IC35LO40AVER fine in MHDD hangs system in windows

[Shortscurcuits]

Hi guys quick question, deathstar was used to keep a backup of about 20GB of work files, started getting a few hangups on my main PC and isolated it to this drive.

The full 37.4 GB shows now as unallocated space, scanning with MHDD shows a perfect surface from start to finish, however when trying to scan the drive with Active file recovery,or R-studio reports multiple bads and hangs at 22%. Spinrite shows cabling errors & ECC errors.

I have a donor Drive and NVRAM shows as OK on the original board. Is this a case of the bad write chip?

I don;t have a scope to check data .

Any ideas?

Thx Andy.

[Shortscurcuits]

Just want to add a lot or ECC errors after 3.0 GB. Do I have a faulty read write chip?

[dick]

I would try to mount the drive in a different pc, also use a different ribbon cable and then see if you can browse or image the drive using r-studio. A different pc motherboard might make the difference!

[Shortscurcuits]

Testing is done in a my testbed rig cables and PC are all good, extra filtering caps on a good enermax PSU. Cabling errors only show up for this drive. Any other good drive works fine. Contacts & pins have been cleaned on PCB interface. IDE connector pins have been checked also. Behaves the same in several PCs. NVRAM OK in PC3K. Will change PCB and post results.

Was just hoping someone had run into a definite fix, didn't want to run the drive if it was writing garbage.

Both sides of platter & heads are OK have checked.

[quasimodo]

check SMART logs.

[Shortscurcuits]

Thx you are correct reallocated sector count shows as -4 !

Have changed 90G chip and is scanning with spinrite -> already has recovered about 22 bads that were not there a week ago

Cabling errors have now dissappeared with new controller chip so I guess that was it.

I do check my smart status weekly and I know that it was good about a week and a half ago

There is about 55MB of data that is ECC corrupted but the surface shows good in MHDD.

This is what was freezing all the cloning software.

Weird as surface scan shows perfect media!

[rameez]

This is a problem of the read write channel also there could be losing in the connection of the preamp and the pcb . The connection is located at the back of the pcb resold it . If same problem occurs change pcb .

[Shortscurcuits]

Thanks Rameez I did re-solder contacts & cleaned the pins on HDA contact block also.First thing I check with IBM/Hitatchi & WD too!

Replaced 90G2018 chip as it was running hot and I didn't have a good donor PCB but had some with good controller chips.

Data was scrambled from 3.0 to 3.4% & 7.8-8.0% on disk and this was stalling Rstudio & Acronis trying to re-read data in this area.Actually locked up the software.

Am now running spinrite to correct as much as possible ( so far all but 2 sectors recovered) so I can recover the complete partition. Data is now being written correctly by drive so hopefully I caught it in time most files are readable at beginning 3 GB of drive now.

Strangest part is that MHDD showed disk OK in both these areas

[Shortscurcuits]

Just to say thanks to rameez & quasimodo. 3 days to scan & run Image 95% data back on a new drive.

What was missing, was not really needed.

Just a reminder folks.... just cause they can be fixed, it's not worth not backing up, not for 5 days of PC scanning and about 4 hrs diagnosis,
soldering & testing

My last backup was about 8 months ago, just got lazy

Thx again, Andy