Hi:
I did some searching first but think I really need to preview what I've done before powering up a SATA drive after pre-prep and board/NV memory swap.
Apologies if I am the 100th person to ask the same thing & for the lengthy details. I personally think details are better but most people do not agree with me. I hope it helps someone else (even if to discourage following a bad path). The more I read about this the more I see what can go wrong if I don't prepare properly.
I offered to help a friend with a non-booting Dell Inspiron (560?) with a Barracuda 7200.12 320 GB drive. I said yes, being a tech (everything but PC's) and having swapped an IDE controller on a puny drive in the mid 90's successfully. As I began reading online, I realize how much things have changed! He wants to recover data to the extent possible. We can reinstall his OS from his media disks when we get to that point, on another drive.
Step 0: Observed drive won't spin up. I looked for all the dumb things like inoperative PC cooling fans etc. Pulled the drive and attempted powering it up with a USB cable drive multi-connector tool but remembered person who gave it to me wasn't sure it worked. I don't know what click-of-death/11-clicks urban legend symptom sounds like. This one makes a periodic attempt to run then stops, waits a few seconds then repeats. I only let it do so 2 or 3 times on each attempt, then tried another power source. The startup sound is to my ears more like an acoustic burst...a more complex sound than a click. I'm sure that helps a lot
. Most confident power source attempt was a previously opened portable HDD so I took that apart & found I had a 7200.11 320 GB drive in there. No change in drive behavior. This gave me the idea to buy more of the same drive for Frankendrive efforts. I don't remember what I could see in the PC BIOS at boot attempt as it was mid December, like whether the drive was recognized.
Step 1: I bought two more of the same drive, erased & tested from an eBay HDD supplier. The only thing I had a clue initially to try to match was the firmware. I said one match would be great; didn't care about 2nd, for Plan B (which I think is a bad idea now, so I won't mention it) They were able to provide one with same firmware and I later found the controller PC boards match both p/n and Rev. Slight variations in semiconductor markings. Inoperative drive is Barracuda ST3320418AS, 7200.12, board is 100535704 Rev. B. Before delivery, seller advised I change the serial memory chip if there was one, or send back for free assessment & possible $60 repair. Told me several 25xxxxxx type p/n's to look for on IC and to look for burn marks.
Step 2: Drives arrived. One of them had identical drive p/n, firmware, board p/n and Rev. Datecode was a few digits off. Original failed drive PCB has absolutely no signs of burning or other distress, just cigarette smoke residue and tarnished silver-plating on board traces. I cleaned them as I considered attempting powering it back up with the same board. Inspected under microscope. Board looks great. I hope that doesn't mean spindle motor is seized. I hope that a tarnished connection only allowing two of three motor phases connected could prevent startup (wishful thinking).
Step 3: I decided to not try reassembly with cleaned board...seemed too simple a solution and probably based on laziness. Yesterday I attempted finding datasheets for the three IC's that differed on the two otherwise same-marked boards. DRAM were different but I didn't care. Slight marking variation on 3rd or 4th line on main controller QFP IC...I don't care. Some slight difference on SMOOTH motion profile small QFP IC...don't care; I can't change it myself anyway. Original board had 4184P SOIC-8 MOSFET+Schottky diode combination. New board had SP8U36 (I think). I couldn't find that one so I stopped caring. Original board had Winbond 25X40ALS33 and new board had 25FW406A. I spent WAY too much time trying to find datasheets in English to compare them. FYI: You need to prefix Winbond p/n with W which is not on device top marking (per datasheets). Closest I could find was 25X40ALSN. There are excess inventory chip vendors sellers out there with 'ALS32 and 'ALS33 but no datasheet found. New board had 25FW406A 29 9YG03 marked part. I eventually figured out a couple prefix letters but only found 403 and 806 family data...maybe too old or customer-specific build...don't care now. I figured out both parts are in 2.3-3.6 VDC range and seem to be 4 MB so I am less worried about compatibility.
Step 4 (today 1/4/18): Swapped both memory chips on ESD bench under microscope & cleaned flux off to inspect my work. I put the 25FW406A back on the failed board and the 25X40ALS33 original memory device on the replacement swap board because...why not? I didn't even think to look up SATA connector pinout but I checked the MOSFET/Diode chips with DMM in diode mode, and 2 or 3 SMT diodes. All measurements extremely close between both boards. Checked both leads polarities to be sure I saw everything I could. I also measured the three motor contacts in resistance mode for some reason instead of diode (remember, motor is not present with board removed from drive frame). Oddly, failed drive started giving me different results on each measurement (DMM charging up MOSFET gate capacitance?). Initial measurements were something like 20, 15, 15 ohms. I checked the replacement board and consistently got 11 ohms among all combinations of pin pairs. The failed board eventually stopped changing and all measured about 50 ohms. Odd but I never place bets on in-circuit measurements. I saw nothing alarming and the discrete diodes and measurements of MOSFET/Schottky chips all matched between both boards within a few mV in diode mode.
Step 5: (Today) Check-in and sanity check questions for anyone still awake after reading this.
Am I ready to reinstall the failed drive and power it up?
Is it OK to do this without expert software or Acelab hardware?
Any risks?
Does this sound more like a non-rotatable motor or some other scenario given board has no sign of distress?
What next?
Thank you
Murray