Hi folks,

A client brought in 2 HDDs that failed at the same time during an electrical storm; a Seagate 2TB & a Seagate 3TB. Neither drive will spin up and one had a protection diode that appears to have melted off of the board (connectors at both ends were still intact so I'm guessing that it must have got hot enough to melt the solder; it wasn't "blown apart"). I believe it's a "crowbar" protection diode across the 12v rail (which now measures 2v) (and the component is now a dead short both directions) ... but I do get 12v across the same component of the other board (but still no platter rotation). So I've reached the conclusion that (a) both boards have been wiped out by some kind of power spike - probably lightning-in-the-area related and (b) there's a good change that the platters, heads etc are still in good nick.

So "plan A" is to get a couple of donor boards from Aliexpress/eBay (both are readily available) - get myself a cheap flash BIOS programmer - copy the BIOS to the donor boards - install them and (hopefully) restore both drives to get the data off them (client can't afford 2x professional recovery). That's the plan anyway.

I've been doing a bit of "Googling" and have come up with https://www.amazon.com.au/EZP2019-Progr ... B0897W5J1Y for a BIOS reader/programmer. I don't have the drives handy, but one of them had a board with the id "100724095 REV A" which I found for a good price here: https://www.aliexpress.com/i/4000314528723.html The other one was readily available too.

So at this point I'm just hoping for a bit of a "sanity check" from those who know more about this than I do (ie pretty much everyone). Does my plan sound workable? Reasonable chance of success? Is that programmer likely to work? Any pitfalls? Anything I've not thought of? Any advice?

Many thanks.

Upload photos of your PCBs.

Check the zero-ohm resistors near the SATA power connector.

https://www.hddoracle.com/download/file.php?id=3120&t=1

TVS Diode FAQ:
http://www.hddoracle.com/viewtopic.php?f=100&t=86

_________________
A backup a day keeps DR away.

Thanks for that - it was very helpful. I think I'll tackle these one at a time - starting with the one you posted the link to the photo identifying the components. Thanks to that, I've now found that the 5V TVS diode appears to be shorted.

I don't have my fine-tip soldering iron with me so I'll lift one end of the diode tomorrow and see how the voltages look from there.

Thanks again for your help.

I've removed the 5v TVS diode and confirmed that it was shorted. When I supply 12v to the board from my bench supply via a SATA power connector I get 12v across the 12v TVS diode - so I'm assuming that my problem is completely on the 5v side.

When I feed 5v onto the 5v rail on the SATA power connector I'm measuring 5v on one side of the zero ohm resistors, but only 0.17v on the "downstream" side - so looks like the zero ohm resistors are close to being open circuit.

When I bypass them by shorting from pins 7/8/9 to what's supposed to be the +5v pad of the (now removed) TVS diode the board still doesn't draw any current.

I've also tried injecting 5v onto what I believe are "downstream" parts of the 5v rail but still no current draw (and at this point I'm struggling to understand where tracks lead to).

Photo of the board can be found at: https://pbase.com/cjsouthern/image/1751 ... iginal.jpg (this is the full resolution version -- it's too big for the 1024px limit of image tags)

Any further thoughts?

PS: One of the Mods has kindly bumped me out of the new-users-requiring-moderation group so I should be able to get replies posted far quicker now.

Measure the resistance between ground (any screw hole) and V1, V2, V3, Vneg. If no test point is shorted, then measure the voltages at these points.

_________________
A backup a day keeps DR away.

No shorts.

I've applied 5v across where the previously installed 5v TVS diode was because the zero ohm resistors are open circuit. I'm not supplying 12v anywhere.

Voltages are as follows:

V1: 1mV

V2: 3mV

V3: 0mV

Vneg: 0mV

Thanks for your continued help. I've ordered a hot air soldering station, but this sized componentry is pretty much on the limit for my 64 year old eyes, hands, and equipment. Since I'm in NZ and you're in Australia, just wondering if you'd be open to having both boards shipped to you for evaluation? And if so, how much you'd charge for looking at them?

I'm older than you and my eyesight and coordination are failing. Sorry.

_________________
A backup a day keeps DR away.

That's OK; I'll just keep trying to work the problem until something works. Is there anything else you can think of for me to try? Any idea where I could get hold of a circuit diagram? I'm guessing at this point that there's some other series component that's gone open circuit. Customer said that both drives went down at the same time when the power went off, but my guess is that that the power went off because of a lightning strike nearby (there was an electrical storm at the time). Given that one TVS diode physically detached itself from the board (customer says he didn't remove it), I'm guessing that the drives have had one heck of a surge ... so I wouldn't be surprised if a lot of components are fried (hence the initial enquiry about just swapping the entire PCB).

Did you test the zero-ohm resistors near the 5V TVS diode? I expect you'll find that they are open.

_________________
A backup a day keeps DR away.

Yes - and they are open - but unless I'm mistaken, they sit between the SATA power connector 5v input and a common rail that the TVS diode is normally connected to; hence the reason I injected 5v directly to that common rail to bypass them since I have no way of replacing them until my hot air soldering iron arrives.

But even after doing that (and verifying that the voltage was now indeed present at the 5v rail), I still essentially had 0v at those 4 points you asked me to measure - so in the absence of a circuit diagram, I'm assuming that there's some other series component that feeds the downstream sub-circuits on the 5v rail that's open circuit ... but I can't figure out what.

I'm guessing that the SMOOTH IC must be damaged. You could try transplanting the Winbond serial flash memory to a donor PCB. If you're not comfortable with that, then a company like hdd-parts.com charges US$50 for a PCB plus firmware transfer.

_________________
A backup a day keeps DR away.

Thanks again for your insight & help (I've learned quite a lot about some of the board electronics thanks to you). The customer doesn't have a lot of money to spend on the problem so - even though he'd like his data back - it's still a case of "if it can't be done cheaply then it can't be done at all" (for me it's mostly just a learning exercise). With that "frame of reference" I'll probably just buy a doner board from Aliexpress and try to move the memory chip myself; on one hand, applying heat from a hot air soldering iron to it makes me nervous, but on the other hand, it's my understanding that that's how anyone else would do it anyway ... so I'll probably just give it a go after practicing on an old throw-away drive.

Before you waste money and time on a replacement PCB, you need to confirm that the preamp IC on the headstack (inside the drive) isn't shorted.

The 3 pins in the attachment should be connected to Vneg, +5Vin, Ground. Use your DMM to identify which is which, and then test for shorts to ground at the corresponding pins on the HDA connector.

_________________
A backup a day keeps DR away.

I just measured the resistance of all 3 with my DMM using one of the metal screw holes on the drive. The first 2 appeared to have some kind of capacitance; the meter reading was in the megaohms and climbed higher and higher. the 3rd pin showed about 0.3 ohms. So from what you said, that sounds about right?

Just be absolutely certain, measure the resistance between the 3rd pin and each of the others.

This is what I'm concerned about:

https://www.hddoracle.com/viewtopic.php?p=22173#p22173

Here are some very old preamps:

https://www.hddoracle.com/viewtopic.php?p=582#p582

_________________
A backup a day keeps DR away.

No shorts found; the readings were all quite high. Fingers crossed I may have dodged a bullet with the pre-amp.

Update: New PCB arrived for the 2TB drive. I was able to swap-over the BIOS chip using the hot air soldering station that I purchased for the exercise, but, alas, still no joy getting the drive to even spin up when installed on a known-working USB -> SATA cradle. I had to set the hot air temp control to 450 deg C to be able to remove parts from a practice board in a timely manner; I appreciate that that's the temperature measured at the sensor in the iron and that what hits the components is lower, but it still makes me nervous. Lower temps just didn't move the part - even at 450 deg C it still took around 10 to 15 seconds to be able to remove the chip - and that's with the outlet for the iron about 2cm away and on the highest fan setting. FYI the same settings took about 5 seconds to melt through a length of lead/tin solder. Makes me uncomfortable, but I don't know what else I could have done.

I measued a few voltages; +5 and +12 were present across the respective TVS diodes now, but still 0v at each of the 4 points that I was asked to check earlier. I did this check with +5v fed into the sata connector from a bench power supply. The supply can only provide 1 voltage at a time so I had to check 5 & 12v seperately. No current draw on either rail from the bench PSU. I guess it's entirely possible that the replacement PCB is faulty (it was the cheapest one I could find on AliExpress -- looked new when it arrived though).

I also measured the resistance between each of the 3 wires heading to the spindle motor - each appeared to be around 0.6 ohm; I have no idea if that's normal or not.

So I'm going to have to call it: Time of Death 11:23 NZST. It's been an interesting exercise though and I've learned a lot, but this is now just taking up too much unproductive time. Thank you again for your help - much appreciated.

Maybe you have corrupted Rom and you are transferring it from one PCB to PCB.
I have had the same problem. The original ROM was corrupted and drive didn't spin. I've transferred this corrupted ROM to a good working PCB. And guess what, drive didn't spin with the new board again.
I have uploaded Rom here and Pepe and fzabkar noticed right away that ROM is corrupt. They told me how to make it work and drive spins normal again.
Get yourself a ch341a v1.7 programmator. You can download ROM from PCB with it, and you can use it also as TTY terminal connection to your drive.
Just be careful how you set voltages, you don't want to burn anything. Think twice, do once.

Thanks for that. In retrospect, I probably should have installed the new board in the drive before swapping the BIOS and see if it started to spin up.

FYI, I've also tried to transplant BIOS with soldering and it didn't work. Don't know what went wrong. Chip for motor (SMOOTH) was very very hot when powered up...
But with the programmer everything went easypeasy. You just have to be careful with voltage, it depends on the type of ROM chip.