Asus A7v133 onboard Temp register too high ?

[Snyk]

I have an Asus A7v133 (1007 bios), and using the onboard diode (flat, doesn't have a little popup tit type sensor) the CPU shows to run a bit hot.
CPU Is a 1GHZ AXIA, at 1ghz it runs 40+ idle, 45'ish under load (with chrome orb and that rinky dink ran pulled out of the center and an 80mm pushing air down it's throat)

I just put in a Cooler master HHC-001 (all copper with Heatpipes).. Clocked at 1280 it runs 45-46 idle under W2k, and 49-50 loaded..

Would almost appear that there is a 10C over compensation in temp.

Thought ?

just from feeling the heatsink while @ 50c it just can't be right, the copper right next to the socket is barely warm to the touch (50C should make the base fairly freakin hot that close to the core)

Thanks

 

[Kelvrick]

I heard the Asus boards read temps a little high. A guy told me this and I don't know how reliable his info is. I don't know whether its 10 degrees celsius or fehrenheit.

 

[anthrax]

my one read little be too hi..but that happened only with the A7V133...my new A7V333 reads fine...

 

[cbuchach]

On a website review a few months back (forget which website) I found an article in which they used they rigged up th thermal diode on the new XP processors using the SMB Bus on the A7V133 to get the ACTUAL processor temps (they used the A7V133 because it is one of the rather limited motherboards that has the SMBus). They found that the A7V133's onboard thermal monitoring was pretty close to the actual temp. It of course trailed any rise or fall in the actual core temp by a half minute or so but when at equilibrium the temps were within a degree celcius.

I know many have complained that the A7V133 reads temps high, but I think it is the case that the Asus engineers compensated for the inherent problems with off-die thermal monitoring. That is, they probably bumped up the temps a bit to more closely approximate the actual die temp of the processor.

Also, it is my opinion that you really can't compare core temps between different models of motherboards let alone different motherboards of the same model. Small things like distance or air flow aberations across the off-die thermal monitoring can alter results.

 

[Mikewarrior2]

cbuchach,

if you're referring to the void yourwarranty tests, the Asus boards still do not read temps all that well... it ranges from several C too high on 2 heatsinks, to matching on one, to 5-7C too low on the rest.

VoidyourWarranty Diode Versus Socket-thermsitor a7v133

So while the Asus is closer to actual DIE temp than many other boards, it often is not reading too high, and could even stand to use some more temp compensation 😀.


Mike

 

[cbuchach]

Yeah that was the website that I saw. OK so they were more than a degree off the actual temp. But still the fact is is that the A7V133 temp monitoring is closer to the real temp than would be boards on which the temperature was read to be lower.

 

[AA0]

Asus boards have always read temps high. The adjust their temps to reflect the core temp, not the temps that are read by a probe outside the cpu. While my temps say 50C for my XP chip, the HSF is barely warm, not much warmer then my k6-2, which runs 15C cooler.


temps are quite useless when they are probes, unless its an actual hardware monitor, don't trust it. Feel how hot the HSF is, if its warm, then you have an issue. Just set your temp max in the bios according to what it reads.

 

[Mikewarrior2]

Chubach,

I agree... I'm not saying that the Asus temps are "too high" or anything. they are teh closest to "die" temp out there. However, they still aren't necessarily accurate, per say. It does mean that other mb's that supposedly read "correct" temperatures (from those poeple spreading the internet myth that Asus boards read 10C too high), are really reading 10+C too low.


Mike