A stupid question: did you check the PS voltages under load? I once had a PS like that, it showed the correct voltage values and started fine without a load, but when connected to the main board it doesn't boot. The clue came in my case from the 5VSB that dropped to 1V when connected. It was a faulty electrolytic capacitor in that rail.
Short somewhere on the backplane? you checked everything i would have. Pull out a drive and see if you can power and access it on a good connection. External USB to SATA/NVME are easily had.
The drives should be in some kind of pool, anything other than RAID1 will require all the drives be connected at the same time. The backplane has power regulators for 8 drives, it would be unusual to do RAID0 or 1 with that many, since that's enough to do RAID5 or 10. That said, older and low end SAS HBAs are pretty cheap, and you can get miniSAS to 4x SATA cables for a similar amount. Then you just need 8 SATA power plugs and an understanding of how to import a foreign pool you don't know details about, probably in linux, which I've never done. As far as I understand this is totally doable, just a bit more involved.
Those slot connectors are PCI/PCIe I just woke up and won't be figuring out which right now, which doesn't necessarily mean anything in cases like this, but it does mean you might be able to get extension cables for it to do diagnostics with the board out. If they are electrically PCIe? then that means there's a SAS or SATA controller on the backplane and actually you could replace the motherboard with anything and use the extension cables to bridge the cap. This will be a horrible mess but it will let you bring up the drives to aid in offloading data to a new device.
@@Lantertronics Since it's an OEM unit, it's probably self-contained design with integrated temperature controlled fan, decoupled from the rest of the device. I had an old Intel server that refused to POST without case fans attached and spinning.
@@sabamacx Hmmmmm... the power supply fan is its own thing. The main case has two fans with standard 3 pin headers. Just for grins I could hook the fans back up and see if it magically started working (occasionally I've had that actually work -- I think just taking things apart and putting the back together worked some dodgy connections).
@@Lantertronics I agree checking out the case fans is definitely worth checking. Also, of that mother board has been around a while, it could be one that suffered from the bad/fake bypass caps that China pit out that shorted many a mother board back in the day. If you have an ESR meter, it might be worth checking the bypass caps on the mother board. Unless the mobo suffered a major power spike, the semiconductors likely are not the cause. I have had bad memory modules cause the power supply not to start, so pulling the memory modules might also be useful. That's about all I can think of I have experienced before. Good luck!
One other thought before you dismiss the power supply---if you have a decent load, like even an appropriate voltage incandescent lamp,(s) you can check the power supply under load. Just because it works open circuit... :P
A stupid question: did you check the PS voltages under load? I once had a PS like that, it showed the correct voltage values and started fine without a load, but when connected to the main board it doesn't boot.
The clue came in my case from the 5VSB that dropped to 1V when connected. It was a faulty electrolytic capacitor in that rail.
Short somewhere on the backplane? you checked everything i would have. Pull out a drive and see if you can power and access it on a good connection. External USB to SATA/NVME are easily had.
The drives should be in some kind of pool, anything other than RAID1 will require all the drives be connected at the same time. The backplane has power regulators for 8 drives, it would be unusual to do RAID0 or 1 with that many, since that's enough to do RAID5 or 10. That said, older and low end SAS HBAs are pretty cheap, and you can get miniSAS to 4x SATA cables for a similar amount.
Then you just need 8 SATA power plugs and an understanding of how to import a foreign pool you don't know details about, probably in linux, which I've never done. As far as I understand this is totally doable, just a bit more involved.
Those slot connectors are PCI/PCIe I just woke up and won't be figuring out which right now, which doesn't necessarily mean anything in cases like this, but it does mean you might be able to get extension cables for it to do diagnostics with the board out. If they are electrically PCIe? then that means there's a SAS or SATA controller on the backplane and actually you could replace the motherboard with anything and use the extension cables to bridge the cap. This will be a horrible mess but it will let you bring up the drives to aid in offloading data to a new device.
I've had power-on head-aches with failed/dead fans in the past which had fan speed sensing: no fan spin registering = no boot.
Would the power supply fan itself not come at all? That's what we're seeing -- it's totally dead.
@@Lantertronics Since it's an OEM unit, it's probably self-contained design with integrated temperature controlled fan, decoupled from the rest of the device.
I had an old Intel server that refused to POST without case fans attached and spinning.
@@sabamacx Hmmmmm... the power supply fan is its own thing. The main case has two fans with standard 3 pin headers. Just for grins I could hook the fans back up and see if it magically started working (occasionally I've had that actually work -- I think just taking things apart and putting the back together worked some dodgy connections).
@@Lantertronics I agree checking out the case fans is definitely worth checking. Also, of that mother board has been around a while, it could be one that suffered from the bad/fake bypass caps that China pit out that shorted many a mother board back in the day. If you have an ESR meter, it might be worth checking the bypass caps on the mother board. Unless the mobo suffered a major power spike, the semiconductors likely are not the cause. I have had bad memory modules cause the power supply not to start, so pulling the memory modules might also be useful. That's about all I can think of I have experienced before. Good luck!
One other thought before you dismiss the power supply---if you have a decent load, like even an appropriate voltage incandescent lamp,(s) you can check the power supply under load. Just because it works open circuit... :P
The only other quick thing I can think to check is the CMOS battery.
Good idea -- it's easy enough to check, so I just went and checked it, and it's reading 3 volts, so it's OK.