA controller. I figured the best course of action for testing would be to connect my 3 WD 14tb drives directly to the SATA ports on my Supermicro X11SCQ to test if the read/write speeds were still an issue.

This led me to a new issue - when plugged directly into the motherboard via SATA cables, the HDDs are unrecognized in both BIOS and unRAID.


Then, I figured maybe the SATA ports on the motherboard were bad, so I grabbed an extra WD 1TB Blue drive I had laying around to see if that would be recognized in the BIOS when plugged into the SATA port, and sure enough it was. I even tested swaping around the SATA cables between the working 1TB WD Blue and the non-working WD 14TB drives and still only the 1TB Blue drive was recognized via BIOS.


Why might this be happening? Do I need to update the BIOS at all? It seems super odd that the 14TB drives are recognized when plugged into my HBA adadpter but not when plugged directly into the SATA ports on the motherboard, even though a regular 1TB WD Blue is recognized. For reference, here's my HBA setup that is at least recognized by the BIOS and unRAID (albeit at 3MB/s read and write speeds).

25 minutes ago, chanzer0 said:

Do I need to update the BIOS at all?

If an update is available I would.

 

If you default the bios settings, will it detect the HDDs? 

 

Is their FW available for the WD drives?

Resetting the BIOS settings does not detect the 14TB WD drives, but the 1TB WD blue is recognized in both BIOS and unRAID. I'm beginning to suspect the refurbished drives I bought were shucked or otherwise need the 3.3v mod done to them. In theory, if the 14TB WD drives were working the the Adaptec HBA card and break-out cables, perhaps that was bypassing the 3.3v issue that these drives sometimes have? so when I plug them directly into SATA power and SATA ports, they aren't powering on at all?

23 minutes ago, chanzer0 said:

In theory, if the 14TB WD drives were working the the Adaptec HBA card and break-out cables

Were you using different power cables?

That's an interesting idea that they may be shucked.  Why they would show up on the HBA and not direct sata... assuming power was coming from the same place, I got no clue.  Regarding the 3.3v mod, what would be the damage to a drive if the mod is done and it did not need it?  If no harm, then give it a shot.  I dont know how that would work either.

36 minutes ago, JorgeB said:

Were you using different power cables?

Negative, the same white cables pictured were used with both the HBA and the direct connection.

 

 

34 minutes ago, Veah said:

That's an interesting idea that they may be shucked.  Why they would show up on the HBA and not direct sata... assuming power was coming from the same place, I got no clue.  Regarding the 3.3v mod, what would be the damage to a drive if the mod is done and it did not need it?  If no harm, then give it a shot.  I dont know how that would work either.

Yeah, no harm afaik. I'm going to give that a shot as clearly the drives aren't spinning up and this 3.3v mod seems to be a common solution for WD white labels.

UPDATE:
 

So the issue ended up being the 3.3v mod most people have issues with on these WD white label drives. I initially tried covering up the 3.3v pins with electrical tape, but had issues as the tape was too thick and the power cable wouldn't seat or would slide the electrical tape off when seated.
 

What I ended up doing was using some SATA power extension cables to connect between the PSU SATA power cables and the HDDs. I noticed that people were snipping the orange wire (3.3v) of their extension cables, but these don't have any so I assume that signal is never being passed through. You can see images of that here.


Doing this resulted in the drives spinning up and being recognized both in BIOS and unRAID. I also solved my initial problem of slow drive speeds when running parity-sync, I'm guessing my controller was thermal throttling or otherwise causing a bottleneck.

Praise be! Back to normal parity-sync times