MaxTheFast wrote:
Another misunderstanding about my tests: I had tried ONLY ONCE to connect the enclosure with both cables (usb, esata-sata) and it worked in usb mode but in EVERY other tests I had used only esata-sata cable or only usb cable.
By the way today I got the CPU led error another time after my last attempt with esata-sata cable, then the bios has reset and booted with windows instead of ubuntu, then I entered the bios and set ubuntu as 1st OS, rebooted and got the CPU led error. It's amazing!

The ESATA port could be off an entirely different chip, than the six ports on the Southbridge. Make sure you're programming the correct chip.
As said I looked at that bios section but I found nothing else than setting sata ports on "hot plug" or not. The section is: settings\advanced\integrated peripherals "sata configuration". There I can set "sata mode" AHCI/RAID too but I guess it doesn't matter here, I can set "onboard lan configuration" and "audio configuration" too but I think these are not the way out. I can't set any other entry on sata ports afaik so I guess the Southbridge is only one.

So, who's the faulty? I'm going to give back the brand new esata-sata cable to amazon but before that I'd like to get a definitive judgement from you.

It's possible to damage an enclosure, by connecting both
data cables at the same time.

Some enclosures use more than one chip, they have
a parallel interface, and chips share a bus. The
chips, when *one* cable is plugged in at a time,
only one output drives the bus.

When *both* cables are connected, two chips
try to drive the bus at the same time. This
can potentially damage the chip pad drivers.

I've had designs in the lab, where this kind
of conflict causes the center of the chip to
crack, and some corny looking smoke to pour out.
The chips in enclosures don't have that kind of
powerful drive, and if the chip suffers, it won't
crack like that. There might not be visible symptoms,
just a loss of function, as evidence.

This can *easily* be avoided for about $0.25
in the design, by adding a logic gate to prevent
contention. A responsible engineer would do that.

Modern designs are even more highly integrated,
there is only one major chip inside the enclosure,
and all of this is handled properly, internally,
in the major chip. It would be an older enclosure
(maybe with SATA I ESATA) that would be kludged like that.

I can't say why your enclosure is not working:

1) Could be a power issue. (Bad wall adapter)
We know this is not the case, because USB works.

2) Could be a disk drive issue.
We know this is not the case, because USB works.

3) No ESATA port driver on the mobo. Is Device Manager clean ?
The USB working case, doesn't tell us whether the
ESATA port has a driver.

4) ESATA port disabled in BIOS.

5) ESATA port hotplug disabled (only prevents detection
while the OS is running, does not prevent detection
while the OS is in the boot phase and is discovering
hardware anyway).

6) Unlikely to be a cable issue, unless the presence of
VBUS (cable with "ears") somehow causes the adapter
to shut down (conflict between +5V VBUS of laptop ESATA
versus +12V provided by wall adapter.

If you have a multimeter, you can try continuity testing
of the cable. I wouldn't bother with that, as my first
step. Cabling termination is done by fancy equipment now,
so humans don't get to stick their greasy fingers on stuff
like they used to do. One cable termination machine, and
a half-asleep operator, can produce a hundred times as many
cables as the greasy finger approach.

A number of these issues, could only be identified by
Googling the product model number and seeing what
other reviewers have discovered during "testing".

The last really bad enclosure I've had here, was probably
fifteen years ago. I don't buy a lot of these (half
dozen total lifetime), but the modern ones are past
the "bad wall adapter" era, and Just Work. I'm really quite
amazed that no bad wall adapters have showed up.

Paul