Motherboard kills memory

  • Thread starter Thread starter Al & Cay Grant
  • Start date Start date
A

Al & Cay Grant

Hi Folks,

I have built a few computers recently with ASUS motherboards.

One of the computers started to hang and eventually would not boot at all.
Replaced the memory and it was fine for about 3/4 months. Then it started to
intermittently hang again, so we again replaced the memory. This time it was
only ok for a day or so and started to hang again.

Obviously something is causing the memory to fail. Just wonder what the
likely culprit is? Motherboard faulty? Power supply?

Anyone?

Cheers

-Al
 
Hi,

If the mobo is still within the warranty period, I would
suggest you to take it for a thorough check. There may be
a faulty power supply to the DIMMS slots.

Peter
 
Al & Cay Grant wrote:
....
Obviously something is causing the memory to fail. Just wonder what the
likely culprit is? Motherboard faulty? Power supply?
Click to expand...
....
Compatibility between mainboard and memory?
You may wish to consult the QVL - qualified vendors list - available
as pfd download on the site of the respective mainboard at www.asus.com.

Roy
 
Roy Coorne said:
Al & Cay Grant wrote:
...
...
Compatibility between mainboard and memory?
You may wish to consult the QVL - qualified vendors list - available
as pfd download on the site of the respective mainboard at www.asus.com.

Roy
Click to expand...

More likely to be a motherboard problem. The QVL lists memory that
has been tested - most likely other memories work but cost too much
for Asus to test. So, absence from the QVL doesn't guarantee a
memory won't work.

If the fault has to do with some signal on the memory bus of the
motherboard, or the motherboard DIMM voltage regulator is putting
out a higher than normal DIMM supply voltage, this could contribute
to a failure. After killing several DIMMs, it is time to RMA the
motherboard.

HTH,
Paul
 
In comp.sys.ibm.pc.hardware.misc Paul said:
If the fault has to do with some signal on the memory bus of the
motherboard, or the motherboard DIMM voltage regulator is putting
out a higher than normal DIMM supply voltage, this could contribute
to a failure. After killing several DIMMs, it is time to RMA the
motherboard.
Click to expand...

Actually overvoltage is about the only thing the board can do
to kill the memory modules. The other possibility is inadequate
cooling, which would not be the fault of the mainboard. Not many
more ways to kill chips whit this speed. Maybe radioactivity....

Arno
Click to expand...
 
If the memory is really dead, then I'd vote that your 3.3V supply is
out of regulation on the high side. Being more than 10% over 3.3V can
significantly reduce the life of a part. It might work for a while
then something dies. You won't see it with the processor because there
is an on-board regulator that reduces the 3.3V down to 1.5 volt or so.

ab
 
In comp.sys.ibm.pc.hardware.misc Arnie Berger said:
If the memory is really dead, then I'd vote that your 3.3V supply is
out of regulation on the high side. Being more than 10% over 3.3V can
significantly reduce the life of a part. It might work for a while
then something dies. You won't see it with the processor because there
is an on-board regulator that reduces the 3.3V down to 1.5 volt or so.
Click to expand...

Actually the memory runs mostly with 2.5V these days. There is
usually a low-drop linear regulator (memory does not take much
power) close to the memory slots. You could have a look into the
PC-Health-Status (or such) in the bios. If the memoru voltage
is > 2.6V, this could be the reason.

Arno
 
Actually overvoltage is about the only thing the board can do
to kill the memory modules. The other possibility is inadequate
cooling, which would not be the fault of the mainboard. Not many
more ways to kill chips whit this speed. Maybe radioactivity....

Arno
Click to expand...

Or user just is buying generic chips that are crap to begin with..
Most memory comes with lifetime warranty's.. Send it back!
If it's user fault (overvoltage) on all of the chips then I would guess the
ram mfgr would say something
 
Actually the memory runs mostly with 2.5V these days. There is
usually a low-drop linear regulator (memory does not take much
power) close to the memory slots. You could have a look into the
PC-Health-Status (or such) in the bios. If the memoru voltage
is > 2.6V, this could be the reason.

Arno
Click to expand...

Is there a tool that can do this? ..
I know there are several dc regulators on a board but as I only keep my main
pc going then I hate to start sticking a probelead around to check stuff :)
 
Or user just is buying generic chips that are crap to begin with..
Most memory comes with lifetime warranty's.. Send it back!
If it's user fault (overvoltage) on all of the chips then I would guess the
ram mfgr would say something
Click to expand...

Yes, quite true! I did not think of that since I usually buy Infinion
Kingston. The problem could be chips that fail the usual tests
but work initially. Such chips are sold very cheaply on spot-markets.

Arno
 
Is there a tool that can do this? ..
I know there are several dc regulators on a board but as I only keep my main
pc going then I hate to start sticking a probelead around to check stuff :)
Click to expand...

System health display in the BIOS setup. Also some vendors
offer hardware monitor software for Windows for their boards.

Side note: I just discoverd that I can increase the memory
voltage up to 3.2V on my board in the BIOS setup! Definitely
a danger!

Arno
 
Back
Top