SAMSUNG: SMART is 'GOOD' in IDE mode, yet 'BAD' in SATA mode

[NRA]

Hi guys)

I'm a bit confused for another (!) SAMSUNG SATA-2 is forgetting its manners:
- tried different working hardware (mobo's/psu's/cables) at no avail;
- the HDD SMART is OK (not cleared; ext-scanned; just 5 reallocates, no pendings; no changes);
- in IDE mode the SMART is GOOD, whereas in SATA (no RAID) mode it's BAD.

Surface scan (including ping-pong/butterfly) is ok, after useless attempts did security erase and... still the same.
In general, the drive is working no prob, BUT it always triggers F1 error message and annoying system warnings, let alone it's rather difficult/impossible to try some OS'es due to SMART reported as BAD; not to mention it's but strange.

WHY?! Where to?

This is my second SAMSUNG weirdo with such symptomatic and I feel somewhat dumb.
Maybe someone more experienced could give me a useful hint? (except captain-obvious using in IDE or throwing it away)

TY

[labtech]

I would confirm with 2-3 more SMART tools. Always.

[ccc]

Please show as a complete SMART attributes report.

[NRA]

If I remember ok, it's very the HDD micro which decides and reports SMART, not some software, right?

This is not my drive, so I took a while. Anyway, different SMART-reporting software (FreeDOS boot and WinXP/W7/W8 boot PE) shows the same SMART values:

SAMSUNG HD321KJ S0MQJ1KQ302374
-------------------------------------------------------------------------
ID Name Value Worst Tresh Raw Health
-------------------------------------------------------------------------
1 Raw read error rate 100 1 51 9 •••••
3 Spin-up time 100 100 15 5568 •••••
4 Number of spin-up times 92 92 0 9064 ••••
5 Reallocated sector count 97 97 10 5 ••••
7 Seek error rate 253 253 51 0 •••••
8 Seek time perfomance 253 253 15 11389 •••••
9 Power-on time 100 100 0 17384 •••••
10 Spin-up retries 253 253 51 0 •••••
11 Recalibration retries 253 100 0 0 •••••
12 Start/stop count 96 96 0 4814 ••••
13 Soft read error rate 100 100 0 755640955 •••••
187 Reported UNC error 1 1 0 8323547 •
188 Command timeout 98 98 0 239 ••••
190 Airflow temperature 76 61 0 24°C/75°F ••••
194 HDA Temperature 163 115 0 25°C/77°F ••••
195 Hardware ECC recovered 100 100 0 755640955 •••••
196 Reallocated event count 97 97 0 5 ••••
197 Current pending sectors 100 96 0 2 •••••
198 Offline scan UNC sectors 253 253 0 0 •••••
199 Ultra DMA CRC errors 200 200 0 0 •••••
200 Write error rate 100 100 0 0 •••••
201 Off-track errors count 100 99 0 0 •••••
202 DAM errors count 100 100 0 1 •••••

Sure, it's rather aged and worn (+2 pendings now), yet I still don't follow why in IDE mode it works as 'GOOD' whereas in SATA mode it either reports 'BAD' or had troubles reporting SMART.

TY

[fzabkar]

All SMART tools access the same data. The difference is in how the authors of these tools choose to assess the data.

I prefer CrystalDiskInfo:

http://crystalmark.info/software/Crysta ... dex-e.html

Your SMART report looks bad. I suspect that the reason that one tool is telling you that everything is OK is that it is simply checking whether the "Value" of any attribute has fallen below its threshold ("Tresh"). At the moment all attributes are OK in this respect.

However, other SMART tools might look a little deeper.

1 Raw read error rate 100 1 51 9
5 Reallocated sector count 97 97 10 5
13 Soft read error rate 100 100 0 755640955
187 Reported UNC error 1 1 0 8323547
188 Command timeout 98 98 0 239
195 Hardware ECC recovered 100 100 0 755640955
196 Reallocated event count 97 97 0 5
197 Current pending sectors 100 96 0 2

The Worst value of the Raw read error rate attribute has fallen below its threshold at some time in the past. Attribute 187 is at rock bottom, and the raw values of attributes 13 and 195 look very bad.

[NRA]

Thank you, fzabkar.

I know the drive is a crap now, yet as I wrote before, it triggers the system SMART err message at mobo POST stage--before any software could run.
You're right the drive is degraded, however why the drive reports 'GOOD' in IDE while having 'BAD' status or troubles with SMART in SATA mode?
Shortly, what makes so big difference?

Don't know if it's a common case, the point is I encountered a ~750GB SAMSUNG with a similar issue, but when I found out another one, I would like to learn about possible reasons. (Just curiosity, for I really doubt it would work for long even in IDE mode.)

TY

[fzabkar]

Is SMART enabled in both modes?

According to the ATA standard ...

7.48.8.2 Description

The SMART RETURN STATUS command causes the device to communicate the reliability status of the device to the host.

The Abort bit shall be set to one if SMART is not enabled.

The drive returns a value of 2CF4h if "the device has detected a threshold exceeded condition".

[NRA]

Some mobos had SMART option as AUTO, forcing it ENABLE made no diff; memclocks/timings (DDR2-800/DDR3-1333) are default, so it must be something about the HDD. The tested mobos (C2D/AMD) are at default settings too with possible overclocking features DISABLEd.
Anyway, the result is still the same: the HDD reports SMART as 'GOOD' in IDE and occasional 'BAD' or unable to read SMART in SATA mode.

Attribute 187 is at rock bottom, and the raw values of attributes 13 and 195 look very bad.

Although according to wiki some values are vendor-specific (rather SAMSUNG-specific?), yet it doesn't explain the difference of interpreting the same SMART report.

"the device has detected a threshold exceeded condition"

So, if the HDD is in SATA mode (using extra protocols), it may find out some issues or consider the same values as 'BAD' for performance mode? Perhaps, it does make sense, because I suspect it has something to do with a firmware or some related.

As far as the HDD is a reject, does it seem ok to flash a new firmware?

TY

[fzabkar]

I don't understand what is happening in SATA mode. Sorry.

As for a firmware update, it won't change anything as far as the bad sectors.

BTW, IDE and SATA modes affect the SATA controller rather than the drive itself. Therefore it must be a BIOS thing, AFAICT.