On Mon, 12 Jan 2009 17:27:13 -0800, Archimedes' Lever
wrote:

On Sun, 11 Jan 2009 01:57:20 -0500, kony wrote:

On Sun, 11 Jan 2009 00:06:31 -0500, kony
wrote:

Here's an example with an ECS GeForce 6100SM-M v. 1.0 motherboard:

http://farm4.static.flickr.com/3397/...b349938d_o.jpg


snip

I don't have an nForce /4xx series board available to check
it, but checking an old A7N8X board that uses a Silicon
Image SATA controller, it has the SATA data pins coupled
directly to the chip inputs with a 2200pF ceramic chip
capacitor in series.

The picture linked above is a little high in contrast so
it's hard to tell but might those empty spots be
corresponding to capacitors on the adjacent used SATA spots?
Further, if you can trace these data lines to the chipset,
are there unused surface mount pads adjacent to it?


I must be blind, upon looking at the picture again obviously
what I was thinking of is supposed to be capacitors as
marked with the C(nnn) silkscreening but now I wonder if you
were looking at the resistor, R(nn) silkscreened positions
above the SATA port in the pictures. They don't seem to be
for SATA?


Ever thought about tracing the local circuitry with a DVM? D'OH!

Actually, I did trace the circuit and also reported what I
measured in a separate post, but on a board I have that is
not the same make and model as his, and if you are familiar
with the extreme economizing ECS often does on their boards,
you would then realize that what is normal is not always to
be taken for granted when ECS is involved.

However, taking measurements with a DVM is a bit less simple
than that in many instances since the parts are still
in-circuit, potentially still in parallel or series with
other parts unless one starts desoldering each one or can
read and decipher the markings with a magnifying glass.

However, on my other post where I listed capacitance values
I am fairly confident there was not need as one end of the
circuit was open at the SATA connector itself. This is
where working with multi-layer boards becomes more complex,
tracing a dense circuit enough to follow it within a
reasonable amount of time, not merely soldering a connector
or SMD cap on the board.


Especially useful at the unpopulated areas for schematesizing the work
in progress. Pretty much leaves out all doubt.

A schematic leaves no doubt... if we can assume it was
followed. Otherwise there are several hints but the first
step is the one not so hard to do, put the port on the board
after populating the missing cap locations with parts
mirroring those on the other implemented ports.