Is there a known "cold-start" problem (motherboards with Nvidia chipsets, Core-2) ???

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

Threaded View
There is an Asus motherboard (P5NSLI) that has been on the market for
a year and at some point a revision 2.00G version of it came out (not
sure when or why).  This board is based on the Nvidia 570 chipset.

I'm speculating that the Rev 2 has something to do with Core-2

Some percentage of these boards have a "cold-start" problem, whereby
when you try to start it when it's fully cooled to room temperature,
it doesn't start (doesn't post) but if you leave the power on for a
minute or two and hit reset then it will start.

Some say that boards that do this will do it when a Core-2 CPU is
installed but work fine if it has a Pentium.

I'm wondering if anyone else has seen this on socket 775 motherboards
that have been modified for Core-2, or perhaps only for boards with
the Nvidia 570 chipset, or only specifically for the Asus P5NSLI.

And before you reply about possible causes, be aware that the P5NSLI
forum on Asus's web site details several people testing problem boards
in various configurations (different power supplies, mounted in a case
vs on a table, different RAM, etc).

Re: Is there a known "cold-start" problem (motherboards with Nvidia chipsets, Core-2) ???


Quoted text here. Click to load it

Unless you find all other equivalent boards from other
brands, same chipset, to exhibit the same problem, it would
be limited to an Asus specific bios or board design or
defect type of problem, as the chipset isn't the variable.

Re: Is there a known "cold-start" problem (motherboards with Nvidia chipsets, Core-2) ???

Quoted text here. Click to load it

Could there be a short somewhere in the circuitry on the motherboard?
If there is a break in the circuitry it may be so small that when
everything heats up and expands it closes the gap between the broken
circuits and everything works as it should.  When the motherboard
cools off the circuits contract and the short comes back.

Re: Is there a known "cold-start" problem (motherboards with Nvidia chipsets, Core-2) ??? wrote:

Quoted text here. Click to load it

This is being seen by several people (on the Asus forum as well as
newsgroup), and I for one have seen it happen on 3 out of 6 P5NSLI
boards purchased 2 months ago.

Are you saying that it could be the same trace problem on all these

Re: Is there a known "cold-start" problem (motherboards with Nvidia chipsets, Core-2) ???


Quoted text here. Click to load it

Have you throughly fixed the variables?  I mean when you
write that it wouldn't work when cooled down, off, but would
after leaving it on and hitting reset, is this always the

If you turn it completely off and wait 10 seconds or so,
does it come on?
If it doesn't, what if you pull the AC power cord (or flip
the PSU AC line switch if so equipped) right after powering
off the system, then after a dozen seconds or so you plug
power cord back in, give it a few seconds and then try it?

One thing I'm attempting to isolate here is the issue of
heat or time off, versus that when you get it going you
aren't turning it on, rather resetting logic while it was
already electrically "on" to a certain extent.

With the system sitting off and cold, you might also measure
the 5VSB voltage.  Check the PS2 and USB power jumpers, and
if jumpered to 5VSB, rejumper them to 5V.  Also try
unplugging all peripherals, so the system is in a minimal
config using only video, 1 memory module, CPU, heatsink/fan.

Is it true that it always works if a prior CPU to Core2 is
installed?  I'd begin to wonder about power requirements, if
none of the PSUs installed had adequately fast voltage rise,
if their 12V rails, regardless of the labeled current
capacity, just weren't beefy enough.  Along this line of
thinking, a system with few (Or better still, none) hard
drives connected and a lower powered basic video card (even
a very old PCI video card if you had one available) might be
put in temporarily to see if it has any effect.

I'm suggesting that it may not be a straight line to the
solution unless it is merely a bios bug and a flash would
resolve it.  Rather, a broader array of attempts might be
undertaken and and results recorded so that some commonality
might be found.

As for heat vs traces or solder joints, one thing you might
do is test after gently warming up the board with a hair
dryer (or heat gun, held quite a bit further away, the goal
being only small heat increase not "hot" to the touch

Site Timeline