Is my PSU at fault here

[Larkspeed]

Yesterday I did a pretty major upgrade to my machine and I have been running into problems ever since.

Previous setup:

Windows 7 Professional 64bit
Asus M2N68-VM Motherboard
AMD Athlon 64 X2 6000+ Black Edition Dual Core (running stock at 3.1 Ghz)
6GB DDR 2 Ram (4BG Crucial Balistix Sport + 2GB Kingston HyperX)
Asus ATI Radeon HD 4870
2 x 3.5 inch SATA Hard Drives (500GB Maxtor Diamond Max 20 + 640GB Western Digital Caviar Green)
1 x 2.5 inch SATA Hard Drive (Samsung drive can't remember model)
1 x 3.5 Inch IDE Hard Drive (don't remember model)
Liteon iHAS124b SATA DVD-RW.
4 x blue LED 80mm fans running on 7v
550w OCZ Fatality PSU

in a cheap and nasty asus case

This machine ran solid as a rock and never had any issues until a week ago when I started getting BSOD's.

Upon checking the event log all there errors were saying processor core so I figured the CPU was starting to fail so time for an upgrade.

Current Setup:

The same HD's, OS, PSU and Graphics card as above plus:

Asus M4N75TD Motherboard
AMD Phemon II X4 955 Black edition Quad Core (Running stock at 3.2 Ghz)
12Gb DDR 3 Memory (4 Gb OCZ + 8 Gb Kingston, can't remember part numbers)
3 x 120mm fans running on 7v

all in a cooler master RC-690 case

after the rebuild I did a clean install of windows

ever since I set this up I have been having random program crashes with the error "The display driver has stopped responding and has recovered"

Initially I blamed the Graphics Card so I swapped it out for an Asus ENGTX275 I had sitting here and it still does it.

I tried a registry tweak I found on the net to change how long windows waits for the GPU before it counts failure and resets the card. all this did was give me screen flickers.

I see not reason why the PSU should not be able to power the machine now so the only thing I can think of is that the PSU is failing.

I don't have a spare PSU to test with nor do I know anyone that has one. They are quite expensive here so I don't really want to go out and buy another PSU if I don't need it.

Am I looking at a duff PSU or have I missed something?

[Phage]

I.m thinking a duff ram stick. It would probably help if they were from a matched set as well.
I'd test with one stick at a time and see if it still does it, or download and run memtest.

[Larkspeed]

The memory has already been tested and is not at fault.

---edit---

I'll expand on that.

The error will always occur within 15 mins of use of the machine.

Memtest passed it and this memory was installed in my sons machine and he ran high load stuff for more than 4 hours with no issues

[c12038]

Try re-installing operating system as this maybe the fault as it got all the old drivers in from your old setup

[Larkspeed]

Try re-installing operating system as this maybe the fault as it got all the old drivers in from your old setup

From my original post:

after the rebuild I did a clean install of windows

[csgohan4]

Try re-installing operating system as this maybe the fault as it got all the old drivers in from your old setup

He did a complete new install of windows so unlikely the problem.

What are your temps on GPU and CPU ?

Try using Prime 95 to test stability of your CPU, did you over clock your CPU or GPU?

What GPU driver you using?

[deejayburnout]

dont think it will be your psu, 550w is enough for your rig.

mixing memory isnt always a good idea. are all the speeds the same and the voltages? i know you say that thay are fine in other pc but try it with each brand separatly and see how you get on.

also check you mobo bios is the latest and fully supports your cpu.

[Larkspeed]

Doubt it is a temp issue as it can happen even when the machine is sitting idle.

currently while I am browsing:

CPU = 37c idle and 59c running prime 95
MOBO = 34c
GPU = 60c (two different GPU's have been tried so kinda rules that out)

nothing is overclocked and CPU stability is fine

As for drivers current as of yesterday

[Larkspeed]

dont think it will be your psu, 550w is enough for your rig.

mixing memory isnt always a good idea. are all the speeds the same and the voltages? i know you say that thay are fine in other pc but try it with each brand separatly and see how you get on.

also check you mobo bios is the latest and fully supports your cpu.

As I already said memory has been tested fully both in this machine and my sons so that's ruled out

I have been mixing memory for years as long as you do it properly there is no problem. As you noticed my previous setup was also running mixed memory and it ran happily for nearly 4 years.

Mobo does fully support CPU

[csgohan4]

Doubt it is a temp issue as it can happen even when the machine is sitting idle.

currently while I am browsing:

CPU = 34c
MOBO = 37c
GPU = 60c (two different GPU's have been tried so kinda rules that out)

nothing is overclocked and CPU stability is fine

As for drivers current as of yesterday

Your GPU idles at 60C, that's a bit high. Try using MSI afterburner to set the Fan thresholds your self

[Larkspeed]

Your GPU idles at 60C, that's a bit high. Try using MSI afterburner to set the Fan thresholds your self

That's only this GPU if memory serves me correctly when the ATI card was in the GPU idle temp was about 45c

This graphics card has always ran a bit warm no matter what machine it's in.

[Phage]

Tested the HDDs ?

[Larkspeed]

yes drives have been tested and I have even unplugged all but the Maxtor 500Gb drive which is the C drive currently.

I did install windows on a different drive that belonged to my son to try that and I still got the error, which meant none of the current drives were actually installed at that time

[Larkspeed]

ok after exhaustive testing and a live chat with the support guys at OCZ I decided to look for a new PSU

I found a really good deal on a CoolerMaster GX-Lite 600W which seems like a decent PSU and should have more than enough power to run my system.

If that turns out not to be the problem at least I have a spare PSU now

[Behemoth]

How old is your current PSU mate ?

[Larkspeed]

just over the 3 year warranty