Data recovery and disk repair questions and discussions related to old-fashioned SATA, SAS, SCSI, IDE, MFM hard drives - any type of storage device that has moving parts
January 30th, 2015, 8:53
For a PCB that "looks" in perfect condition but dead when attached to a compatible HDD...how do you diagnose what is wrong?
January 30th, 2015, 22:00
First thing I would try is probably testing TVS Diodes.
January 31st, 2015, 12:13
Thanks for your reply bcometa.
And if the TVS diodes test ok - what next?
January 31st, 2015, 14:52
After the tvs then check to see if there are any shorts across capacitors or diodes. Check coils for continuity.
You can connect the drive and pcb to a spare power supply and take voltage readings. you need to at least find +5v for logic or sometimes +3.3v and +12v for the motor. Trace the voltages and see how far you can get. Preparing a replacement pcb is often the best and only way if the goal is to recover the data.
January 31st, 2015, 15:29
>>>You can connect the drive and pcb to a spare power supply and take voltage readings
Thanks Dick, how do you take voltage readings from a live PCB that is also connected to a drive?
Most circuits of PCB are inward facing i.e. you have to remove them first from the drive before you can
get a multimeter to them?
January 31st, 2015, 17:16
January 31st, 2015, 22:26
The easiest way it to just read the ROM (or NVRAM) code and then write it to a working compatible PCB. If the drive works you know the PCB was bad.
If the drive doesn't work, I then take ithe original PCB and program it with the ROM code from the donor drive and see if it will work.
Works every time.
February 1st, 2015, 16:45
If the PCB has a dead Marvell MCU with embedded "ROM", then a ROM swap will be a waste of time. Even in thoses cases where the ROM is a discrete flash memory chip, a quick check of the resistance between its Vcc and Ground pins would be advisable, as a short circuit would make a ROM swap pointless.
As for donor PCBs, a competent tech should be able to, in most cases, diagnose a PCB in less time than it takes to order a replacement. An examination of the terminal log will often tell you where the problem lies. In fact there is a current thread at the HDD Oracle where the OP's Samsung drives (2 x HD204UI) have been rendered inoperable by what appears to be a PUIS firmware bug. The solution in that case would be to locate the PUIS flag and reset it, then recalculate the checksum. A ROM swap would be unnecessary.
In any case one should not blindly swap a ROM. At the very least one should test its integrity by verifying its checksum.
February 1st, 2015, 18:03
thanks fzabkar for those great links.
>>>The easiest way it to just read the ROM (or NVRAM) code and then write it to a working compatible PCB
how do you read the ROM code?
Writing to a compatible PCB - is that performed with PC3000?
>>>a quick check of the resistance between its Vcc and Ground pins would be advisable
how do you find where the GND and Vcc pins are and what type resistance parameters indicate a failed ROM / working ROM?
February 1st, 2015, 18:40
If the PCB is dead, then the only way to read the ROM may be by desoldering it and using a chip reader. Sometimes you can read it in-circuit with an SOIC clip and chip reader, but you would need to be mindful of the supply voltages so as not to damage the PCB. Some chip readers strap the Hold pin to Vcc, so these may damage the MCU.
The Vcc and Ground pins of 8-pin serial flash/EEPROM devices are usually 8 and 4, respectively.
As for suitable software tools, that would depend on the drive.
Powered by phpBB © phpBB Group.