A7N8X-Deluxe 5 Volt possible problem

[dyno]

My son has an Asus a7n8x-deluxe with a Athlon 2500 cpu, 1 gig of ram, SB
live audio and winTV tuner card. The power supply is an Antec SL350 (350
watt). The system also has five 12V fans, two 80 gig HD and a pair of
CDROMs (infrequently used). Video card is an ATI 9800.

This system has been running without a hitch for almost two years. But,
in the last couple of weeks the system has started acting flacky;
noise/distortion on the TV audio and some stability/boot problems.
Because of this he brought the system home over the weekend to try and
determine the cause of the problem. In inspecting the system I found
nothing wrong until I checked the motherboard 20 pin power connection.
After unplugging the connector, I found the white plastic around all
four +5V pins significantly scorched. This is obviously not good. These
connections are overheating due to either excessive current draw,
corrosion (high resistence), or what? I'm trying to understand what can
cause this to happen. I believe the PCI slots are powered by the +5V
rail. Inspection of the two PCI cards (tuner and SB Live) didn't reveal
any obvious signs of a failed components. When I removed the MB and
looked it over, other than the discolored power connector socket there
were no obvious signs of overheated parts. Of course power regulators
and such can be bad without visible signs, but I kind of expected a high
current condition to leave some trace on the MB itself.

So far I replaced the power supply with a no-name 350w that I had on
hand and cleaned the light layer of dust off the motherboard and case.
For the last two days the system seems to working. But, for how long?

With the "new" power supply Motherboad Monitor shows the +5V supply
ranging from 4.78V to 4.88 V, perhaps a little on the low sigh but still
with the 5% tolerance for +5V. The 3.3V and 12V lines are fine. System
temperature doesn't exceed 50 deg C for the CPU. I did see one transient
event on the processor core voltage where it dipped to 0.7v from the
normal ~1.6V. This was a bit disturbing too. Does this point to a
voltage regulation problem on the MB?

My questions are:
What else on the MB is powered from the +5V supply? (I know the +5V goes
to the two HD installed on the system, but non of the 5V pins elsewhere
show any signs of overheating) I believe the CPU is powered by the 3.3V
rail(which showed no signs of distress).
Is the +5V indeed regulated via circuitry on the MB? If so, could a
failure on the MB cause an excessive current draw, but only on the +5V?

Before I/he just buys a new MB, I would like to narrow down (if
possible) the problem to one component. In the past I have played the
parts change game and ended up replacing parts that were perfectly good
(e.g., the spare 350W power supply).

 

[Paul]

My son has an Asus a7n8x-deluxe with a Athlon 2500 cpu, 1 gig of ram, SB
live audio and winTV tuner card. The power supply is an Antec SL350 (350
watt). The system also has five 12V fans, two 80 gig HD and a pair of
CDROMs (infrequently used). Video card is an ATI 9800.

This system has been running without a hitch for almost two years. But,
in the last couple of weeks the system has started acting flacky;
noise/distortion on the TV audio and some stability/boot problems.
Because of this he brought the system home over the weekend to try and
determine the cause of the problem. In inspecting the system I found
nothing wrong until I checked the motherboard 20 pin power connection.
After unplugging the connector, I found the white plastic around all
four +5V pins significantly scorched. This is obviously not good. These
connections are overheating due to either excessive current draw,
corrosion (high resistence), or what? I'm trying to understand what can
cause this to happen. I believe the PCI slots are powered by the +5V
rail. Inspection of the two PCI cards (tuner and SB Live) didn't reveal
any obvious signs of a failed components. When I removed the MB and
looked it over, other than the discolored power connector socket there
were no obvious signs of overheated parts. Of course power regulators
and such can be bad without visible signs, but I kind of expected a high
current condition to leave some trace on the MB itself.

So far I replaced the power supply with a no-name 350w that I had on
hand and cleaned the light layer of dust off the motherboard and case.
For the last two days the system seems to working. But, for how long?

With the "new" power supply Motherboad Monitor shows the +5V supply
ranging from 4.78V to 4.88 V, perhaps a little on the low sigh but still
with the 5% tolerance for +5V. The 3.3V and 12V lines are fine. System
temperature doesn't exceed 50 deg C for the CPU. I did see one transient
event on the processor core voltage where it dipped to 0.7v from the
normal ~1.6V. This was a bit disturbing too. Does this point to a
voltage regulation problem on the MB?

My questions are:
What else on the MB is powered from the +5V supply? (I know the +5V goes
to the two HD installed on the system, but non of the 5V pins elsewhere
show any signs of overheating) I believe the CPU is powered by the 3.3V
rail(which showed no signs of distress).
Is the +5V indeed regulated via circuitry on the MB? If so, could a
failure on the MB cause an excessive current draw, but only on the +5V?

Before I/he just buys a new MB, I would like to narrow down (if
possible) the problem to one component. In the past I have played the
parts change game and ended up replacing parts that were perfectly good
(e.g., the spare 350W power supply).

The processor is powered by +5V, at least it looks like it
from the power numbers I measured.

I own a clamp-on DC ammeter, and when running Prime95, with
an ATI 9800pro installed, consumption is [email protected], [email protected],
[email protected]. There isn't enough juice drawn on 3.3 or 12V, to
run the processor, so it has to be powered by +5V. (Note: Other
manufacturers use +12V to power the Vcore for the AthlonXP, so
how this is designed, is not a constant across the industry. It
is the whim of the designer.)

The video card has a separate power connector, and there is
additional current on there, which may not be applicable
to your situation. (The video card would add 3 to 5.5 amps or
so, to the load on +5V, but that current doesn't flow through
the 20 pin connector.)

When I measured the 16.6 amps on the motherboard connector,
that is determine by stuffing all four +5V wires into the
jaws of the clamp-on DC ammeter. That causes the currents to
add together (measurement is based on magnetic field, which
is additive). If I measure the wires separately, there is a
one amp difference between the wire with the most current,
and the wire with the least current. So, that should give
you some idea of how the current splits between leads. The
current is easily within the 6A max current allowed on the
20 pin connector, on any individual pin. (8 amps is allowed
on disk drive connectors.)

It could be the connector was loose, worked itself loose,
was a poor fit to begin with, and the connector pins went
into thermal runaway (pins heat, contact resistance increases,
pins heat some more, and so on).

Generally, when the surface of the pin is damaged, it is
only a matter of time before you'll begin seeing symptoms
again. At least, that is my experience with having to
repair, over and over again, a piece of equipment with this
kind of problem.

There are brute force solutions that can help. On a Tyan
motherboard, where the idiots who designed it, did not
calculate the total number of amps to expect on a dual
processor board, some users corrected a situation like
this, by soldering a disk drive cable to a +5V and a
GND contact on the motherboard (cut up a "Y" cable
to get parts). This gives you an additional connector
and pin to carry the +5V current, and helps take the
stress off the motherboard connectors.

An even easier solution is to RMA the board. It should
be easy for Asus to remove the old 20 pin connector and
put a new one on there, so the board will be salvagable.
For a home user, this is not an easy type of repair to do,
as a lot of heat would be needed to remove that connector.
The ease of removal, really depends on how tightly the
holes in the motherboard fit the pins on the connector.

Paul