MaxTheFast wrote:
Hi Paul, sorry for delay.
I've made other researches to get deeper into both sata and esata operating mode and these are my results:
- sata/esata electrical signals are slightly different: Tx sata 400-600mV; Tx esata 500-600mV; Rx sata 325-600mV; Rx esata 240-600mV;
- maybe it's my mobo's fault because of the following taken from https://sata-io.org/sites/default/fi..._WP_11-09.pdf:
"Since many of the existing disk drives and Serial ATA chipsets were originally designed with only the internal signal levels in mind, they may not be able to meet the more rigorous requirements of the external interconnect levels."

In other words my theory is: for sure my fantec enclosure affords an esata signal from its esata ports; fantec can't "understand" there's an esata-sata cable connected so the esata signal is carried to the mobo's sata; my mobo's chipset (maybe fw?) doesn't recognize the slightly different signal so it's as the fantec is not connected.
What do you think of that? If it's correct do you think there could be a way to set the bios properly? I looked into the bios but I found nothing about that.
Anyway I guess it's not a good way to sold something to cable's wires to try to modify the transmitted voltages, is it?

There *is* no ESATA mode.

I suspect that ports with so-called dual support (SATA or ESATA)
are *always* run in ESATA mode. This allows, say, an Intel
Southbridge to actually work with a longer SATA cable inside
the PC casing.

What the ESATA launch amplitude and receive sensitivity allow,
is SATA II rates on up to 2 meters of cable. It's something
like that.

And that's all that ESATA TX/RX capability does, is extend
the reach (double the cable length). Any other layers in
the hardware stack are unaffected. And the operating
system cannot look at that port and say "that's ESATA".
It really doesn't know.

I pulled datasheets before, trying and trying to find an
ESATA register setting in the PCH specs, and there is
nothing there. My conclusion is, it's always running
in ESATA mode.

*******

ESATA works without line buildout. Long transmission
lines normally need "compensation", otherwise known
as pre-emphasis. This allows long lines to have
compensation at the driving end, to make the signal
look better at the receiving end.

Now SAS has line buildout. It has a single bit for
selecting "short" or "long" lines. And SAS is similar
to SATA, except a *much* longer cable is supported.
And that's because the PHY has a more complex driver,
with a one-bit line length setting.

SATA/ESATA on the other hand, has no line buildout
option. Whether you use a 1 meter cable, a 12 inch
cable, a 2 meter cable, the SATA scheme has to work
for all of them, without any line buildout to compensate.
This means the eye diagram for the I/O looks different,
depend on cable, and the scheme has to work in every
case.

As near as I can determine, anything with ESATA capability,
simply stays in ESATA mode all the time. The device may
allow programming the Hot Plug capability, to avoid
annoyances if the user isn't actually Hot Plugging
the hardware. And this must be set correctly, for
external ESATA enclosures, if you want them to be as
flexible as USB3.

On "good" designs, the hard drive platter is in
spin-down state, before you pull out the cable.
That's what a Safely Remove should be doing for
an ESATA device. The same would be preferred
behavior for hard drives in USB3 enclosures.

WinXP had the best setup, in that the power
LED on the enclosure would go off when Safely
Remove was used. It was easier to tell there
was a state change. The power LED in that
case, was a way for the enclosure interface
chip, to signal "I've entered a low power state".

Paul