Ed M. wrote:
> "Erratic" <Erattic RemoveThis @no.spam> wrote in message
> news:46f32e26$0$8426$db0fefd9@news.zen.co.uk...
>> While finally close to rock stable my latest upgrade/rebuild hasn't gone
>> too well and am coming to the conclusion my P5N32-E SLI may be to blame
>> (almost everything else has been swapped/replaced or removed). Anyhow...
>>
>> Components are :-
>>
>> E4300 Core 2 Duo (o/c to 3.0GHz)
>> 2GB PC6400 Crucial Ballistix Tracer
>> 320MB 8800GTS
>> 3xHDD and 2 DVD-RW Drives
>> 700W PSU
>> Kandalf LCS Case (Water cooled CPU)
>>
>>
>> I don't need SLI/Crossfire so motherboards based on P35 chipset are fine,
>> I was going to get the P5K but a number of suppliers have the Blitz
>> formula only £10-20 more and as I've just dipped my toe into watercooling
>> the Blitz Formula is just a little more appealing with its water cooled
>> chipset option.
>>
>> Niether board is cheap and both have good reviews, I need the board to be
>> a good O/C and to be able to take a quad-core or better in the future.
>>
>> Any advice appreciated.
>
> That is a pretty good OC on the 4300. As Phil says, a little more info might
> help a little. I had the same MB on my E6600 build at first and it was fine
> until the SATA controller failed. That is a known issue, but most worked out
> fine. My replacement was a Striker Extreme with the same chipset and a few
> extra bells and whistles. I have had pretty good luck overall with the 680i
> chipset. I took my E6600 out for a build for my son and now have a Q6600 (C0
> stepping) in the Striker running at 3.37Ghz and it is fast.
>
> Ed
>
>
Problem is random lockups when the graphics card comes under load,
either games or video (after 5-30 seconds). This is regardless of video
card, tried 6800, 7900GT, 8800GTS and couple of ATI cards.
I suspect something to do with the PCI-E part of the chipset when put
underload. I dont think its heat as powering off the system (but
keeping the board powered) and allowing everything to cool will *always*
result in the same lockup next time video is put under load. However
just switching the power off from the board for 5 seconds and back on
will usually cure it. To me this looks like some incorrect internal
setting on the board is surviving the crash and only powering it off
completely resets it to the working state.
Apart from this issue the system is rock stable even under sustained
load 24/7 (BOINC) and passes various stability tests without missing a
beat (48hrs+).
I run the CPU at 325x9 (2925Mhz) at 1.4125v in BIOS (1.39v reported),
RAM at 940Mhz (4-4-4-12) at 2.25v.
CPU will POST ok up to a little over 3150 (350x9) but fails prime or
simply reboots after at most a couple of minutes when over about 3060
(usually in less than 10 seconds), increasing voltage makes no
difference. 356FSB seem to be about the limit for this CPU on this MB
even dropping the muliplier it wont POST.
RAM is actually ok up to around 972Mhz at same voltages and timings
above but I get strange black screen flashes at random, pauses lasting 2
to 20 seconds where the screen wont update and occasional BSOD. This is
despite passing 72 hours memtest86+ without problem and working fine on
a friends board.
PSU(s) I've tried are Thermaltake 700W Toughpower, Antec 460W and
Jeantech 700W STORM. System is pulling only about 260W max (according
to the meter on the STORM PSU). Voltages are very stable.
MB voltages are HT=1.3v, NB=1.45v, SB=1.55v and VTT=1.55v, all other
voltages are auto.
Temperatures (CPU/MB) are 58C/45C under full load and 30C/42C idle/low load.
First P5N32-E SLI I had died after about a month, first it got VERY
flakey then got some strange voltage readings including warning of
overvoltage on the 12v line (BIOS said 27v!!), not surprisingly a week
after this the BIOS chip physically burnt up!).
Another week after getting the replacement board one of the RAM sticks
died completely.
About 2 weeks after this the 7900GT i was using started to flake out.
>> Stay informed about: Asus P5K Deluxe Wifi or Asus Blitz Formula? 
FAQ
Profile
Private Messages
Log in
