Help me understand HDD failure

[netguino]

Inspired by the recent-ish video that showcased the Jonsbo N1, I decided to upgrade my home server as I was looking to do so for a while.

Case got here, and I proceeded to migrate everything from my old case. I plugged the mobo, and used a molex connector for the sata backplate. I did notice that it had two molex connectors, so I plugged another one from the PSU.

Connected hard drives, and pressed the power button. PSU spins, then stops. I think I may have seen a blue LED flash from the backplate case, but unsure.

I am a bit dumbfounded and proceed to take everything out of the case, to put it on the desk to start part by part. I eventually notice, that everytime the backplate is connected to the PSU, the systems wont boot.

Head scratches later and I decide to remove the backplate completely, and connect the drives to the mobo directly. System boots from my Unraid USB just fine. Wheh.

I connect to the web console, and notice none of the drives are recognized. I reboot to bios and notice the same. I start to panic.

I take all drives out and put them on the old case, boot from unraid and find out the same. None are recognized.

I take one drive and put it on an external caddy. No spinning. Oh no. Each drive is the same. No spinning. Did I really somehow fried 4 pcbs?

So, is this really possible? I must have messed up something really bad. I dont see any signs of fried components neither on the jonsbo backplate, or the drive pcbs, but I understand it could just be a diode, or a resistor, or something small.

Does anyone know what the correct connections to the backplate are? Did I really need two molex connectors? Is that what could’ve caused this? How rare is it that all 4 drives pcbs die?

The 4 drives were part of my unraid setup ( 2 disks and 2 parity ) so I bought 2 new pcbs, and I am going to try doing the BIOS chip hotswap. I bought two pcbs, so once they get here I will attempt to do it. Really gutted by this but I am quite unsure what to do.

[alkafrazin]

I’d check the voltages on the sata pins of the jonsbo. It might be bridging a 12v rail to a 3.3v or 5v somewhere early on, and delivering completely the wrong voltages to every drive attached simultaneously, insta-toasting the lot, and then causing a safety poweroff.

[netguino]

Thanks for the suggestion!

I have a spare PSU that I can use to test things.

The PSU doesn’t reset with the SATA backplate connected, so I was able to read the voltages on all 5 sata pinouts.

I do get clean 5 and 12 volts, but none of the 3.3 ones are live.

I’ve measured voltage across the capacitors outlined below, and that’s presumably where the 12 and 5 come from. All measure fine.

I’m guessing the board doesn’t support the old 3.3v style drives, hence why I can’t get pins 1,2,3.

I am still incredibly dumbfounded as to why I’d fry all 4 drives…

[shadragon]

The 3.3V lines are used to engage and disengage power save modes on a few models of HDD. They are not commonly used and may simply not be utilized on your system.

“If you should happen to put a DEVSLP enabled drive in a system that provides 3.3V power, the drive will go into low power mode and stay there.”

So your drives may just be in low power mode and only look dead. Impossible to tell without detailed hardware specs. The drive manufacturer tech specs would be a good place to start. (There is a decent technical article in German on epaper.heise.de named ct.14.15.152-155.pdf)

Why Do We Need SATA DEVSLP - Quarch Technology.

DevSlp