bsod help

Help!
I have built my new system around my new ecs C19-A SLI mobo but after installing windows I have periodic bsod errors that happen normally at periods of high load.
The errors are different each time and range from page "error in non paged area" bsods to "IRQ not less or equal"
My system is:
C19-A SLI mobo
p4 630 3.0ghz 800fsb 2mb cache
1gb infineon 533mhz 4300 ddr2 ram
160gb seagate hd
xfx 7900gs 256mb vga
570watt branded psu

The processor runs at 65 degrees at full load and 49 idle and system temp is low at about 40.

I have used memcheck which found no errors in the ram, the hard drive works fine with no errors in another system, the errors occur without the vga card plugged in and on a previous mobo I was having similar problems. These errors also happened during windows xp install so it is not an issue with any drivers I have installed.

What I need to know is whether these errors could be caused by a faulty processor and how to check without obtaining a new one. Is there a way to isolate exactly which component is to blame by reading the error logs in event viewer? Could the temp of the ram be to blame as errors occur at high load after about 30 minutes.

Any help with this would be greatly appreciated as I want to play Medievel Total War for more than 10 minutes!!!!!

Thanks guys!

65C is way too hot, you'll need to beef up the cooling.

Quote:

---

Originally Posted by aidanjt

65C is way too hot, you'll need to beef up the cooling.

---

I disagree. It is a bit higher than ideal, but not high enough to be the source of the problems.

In any case, if heat is the cause it is easy to test, just run your system with the cover off, and direct a large desk fan at the CPU cooler.

I'm sure the cpu overheating is not the problem, I am more worried about the ram temp being the problem. Is this a common problem?? The thermal spec of a p4 630 chip is about 80 degrees, does anyone know what the ram is?

Run memtest86 and find out, you can find it as part of most linux distro cds and livecds.

or just download the small memtest ISO so you don't have to download a 600 meg linux distro just to use memtest.

Quote:

---

Originally Posted by aidanjt

Run memtest86 and find out, you can find it as part of most linux distro cds and livecds.

---

Read his post, before posting. he says that his memory has already passed the memtest errors.

Have you changed anything in your system recently or updated any drivers? Can you remember your config the last time the pc was stable?

Quote:

---

Originally Posted by mathewdbarlow

Help!
I have built my new system around my new ecs C19-A SLI mobo but after installing windows I have periodic bsod errors that happen normally at periods of high load.
The errors are different each time and range from page "error in non paged area" bsods to "IRQ not less or equal"
My system is:
C19-A SLI mobo
p4 630 3.0ghz 800fsb 2mb cache
1gb infineon 533mhz 4300 ddr2 ram
160gb seagate hd
xfx 7900gs 256mb vga
570watt branded psu

The processor runs at 65 degrees at full load and 49 idle and system temp is low at about 40.

I have used memcheck which found no errors in the ram, the hard drive works fine with no errors in another system, the errors occur without the vga card plugged in and on a previous mobo I was having similar problems. These errors also happened during windows xp install so it is not an issue with any drivers I have installed.

What I need to know is whether these errors could be caused by a faulty processor and how to check without obtaining a new one. Is there a way to isolate exactly which component is to blame by reading the error logs in event viewer? Could the temp of the ram be to blame as errors occur at high load after about 30 minutes.

Any help with this would be greatly appreciated as I want to play Medievel Total War for more than 10 minutes!!!!!

Thanks guys!

---

You say its a 'branded' PSU.....which one and what are the stated rails on it?

Quote:

---

Originally Posted by chrestomanci

I disagree. It is a bit higher than ideal, but not high enough to be the source of the problems.

In any case, if heat is the cause it is easy to test, just run your system with the cover off, and direct a large desk fan at the CPU cooler.

---

I disagree aswell.
Although thats hot, its still not overheating enough to cause BSOD.
The thermal throttling would occur and shut down the system before a BSOD would appear.

1. WHat memory timings have you got (is it OC'ed as this is classic memory timing symptoms whether memtest has passed or not).
2. Tell us about this PSU you have.

Its definitely one of the two thats causing this.

Quote:

---

Originally Posted by s_kinton

Read his post, before posting. he says that his memory has already passed the memtest errors.

---

No, he said memcheck, not memtest86. Read his post before posting.

If memtest cant turn up any errors then you don't have problems there.

Quote:

---

Originally Posted by aidanjt

No, he said memcheck, not memtest86. Read his post before posting.

If memtest cant turn up any errors then you don't have problems there.

---

I did read the post......and you have said this 3 times needlessly already.

Like i said, and forget memcheck for a minute if you can.
Start swapping modules in and out.

Mem tested or not,.......this IS classic memory syptoms.

If you think its heat related fella you are way off the mark.

memtest86 not memcheck.

They're two entirely different programs.

memtest86 isn't operating system dependent/limited.

Right!
The power supply is a trust pci-e 570watt. Could this be the problem???

The ram is not overclocked and bios shows it is being supllied with 1.8v.

As for installing new drivers etc, this cannot be the problem as it is a clean install on a new hd and problems also occurred during xp installation

http://www.trust.com/products/product.aspx?artnr=14888
This is the link to manufacturers spec page for psu

run memtest for a good few passes (take a few hours) just to gain confidence in it.

Exactly

You can get a skinny memtest ISO at http://www.memtest.org

Just burn it to disc, and reboot.

Have you any usb devices connected to your system, eg.broadband modem. Does the fault still occur with them disconnected?

Quote:

---

Originally Posted by egg

Have you any usb devices connected to your system, eg.broadband modem. Does the fault still occur with them disconnected?

---

Good point egg. The Alcatel speedtouch 330 is particulary notorous for causing BSOD.

To be sure, I would disconnect all USB periferals except for your keyboard and mouse

Unfortunately it can't be usb because I tried system on minimum set-up and even disconnected the usb leads from the mobo.

Getting annoyed now. Blue screens seem to be getting worse, especially at high load.

Is it possible that this could be a problem with the processor?? I know page faults and irq errors are normally associated to RAM but I think I have eliminated the ram from being the problem as the same happens if I remove one stick or the other (it is highly unlikely the same fault could happen with 2 different sticks of ram)

You know BSoDs do happen just because of a crap driver, in fact most BSoDs happen because of that, it could also happen if the kernel is corrupt, when was the last time you reinstalled windows?.

It is a fresh install of windows and it happened during windows install also so i don't think it can be a driver issue. Can anyone suggest a decent diagnostic utility or a way I can check that the cpuis not faulty without having to replace it.

Whats the BSOD error code? Sometimes you can get an idea of whats wrong from putting that into google.

This is the problem, the error codes are not consistant and I am recieving a different one each time, ranging from 'page fault in nonpaged area' to 'irq not less or equal'. The stop codes are invariably 100000050, 100000008e, 100000007e etc but no one is more frequent than another.

Does this seemingly random pattern of errors point towards dodgy or incompatable ram? I am going to take a chance and replace the processor because I have tried removing a stick of ram at a time and using different slots and the errors still happen. ITS STARTING TO PISS ME OFF!

have you run memtest86+ yet?

You can download a small iso image and make a bootable CD. I would not trust any other memory tester, especialy not anything that runs in windows.

It will find most memory errors very quickly (less than 5 minutes) but to be sure you should let it run for some time to make sure it finds everything. At least overnight, preferably 24 hours or more.

Let us know the outcome.

Yes, have run memtest 86 for 5 passes. Will try running for longer and let you know the outcome.

Do you think that these type of faults could be the result of a faulty processor? I have heard that defective cpu cache, of which my processor has 2mb could be the culprit.

Thanks in advance!

If memtest86+ has run for 5 complete passes and not found anything, then it is unlikely you have faulty memory. The other possibilities are your CPU, motherboard or hard drive.

The only way you can eliminate your CPU or motherboard are to try swapping them out. It is fairly irrelevant to discuss which parts of your CPU might be faulty as you cannot replace or repair those parts.

You might try booting from a knoppix disc or the like and seeing if you get similar crashes from linux. If you don't it would suggest that either your hard drive is somehow corrupting the windows files stored on it, or alternatively a bit of faulty hardware is confusing windows and making it crash, and that linux is less susceptible.

Unfortunately it is very hard to work out what the problem is. If you brought your CPU, motherboard and memory from the same supplier, I would consider getting an RMA on all three, as it will be very hard to identify the culprit.

Quote:

---

Originally Posted by mathewdbarlow

This is the problem, the error codes are not consistant and I am recieving a different one each time, ranging from 'page fault in nonpaged area' to 'irq not less or equal'. The stop codes are invariably 100000050, 100000008e, 100000007e etc but no one is more frequent than another.

Does this seemingly random pattern of errors point towards dodgy or incompatable ram? I am going to take a chance and replace the processor because I have tried removing a stick of ram at a time and using different slots and the errors still happen. ITS STARTING TO PISS ME OFF!

---

It does sound like a memory problem too me. Although it could be caused by all the thing mentioned before. Have you tried upping the voltage slightly to the ram?

I haven't messed around with voltages at all. My knowledge of voltages / overclocking is pretty limited and I didn't want to do something wrong and fry something.

My voltage control is set to automatic in the bios, how much would be reasonable to increase this by before I am risk of doing any damage?

It might be worth mentioning that I tried adjusting the speed down to 400mhz but this causes the machine to restart and then it is unable to post. This can only then be rectified by clearing the bios cache.

This Branded PSU - how much did it cost?

Lots of people saying the psu might be the culprit. Its a cheapy i'm afraid. It only cost about £30 which is stupidly cheap for a 570w. It only gives out 16 amps on the 12v and I had a feeling that this might be the problem.

However, the errors still occur with a thermaltake 450w supply that works fine in another system so I have eliminated the psu from the investigation.

Whilst we're on the subject, I am going get a psu. Can anyone recommend a good quality one that comes with a 6 pin pci-e connector for less than £50?

£30 for a 570 watt PSU! There is your (most probable) answer - your PSU. If a £30 PSU can do the job well why do you think other manufacturers sell a similar capacity one for £90 or even £110? There must be a good reason or no one would buy them.

I believe more hangs and crashes of consumer PCs are due to poor PSUs than probably any other hardware related issue.

A PSU is vitally important to a PC, the PSU is it’s heart. They not only supply electricity, they have to provide the right amount, and reliably and stability is hugely important.

Get a really good (not half good) PSU. I recommend Seasonics, others will recommend such as Corsair (built by Seasonic but to a different recipe and not quite as good.... but nearly). There are a good half dozen quality PSU manufacturers.

I suspect you might be right. I turned up the voltage in the bios for the ram and it has made it alot better (although the problem is still there) I shall have to wait til payday to purchase a new psu and then I will let you know of my findings.

I have never had this problem before even though I have built many systems, its typical of the way things are going I suppose, with higher spec machines soaking up more and more juice. In future I will never be seduced by high wattage, cheap psu's and will always go for a established brand.

Thanks very much for your help anyone who replied, without such a good response I think I would have replaced every part but the power supply and would have thrown the whole lot out of the window by now.

Quote:

---

Originally Posted by mathewdbarlow

Lots of people saying the psu might be the culprit. Its a cheapy i'm afraid. It only cost about £30 which is stupidly cheap for a 570w. It only gives out 16 amps on the 12v and I had a feeling that this might be the problem.

However, the errors still occur with a thermaltake 450w supply that works fine in another system so I have eliminated the psu from the investigation.

Whilst we're on the subject, I am going get a psu. Can anyone recommend a good quality one that comes with a 6 pin pci-e connector for less than £50?

---

The Thermaltake 450W is historically a VERY poor PSU and no better (if not worse) than the original one you were using.

How many amps do you need on the 12v rails to power a system set up like mine (in system specs) reliably?

I have had a look at a few units and they range form 14 - 24amps? Is there a minimum I should be looking for? The same goes for the 6 pin pci-e connector, is there a minimum spec that I should be looking for to power my 7900gs?

16 amps should be enough.

Someone mentioned ripples or something before, is this basically fluctuations in the power suplly getting through and is this what causes the errors. This may expalin why the errors occur more frequently when playing intensive games etc.

Ok if I was looking for a new powersupply for around £50 I would get a PSP Blue Storm 500W £51 from komplett.

http://www.komplett.co.uk/k/ki.asp?s...798&view=links

Here's the manufacturers info page :

http://www.fsp-group.com.tw/english/...d=49&proid=179

Here's a review :

http://www.neoseeker.com/Articles/Ha...bluestorm500w/