Is my Motherboard Frying my RAM?

[k@rt]

Hi All.

I recently replaced a dead motherboard in my flatmates Packard Bell iMedia x9166. The specs of the original PC are as follows:

Intel C2Q Q9300
nVidia 9500 GS
4Gb DDR2 @ 1333Mhz (pretty sure speed is correct here)
FoxConn Mainboard - LGA755 - DDR2 (exact specs hard to find, but this is the same)
500Gb WD HDD
250 Watt PSU
DVD drive, card reader etc.

I replaced the broken FoxConn Motherboard with a Intel DG31PR Mainboard

At the same time I also installed the graphics card from my old PC which was a 8800 GT (better than the 9500 GS) which also necessitated using the more powerful 500Watt PSU from my old PC as well.

So the specs of the new fixed PC are:

Intel C2Q Q9300
nVidia 8800 GT
4Gb DDR2 @ 1333Mhz
Intel DG31PR
500Gb WD HDD
500 Watt PSU
DVD Drive, card reader etc

So I rebuilt the PC a couple of weeks ago, installed windows vista (32 bit Home Premium) and since then the computer seems to have been working absolutely fine with no problems. Until today.

This morning my flatmate tries to start the PC and it wont boot. I try and get a 3-beep bios code which after confirming on Intels website points to a problem with the memory.

So I pulled out the 2 memory modules and replaced both of them with the memory modules from my old PC (Kingtson DDR2 - but only 3Gb, one stick 2Gb and one stick 1Gb)

The PC started immediately so I assumed that simply one of the old memory modules that were still from the original PC died. I didn't have time to look any more at the problem and assumed it was basically okay.

However the PC worked for another six hours and then it died again!!!

I have managed to boot the PC again now using only one of the original 2Gb memory modules in slot 1... However what I am now really concerned about is why this happened. It seems REALLY unlikely to me that 2 completely different memory modules of different ages and histories would randomly choose to die within 6 hours of each other. So I am worried that it is the system that is killing them in some way.

So my main question is to ask if it is possible that the Intel Motherboard I used to fix the PC is somehow damaging the RAM?? I am not an expert but I checked the specs pretty carefully for the mainboard and I think that everything was correct. I am really worried about just going out and buying new RAM in case that gets fried as well.

Can I test this in some way? Is there any way for me to diagnose at home whether the mainboard is burning out the RAM? I looked in the bios but there is no info there about voltages and such for the RAM, only the fact that it's present, how much there is and the frequency it's running at.

Is there anything else that could be killing RAM modules? I obviously also put in a more powerful PSU - but assuming it is not functioning incorrectly that shouldn't be any kind of problem should it?

Can I use CPU-Z (or something similar) to monitor the RAM and see if it is being stressed in some way or functioning incorrectly? (as it stands I can still get the PC to boot with a 2Gb module in place) If it is possible monitor the RAM like this what should I be using and what problems should I be looking for?

If the info helps I got the mainboard from an ebay dealer who seemed to have a quantity of them and is advertizing them as "Reconditioned". Is it possible that simply the board itself is faulty?

The first two modules worked fine for two weeks and the other two burnt out within a few hours, but the ones that burnt out within a few hours were 2+1 Gb which may have had an effect?

I am really at a total loss to know what to do here 🙁

I don't know whether to try getting new memory modules, or to try to contact the ebay selller and return the motherboard if I can. I don't see what (if anything) I could have done wrong to cause this problem, but I admit to not being an expert either. All the components that I used from the old PC were basically working correctly when I tested them.

Any advice, suggestions or information would be greatly appreciated.

Thank you!!

 

[ingtar33]

a bad power supply will kill your ram. not so sure about a bad motherboard. typically only the power supply has that type of effect.

 

[k@rt]

Great thanks. Is there any way I can actually diagnose that to confirm whether it is the problem or not? Maybe some piece of software that can monitor the power supplies directly and tell if there is a problem? The PSU is fairly old, but seemed to be working okay until recently (obviously proves nothing). I also gave it a very good cleaning when I put it into the new, fixed PC.

Will anything show up in CPU-Z if something is wrong? Thing is I would rather not have to go out and start buying new RAM modules and PSU in the vague hope that the problem comes from that... it would be really nice if there were some test that could be done which, if not proving conclusively that the PSU is at fault, would at least point more in that direction.

The reason I thought this may have something to do with the Motherboard is that although the power for the RAM comes from the PSU I thought it was the Motherboard that controlled how that power is delivered to the RAM itself... all of the voltages running through the ATX cable are much higher than RAM itself requires, after all.

 

[k@rt]

IMPORTANT UPDATE - I have just discovered that the RAM modules I thought had been fryed and broken are actually not. I have just tried putting each of the four DDR2 RAM modules I used into Slot1 and the PC booted each time. This is strange because earlier I tried exactly the same thing with one of the RAM modules and it wouldn't boot, which is why I assumed it was that module which had died.

I am now really confused and will have to run more tests tomorrow when I am fresh, but could this be a problem when both slots are used? Is it possible that the memory was never part of the problem at all? I only went to memory because of the bios beeps I got this morning, and that only happened once. What else could cause the PC to fail to boot, then work fine, then cause it to fail to boot again?

 

[ingtar33]

IMPORTANT UPDATE - I have just discovered that the RAM modules I thought had been fryed and broken are actually not. I have just tried putting each of the four DDR2 RAM modules I used into Slot1 and the PC booted each time. This is strange because earlier I tried exactly the same thing with one of the RAM modules and it wouldn't boot, which is why I assumed it was that module which had died.

I am now really confused and will have to run more tests tomorrow when I am fresh, but could this be a problem when both slots are used? Is it possible that the memory was never part of the problem at all? I only went to memory because of the bios beeps I got this morning, and that only happened once. What else could cause the PC to fail to boot, then work fine, then cause it to fail to boot again?

I think you need more information. I was working with a client today with some strange issues on her computer. the problem was i couldn't replicate the errors. once i started to play around on the system all the problems went away. Nothing i did other then rebooting the system.

Well once the symptoms go away you have to make them come back to figure out what's wrong. i never could get them back so i couldn't figure out what was wrong (well that's not true. i did check for the usual suspects, came up empty each time, but at least i checked out what i could)

in your case without the system failing to boot you can't find out what's causing the issue.

 

[k@rt]

This is my feeling exactly ingtar33, when something simply stops working then you can go about the process of eliminating things one by one until you find what is broken and replace it. However when one time it works and one time it doesn't then it is really hard to know what to do.

But I did collect some information this morning that (may) be helpful, there is certainly one thing that came out of it I would like to ask about.

Firstly after confirming individually that each RAM module was working, I simply put the PC back exactly how it was yesterday when it stopped working the first time (with the 2x2Gb original RAMs) and (of course!!) it booted fine. Then using CPU-Z and a program called SpeedFan I collected some data about the memory and voltages.

Here is a dropbox link to a jpeg showing all the information I collected that I think maybe relevant: https://www.dropbox.com/s/urjcr5exjn2w3fu/PC_info.jpg?dl=0

I am not an expert but I think that all the voltages that SpeedFan is showing for the Vcore etc are within acceptable tolerances.

The number that surprised me a lot was the voltage readings for the memory modules. In both SpeedFan and CPU-Z they are reading as running at 1.80V. Is this normal for DDR2 667?! It seems high, considering my 4Gb DDR3s 800s are only using max 1.65V, and on lower frequency 1.5V.

If it is wrong, why is it wrong and how do i fix it?? Like I said in my previous post the bios simply doesn't seem to have anywhere I can even see directly, let alone control the voltages. It is possible I may be able to update the bios, I am not really sure which version I have compared to the latest on Intels site (their last version dates back to 2010 which is pretty old).

If I am wrong and 1.80V is in fact fine for the DDR2 memory I am using, are there any other numbers here that look incorrect or out of place to anyone?

At the moment my only strategy is to wait and see if another problem occurs. For the moment nothing seems to have suffered any damage from these failed boots, although that may not continue of course.

Also when the PC is swtiched off there is a green LED lit on the Motherboard which according to the site means:

"The +5 V standby power indicator LED shows that power is still present even when the computer appears to be off"

Now I am not 100% sure about this, but I think that is normal no? It should be on? It also says in their documentation:

"For Instantly Available PC technology, the +5 V standby line from the power supply must be capable of providing adequate +5 V standby current. Failure to provide adequate standby current when implementing Instantly Available PC technology can damage the power supply. Instantly Available PC technology enables the board to enter the ACPI S3 (Suspend-toRAM) sleep-state. While in the S3 sleep-state, the computer will appear to be off (the power supply is off, and the front panel LED is amber if dual colored, or off if single colored.) When signaled by a wake-up device or event, the system quickly returns to its last known wake state. Table 7 on page 32 lists the devices and events that can wake the computer from the S3 state."

So I think that the green LED is simply saying there in enough power to provide IA PC Technology. But I am not sure whether it should be lit when the PC is completely powered off and not just in sleep or hibernate mode. Does it mean there could be a problem if it is lit when the PC is completely off? Could the residual power in the board or going to the RAM be corrupting the boot in some way?

On intels site for the DG31PR website it says "Max Memory Size (dependent on memory type) 4Gb" and then specifies "DDR2-800/677" as the memory types... it doesn't seem to explain what types effect the max size and how... I haven't simply put two things together that shouldn't be working together have I? Does CPU-Z show anything else out of the ordinary I can look at?

Thanks for taking an interest by the way... I am little bit lost here.

 

[ingtar33]

1.8v is pretty normal for ddr2 memory.

the green light just means there is some sort of power flowing into the system, it's working correctly.

 

[k@rt]

Well the PC worked fine all day yesterday... just waiting to see if anything else happens. 1.8V sounded high, but I trust you if you tell me that is fairly normal... apart from that all the other figures looked normal to me.

If anything else happens I will try putting back in the old PSU and see if that changes anything. May be one of those stupid things that never repeats. I did recheck all the connections inside the PC and was pretty embarrassed to notice that I mounted the CPU fan incorrectly, although I don't think that had anything to do with the problem.

Does seem a bit silly that the green LED is always lit when the PC is off, but if it is supposed to be then that is no problem. I thought maybe it was only meant to be lit when the PC was sleeping or hibernating, but not physically switched off.

Thanks again for your time and help... if anything else goes wrong I will update the thread, for now I will just wait and see.

 

[ingtar33]

ok. good luck!