"Disk Failed", then "Disk Unplugged" errors

[BlueSkye]

I think at this point you have nothing to loose, at least if you replce the MOSFET`s you can rule them out of the equation and know they are now good. I cant be 100% sure if this will be a cure, but best place to start i think as they seem to be the most likley fault at this point.

If you have access to a hot air rework station its fairly easy, use a small tip set fairly hot but low air flow so as not to blow the surrounding componens off, pleanty of flux on the old surface mount IC, key is not to get the board to hot, keep moveing the tip around the IC so you warm the board faily evenly and you shouldn`t have a problem and fine pair of tweezers is very handy to very gently lift the MOSFET whilst you heat it. You can do it with a standard soldeing iron, but you can risk over heating the PCB and the tracks lifting + knocking other componens off, but is possible if very carfull.

Well I am half way there. I replaced the MOSFET devices for slots 3 & 4. Slot #4 now works like a champ. Slot 3 is still dead even though I replaced the MOSFET twice thinking I may have had a bad one. No joy. In the process I think I buggered the board at pin 3 of the MOSFET by overheating and "removing much of the pad and there's probably no way to recover that without wiring that pin with a jumper to wherever it is supposed to go. Pin 3 is the S2 source so it is either +5 or +12 and I'll try to figure out which one that is and jumper directly to the power supply.

I am still not sure why I was unable to get either of these slots working by plugging in the disks externally via a M to F SATA extension cable and an external power supply yet the external disk would work fine in any of the other six slots.

Thanks to all of you who have brought me along this far! This is a glass half full sort of day!

[maeffjus]

Hey Guys, I have to write here too:

I do have a TS-473e and disks 3&4 (one HGST and one Seagate) have been disconnected.
Then I did the work to bridge the MOSFET as it is supposed here.
Now the weird part:

Disk 3 does spin up like a charm rightaway now (initially only disk 1&2's led was on during the initial moments of booting), so basically #3 powers on with the others.
But disk 4 does not! It will power on when the NAS says "spinning-up drives" and then works for a while only to go off again.
Do you guys have any idea what to do?
I have also checked that the MOSFET (and its bridging) is connected to each disks connector.
So how can this happen and what could I possibly do about it?

I can even hear and see when disk 4 quits its job: it clicks like a 2.5" USB disk that can not draw enough power from the port.
It does that a few times and then is gone until another reboot.

Any help is highly appreciated!
Regards,
Maeffjus

Something in addition:
Just for fun I have turned-up the disk-fan to 100%, the drive comes back online.
Also when I previously plugged a brand-new WD-HDD, it did not help and it was not recognized.
Any other help / information on this new point?

[Beddhist]

It sounds like disk 4 is still behaving like it was originally, since the fan makes it work. I would check the bridging again. Perhaps there is another component involved?

[BlueSkye]

It sounds like disk 4 is still behaving like it was originally, since the fan makes it work. I would check the bridging again. Perhaps there is another component involved?

I have to agree. I believe the SATA back planes in these units have problems the go beyond just getting DC power to the drives. Interestingly, before purchasing another currently technology NAS from QNAP I asked their support staff if their latest models also suffer from back plane issues. His honesty was surprising when he did say that he's seen back plane issues with current models as well. In spite of the warning bells going off in my head I went ahead and purchased another unit.

Knock on wood, but I have an extremely old TS-459 PRO II that has never hiccuped nor has ever lost a single disk purchased all the way back in 2011.

[jackams9@aol.com]

I think at this point you have nothing to loose, at least if you replce the MOSFET`s you can rule them out of the equation and know they are now good. I cant be 100% sure if this will be a cure, but best place to start i think as they seem to be the most likley fault at this point.

If you have access to a hot air rework station its fairly easy, use a small tip set fairly hot but low air flow so as not to blow the surrounding componens off, pleanty of flux on the old surface mount IC, key is not to get the board to hot, keep moveing the tip around the IC so you warm the board faily evenly and you shouldn`t have a problem and fine pair of tweezers is very handy to very gently lift the MOSFET whilst you heat it. You can do it with a standard soldeing iron, but you can risk over heating the PCB and the tracks lifting + knocking other componens off, but is possible if very carfull.

Well I am half way there. I replaced the MOSFET devices for slots 3 & 4. Slot #4 now works like a champ. Slot 3 is still dead even though I replaced the MOSFET twice thinking I may have had a bad one. No joy. In the process I think I buggered the board at pin 3 of the MOSFET by overheating and "removing much of the pad and there's probably no way to recover that without wiring that pin with a jumper to wherever it is supposed to go. Pin 3 is the S2 source so it is either +5 or +12 and I'll try to figure out which one that is and jumper directly to the power supply.

I am still not sure why I was unable to get either of these slots working by plugging in the disks externally via a M to F SATA extension cable and an external power supply yet the external disk would work fine in any of the other six slots.

Thanks to all of you who have brought me along this far! This is a glass half full sort of day!

Some good new then, keep going, i would trace the broken solder pad and put a jumper wire in from the leg of the MOSFET, you can do this by using a multimeter, hold one end on the end of the broken track and then probe around until you get a dead short, then try and trace the track and make sure it run back to the MOSEFET to be sure your in the right area, that should hopefully sort slot 3 as well. As for the extension cable not working, there will be a controller on each bay which looks for a supply voltage from the MOSFET amongst other things, the controller needs to see this supply voltage be for it will allow any for of access to that bay, no voltage, the unit assumes their is no drive in the slot, in tern means no access to that socket (even from a data point of view)

[Elevel]

Reading through the threat and considering the errors I get I am quite certain my TS451 has a broken back plane. How can I organize repair or order get a spare part? I can't find any replacement in the spare parts section on shop.qnap.com
Of course plan B is to buy a new NAS.

[BlueSkye]

I came up blank finding a repair facility or spare back planes for my 879. I suggest, if you really want to keep the 451, look for a well-priced used one and cannibalize from the two to make one good one.

[Al3x]

So really struggled with QNAP support over this and the in warranty/ out of warranty service is by far one of the best i have experienced.

Those units that are in warranty have a long wait time without a unit before you get one sent back. Those that are out of warranty have postage to NL and the 110USD fee on top to cover labour, and other fees.

The way i see it, you have the option of either solder the circuit and forget about the power down feature, use an external psu and SATA leads or just as others have said buy a backplane from ebay or other place and cannibalise it.

If only there was more information on what exactly has broke so it can be removed/ replaced on the board at a circuit level.

[BlueSkye]

I replaced the offending parts, the surface mounted 8-legged MOSFETs and that really didn't help. I also got the SATA data extension cable and the external power supply and also had no luck. In my case, I suspect the problem was even more than just the power on/off circuitry that fed the drives.

[Al3x]

I replaced the offending parts, the surface mounted 8-legged MOSFETs and that really didn't help. I also got the SATA data extension cable and the external power supply and also had no luck. In my case, I suspect the problem was even more than just the power on/off circuitry that fed the drives.

Thanks, based on this i think it might not be worth undertaking the gamble and look for another alternative.

[BlueSkye]

Good luck! I was able to find another one on an auction site and also purchased a newer model.

[Danixu86]

Hello again!,

I'm here just to report my status. After bypasing the mosfet all started to work without problem and was working for a month without problem.
I bougth the mosfet on LCSC for about 1€ and I've received the mosfet a few days ago (mouser also have it). Today I've proceed to change the bypass (it still working, but I preffer to keep the original operation mode), and all seems to be working as always without a problem, so I can't believe that for just a component of about 1€, you have to discard a 400€ NAS because there's no replacement parts.

Again, thanks to all who has debugged this problem and has realized that the problem was the mosfet, because for sure I would still changing drives thinking that maybe they are not compatible with my NAS or maybe I would have abandoned the usage of that slot.

Best regards.

[Beddhist]

Consider fitting heat sinks or running the fan at higher speed, to avoid a repeat failure. It is a design flaw, after all - or planned obsolescence.

The parts (back planes) are available for those units not EOL, but QNAP refuses to sell them to you. Instead, you have to send your unit in and pay $$$. Commercially understandable, legally questionable (depending on location), morally and environmentally reprehensible.

Well done and thanks for reporting back.

[Serruxo]

Hello

I am trying to repair my TS-453A as I have published in other thread called "interface fatal error" viewtopic.php?f=50&t=156999, and I have some doubts about the procedure to check the components or discover the problem with a multimeter. I have replaced the mosfets and a pair of transistors in the backplane, but SATA errors are still present in the logs (tested with different disks).

Could you please give some advice or help?:

- Can the components be checked with the NAS powered off?
- Do I need to switch on the NAS?
- Do I need to insert the HDD into the sata2 port (the one is working bad) and check with a multimeter?
- Is there any "basic" tutorial you might know to discover problems on a board?

It is said to check the continuity between the pins of the MOFSFET and the pins of the SATA connector, but I do not know if this must be done having the HDD inserted and the NAS powered on.

Thanks in advance.
Regards

@microsolder..could you mark continuity testpoints in the pic ? (at least the ones visible)
so the person asking could make sure what components are what ?

TS-253-pro-bp.png

Continuity testing an unpowered board will not detect the problem. What one needs to do is to
monitor the +5 V and +12 V lines at the SATA power connector. The related pins are marked in the
picture. The critical line is the +5 V. While +12 V tolerates 10% (1.2V) drop and 10.8 V is OK, the
+5 V has only 5% tolerance. The disk begins to reset when the +5 V drops below 4.75 V. So the
right method for the analysis is to monitor the +5 V during normal operation. A DMM with min/max
sampling is helpful to catch the voltage drops.

One can also measure the gate pins of the switch MOSFET, the ones on the input side next to the
input voltage lines. They are connected together and driven by the same voltage. When on the
gate voltage should be below +1 V, which I suspect it isn't before modifying the driver circuit.
In TS-431 it was +3.2 V which results in only -1.8V gate drive -- not enough.

The method to locate the resistor to replace is to check which resistor is the one that is connected
directly to the gate pins. In the picture it may be R70. Then measure its resistance. If it is 100 kΩ
replace it with a 10 kΩ one and recheck the gate voltage. It should then be much lower.
This replacement without damaging anything around it requires some skill, though,
due to the size of the component and little space around it.

bridging was already shown earlier (either accorss the component or be removing the mosfet alltogether and bridging the traces)
even thought replacing it would be safest..right ?

Replacing the switch MOSFET is useful only if the driver is also modified so that the gate voltage is low enough.
At least that is the case with TS-431. Jumpering the MOSFET always works even though the power switching
functionality is lost.

[Serruxo]

The TS-235 pro back plane you show does not have a second MOSFET switch. Like other QNAPs the disk 1 is fed directly from the connector. Therefore either there is a failing switch on the main board (can't tell if there is any without seeing the main board) or the 5 V buck regulator, also on the main board, is failing. Next, photo of the main board, please. (This gets interesting.)

Do you mean that probably there is another identical MOSFET+DRIVER on the mainboard that fed the SATA1 port?
As I said in a previous post, my TS-453A (4x sata ports) has the SATA2 failing. I tested the 12V pins of the SATA port and they come from the PCI connector. Perhaps it could be the reason of my problem because problem persists in spite having changed the two MOSFET+DRIVER that I think are only related to the SATA3 and SATA4...

Thanks in advance