Re: Recent A8N-E installation crashes frequently

In message <hogOg.26843$G72.25113@xxxxxxxxxxxxxxxxxxxx>
"Unicorn" <unicorn@xxxxxxxxxx> wrote:

Thanks for your response!

A 512MB PC2100 memory stick from my old system.

This could be the problem - 512 RAM PC2100 is not the best for this mobo,
and it's not enough. I'd increas it to at least 1 GB RAM PC3200 (?)

Which is exactly what I have done, if you read down, - as two 512MB
modules from Crucial, to gain the benefit of the interleaving.

I have been carrying out some more runs of Memtest (the plus version
now). First of all I set the memory voltage back to 'auto' and ran
Memtest on test #5 only. Within an hour, it had logged four pairs of
similar faults to those logged last night. I then removed one of the
modules, and ran the same test for an hour and a half without any faults
being logged. I've just swapped to the other module and after 10 mintes
there are no faults logged.


Check here DDR SDRAM memory Dual Channel Memory Architecture
http://www.asus.com/products4.aspx?l1=3&l2=15&l3=171&model=455&modelmenu=2

I will look that out - thanks.

And check Power Supply just in case.

A good idea but, short of buying another PSU, what can I do? If I do
buy another PSU, are there any recommendations?



"Peter Bell" <peter@xxxxxxxxxxxxxxxxx> wrote in message
news:ef8c2c664e.peter@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
About a month ago I built up a new system comprising:
Asus A8N-E Mobo
Athlon 64 3500+ processor
MSI GeForce 6200TC video card
Samsung SATA 160GB drive
A 512MB PC2100 memory stick from my old system.
A 450W PSU, branded as 'X-Power'

Some of the peripherals have been retained from a previous
configuration, but I don't believe that they're implicated in the
problems I'm about to describe.

When I initially put the system together I attempted to load from a Win
2k, SP2 distribution. This would fail very early in the install and I
suspected some incompatibility between W2k/SP2 and the SATA disk system.

I then moved on to loading the system from a WinXP home SP2
distribution. This was not without problems, but I did manage to get
the system up and running.

However, the system was producing lots of BSODs, frequently
'DRIVER_IRQL_NOT_LESS_OR_EQUAL', implicating various system drivers (eg
HTTP.SYS, IPNAT.SYS).

First of all, I replaced the memory with 2 512MB PC3200 sticks, from
Crucial, using their recommendation for this particular motherboard.
This didn't appear to have any significant effect on the problem.

After some research I discovered that the NVIDIA firewall which comes
with the system was not the most robust of software and de-installed it.

This made the system almost usable - it will run for more than an hour,
but it is still a little fragile and certain operations are quite likely
to generate a windows error (for instance, Firefox will crash
frequently, whilst IE6 is reasonably robust) with the occasional BSOD
still occuring. The MS error reporting system almost always blames a
device driver for these problems.

More research turned up a recommendation of setting the memory voltage
to 2.7V, which I did yesterday, and I was still getting program crashes
and a BSOD.

I found Memtst86 and ran this overnight with the following results:

Tst Pass Failing Address Good Bad Err-Bits
5 2 0002eb4acf8 - 747.6MB ffffffff 7fffffff 80000000
5 2 0002fb4acd8 - 763.6MB ffffffff 7fffffff 80000000
5 7 0000eac64b8 - 234.3MB ffffffff 7fffffff 80000000
5 7 0000fac6498 - 250.3MB ffffffff 7fffffff 80000000
5 8 0001eee22b8 - 494.1MB ffffffff 7fffffff 80000000
5 8 0001fee2298 - 510.1MB ffffffff 7fffffff 80000000
5 23 0001ef073f8 - 495.4MB ffffffff efffffff 10000000
5 23 0001ff073d8 - 511.4MB ffffffff efffffff 10000000


Now, I find these results interesting:
Many passes are successful, but when there's a failure, there are always
two together, always ffffe0 locations apart, and it's only test 5 which
fails - the 'all ones' test. I was ready to believe that it was only
the highest bit which was failing until the pass 23 failure popped up
this morning.

Now, I have to believe that the operational problems I'm having are
related to this apparent memory problem - at least it's not worth
further investigations before I've fixed this.

So, I'm looking for some advice .... first of all, I trust that these
failures cannot be related to any peripherals connected via the PCI
system - this must eliminate the expansion card bus, as well as any
motherboard systems such as SATA, IDE, sound, network etc.

I guess the next thing to try is swapping the memory sticks over and
seeing whether the pattern changes. I could even go back to the old
single PC2100 stick and see what results I get with that.

However, I wonder whether anyone has any other suggestions - is this a
recognisable problem/pattern, is it likely to be a faulty memory stick
or a mobo fault (or, maybe, processor, or PSU)? I can't see anything
else being implicated here. All fans are running and system
temperatures appear to be low - usually reported as being around 35C.

All advice gratefully received, and thanks for reading this far.

--
Peter Bell - peter@xxxxxxxxxxxxxxxxx




--
Peter Bell (Note Spamtrap - To reply, replace 'invalid' with 'bellfamily')
.

Relevant Pages

  • Re: Dead PC - help please !
    ... Further diagnosis showed that one of the power supply fans was seized solid - so, on a wild hope, I purchased and installed a new PSU. ... The minimum system approach is a good one - but minimum is mobo processor and memory ... If that combination doesn't give any signs of life, other than HDD and PWR both being on permanently, then it is usually the mobo - unless the processor heatsink has fallen off... ... Perhaps better than putting it in the replacement system, where it will only add to the heat and probably not much to the functionality. ...
    (uk.comp.homebuilt)
  • Re: Your next Computer upgrade
    ... Haven't noticed the memory holding me back yet but obviously need to mess around a lot more yet. ... Installed the new PSU and planned out the wiring, fitted the CPU to the mobo and installed the mobo to the case - around another hour/hour and half. ... What's such a bitch about MS's updated page is that you have to DL like 62 patches BEFORE it even lets you install SP2! ... I DL'ed the BIOS and put it on my MUVO MP3 player. ...
    (comp.sys.ibm.pc.games.action)
  • Re: Multi CPUs Not Shown
    ... instabilities and operations after upgrading my MoBo, ... I was hoping to cure the instabilities that had been ... and ECS MoBo. ... number of CPUs recognized or the timing of memory or video display ...
    (microsoft.public.windowsxp.configuration_manage)
  • RE: Upgrading a MS 6382 Mobo System
    ... the Motherboard. ... You will also need to ensure that the new MOBO is compatible with the Power ... look at a picture of the MOBO and eyeball the number os scokets for memory. ... Most cases will provide for two or more HDD to be mounted above each other. ...
    (microsoft.public.windowsxp.hardware)
  • Re: DDR333 vs DDR400
    ... my mobo requires a low density module i.e unbuffered of any type ... "Although this motherboard specifies DDR400 compatibility, ... They are deliberately avoiding selecting any memory that is DDR400, ...
    (alt.comp.periphs.mainboard.asus)