Advent 3117 PC mobo (Trigem 865PE)

[JEWilson]

Rec'd this mobo as known dead.
Was formerly used with a Bestec ATX-250-12E Rev F1

Have tried to start with known good PSU which does not have an on/off switch

Symptom - PS-ON will not go low.

The mobo front panel header for Power on will not start mobo.
Took reading with mmeter across the two pins, only drops to 4.88V

Shorted pins 13,14 on ATX (2 x10) connector - can get the PSU,
in turn, mobo to start but the southbridge is getting very hot. Albeit I shut
the system down from the wall outlet before the s/bridge gets
totally fried, the mobo did appear to POST.

Next, using freezer spray, cooled down the s/bridge for sufficient
time to observe mobo behaviour. Only had diskette drive attached
with Memtest86 v3.3 utility - this all goes smoothly and will load
the utility from diskette into memory and starts testing memory.

I have tested the mobo to the extent, tested ATX (2x10) connector
for ground fault on pins 3,5,7,13,15,16 and 17 all ok

The Intel part FW82801EB SL73Z - is it known to have a thermal problem?
Is the board exhibiting problem consistent with over voltage problem per
Bestec failires?

Any insight, help, advice and guidance appreaciated

[Harvey]

Re: Advent 3117 PC mobo (Trigem 865PE)

Was

formerly used with a Bestec ATX-250-12E Rev F1

Says its all really, its going to be a killed southbridge, the most obvious symptom being the overheating.

FW82801EB SL73Z

Hmmm, I've just recently been looking at a MSI 865PE NEO2-P board, that has a blown southbridge, seemingly from a design fault in the USB if http://www.devhardware.com/forums/mo...b-s-33530.html is accurate. It's southbridge is a FW82801EB SL73Z too... I wonder if there's a design fault in the chip??

[JEWilson]

Re: Advent 3117 PC mobo (Trigem 865PE)

Maybe

However, so far as I have forced PS-ON to start the PSU and
supply power to the board, the first time I did this, I got a
clicking sound consistent with starting into a short. The second
time I started the mobo, as I recall, the clicking sound disappeared
and has remained consistently so since. The sound (FWIW) came
from around the s/bridge and at first, I figured it might be the xtal.

Such a first short given the Bestec PSU may have resulted from an
over voltage condition which did for the mobo in the first instance.

So...

Given I have tested ground faults o/c on the mobo and have none,
I suspect a short had developed in the first instance which accounts
for the PS-ON being unable to be pulled to ground.

The clicking sound may have accounted for the original fault and that
further, shorting PS-ON to ground on the ATX connector to get the
PSU and mobo to start may have generated a further short such as
with a supply cap on a rail which delivers supply to the s/bridge.

This may account for the established (that being so) and/or further
short as suggested, developing the heat in the IC.

Looking at the part
see further http://www.alldatasheet.com/view.jsp...word=FW82801EB

Having investigated the part, the IC can tolerate temp up to 109 deg C.
IC supply voltage should be 1.8V albeit it can supply 3.3V I/O and tolerate up to 5V with some of its buffers (see p1 of 84)

The power pins detail is shown at p48 of 84 @ 3.2.20 Power and Ground at Table 26 as well as singalling for power at p42 of 84 @ 3.2.10 Power signals at Table 16
Note VRMPWRGD - should be connected to the processors VRM Power Good

Turning to regulation, p51 of 84 @ 3.4.1 Power planes and Table 29
Vcc3_3 IC I/O is powered at 3.3V by PSU
Vcc1_8 IC Logic is powered at 1.8V by PSU
V_CPU_IO is powered by the main PSU via CPU VRM

So far as the PS-ON failure points to a mobo logic problem and
the s/bridge IC develops a thermal problem when the first issue
is circumvented, my attention is drawn to regulation of the IC.

Where and how is 1.8V synthesised? The ATX PSU does not deliver
a voltage such as this. If any cap are s/c- really need an ESR meter
to do this job.

Any suggestions?

[JEWilson]

Re: Advent 3117 PC mobo (Trigem 865PE)

Sorry

Got distracted

I do have one of those boards and was planning to work it
up for production use. These are regarded as being a high-spec.
mobos so I believe. Of course, the matter with repeated failures
on MSI 865PE NEO2-P MS-6728 re - USB, tells me at this stage;
just a guess please note, due to, very probably bad caps vis-a-vis
a cap or caps failure.

Get a BOM of the electrolytics and establish whether (not) they
are of an acceptable quality. To do this, refer to the good
https://www.badcaps.net/forum/showthread.php?t=414
and bad
https://www.badcaps.net/forum/showthread.php?t=388
caps lists respectively.

You may also want to consider the power sunk by a board of this
spec. under light and full load conditions as well as the number of
devices attached such as with;
USB
1394
LAN
HDD
FDD
Fans
and what PSU you decide to mate these with.

For a quick giude to power usage, you could go to
http://www.extreme.outervision.com/p...ulatorlite.jsp
This is the free, lite version - you will have to pay to get the
pro version which details guidance re - current draw on the
relevant PSU rails.

HTH