X5650 sounding temp alarm .

[Durvelle27]

No. Sorry. Comment was in response to two points made above. In post #21, David mentioned 10°C difference between two different TIMs. My point is, as long as the higher temp is still good, it does not matter. Lower temps does NOT automatically mean better. For example, there is nothing to suggest a CPU running at 40°C or 30°C will be more stable, perform better, or have a longer lifespan than a CPU running at 50°C. They are designed to run within a rather wide "operating temperature range". As long as it is comfortably within that range, all is good. All the 30°C really gets you is bragging rights.

The point I was making about the dirty pad was in response to post #17. If the TIM pad and cooler were properly applied in the first place, and if the cured bond was never some how broken, there is no reason for the TIM to get dirty. Consequently, memory channels would not somehow mysteriously be restored simply by replacing TIM. The memory channels were restored because the "dirty" TIM was replaced with "clean" TIM. TIM does not just get dirty and it does not go bad just because it gets old.

I note there is not one single TIM maker, CPU maker, GPU maker, cooler maker, computer maker or motherboard maker who claims TIM needs to be replaced just because it is X years old. Yes, it can lose a "few" (typically less than 5) degrees of efficiency over many years. But if a processor "needs" those 5° to keep from crossing the threshold into excessive thermal zones, then something else is wrong that needs to be addressed first - like a failing CPU fan, case cooling or something else. Not the TIM.

no not the TIM

The actual contacts on the CPU is what he referenced being dirty

Dirty contacts or misaligned contacts can indeed cause memory to not be seen or slots not to work

 

[Bill_Bright]

If that is what he meant, then my apologies and I agree. But I didn't read it that way since he was specifically talking about the OP who specifically said he "refitted the Heat sink and fan". There is no need to pull the CPU from the socket to refit/remount a CPU cooler so I don't see why the CPU would need to be pulled. But, the OP swapped CPUs with a spare and then started getting temp alarms - not memory problems.

That said, dirty CPU contacts should never happen. Period. If they get dirty, that is from careless handling/storage by the user. And if properly mounted, there is no way the pins/pads can be misaligned either. That again would be from carelessness. That said, I don't see how only "some" pads could be misaligned.

Either way, I think those points are moot since now we are talking about over temp conditions in a hot room.

 

[Solaris17]

memory channels would not somehow mysteriously be restored simply by replacing TIM

Actually yes they could, this was a big deal on x58 systems. Numerous threads on the web at the time for it. The issue specifically had to do with the amount of board flex incurred with the then new heavy tower coolers but mostly the over tightening caused by users along with manufacturers not having stronger backplates.

Alas that is education meant for a different thread. I am sure you would find the resources you need by googling however. Thankfully its generally an issue of the past on more modern systems.

 

[Bill_Bright]

Actually yes they could, this was a big deal on x58 systems. Numerous threads on the web at the time for it. The issue specifically had to do with the amount of board flex incurred with the then new heavy tower coolers but mostly the over tightening caused by users along with manufacturers not having stronger backplates.

No! Please read and take a moment to understand what I have been saying from the beginning. And note where I said, "simply by replacing TIM" and "if properly mounted".

In fact, you just verified what I said when you said,

The issue specifically had to do with the amount of board flex...over tightening caused by users

That clearly indicates the problem was caused by improper mounting by the user. Not because the old TIM needed to be replaced. And the problem did not go away because the TIM was replaced. The problem went away because the cooler was remounted properly this time.

As for your stronger backplate comment, I contend that was not the problem either. It was the PCB itself flexing, and that was due, in part, by the user (once again - the user) choosing too tall/heavy a cooler and/or over-tightening or tightening unevenly. Note monster coolers causing board flex would not be a problem with a desktop/horizontal case.

What probably needs to happen is the ATX Form Factor standard needs to be revised again to accommodate more motherboard mounting holes that would do a better job of evenly distributing the weight of big, heavy HSF assemblies hanging off motherboards when in a tower/vertical configuration. But again, that has nothing to do with the TIM - which was the whole point of all my comments.